Merge remote-tracking branch 'origin/master' into substitute-other-storedir

src/build-remote/build-remote.cc

@@ -33,7 +33,7 @@ std::string escapeUri(std::string uri)
 
 static string currentLoad;
 
-static AutoCloseFD openSlotLock(const Machine & m, unsigned long long slot)
+static AutoCloseFD openSlotLock(const Machine & m, uint64_t slot)
 {
     return openLockFile(fmt("%s/%s-%d", currentLoad, escapeUri(m.storeUri), slot), true);
 }
@@ -119,7 +119,7 @@ static int _main(int argc, char * * argv)
                 bool rightType = false;
 
                 Machine * bestMachine = nullptr;
-                unsigned long long bestLoad = 0;
+                uint64_t bestLoad = 0;
                 for (auto & m : machines) {
                     debug("considering building on remote machine '%s'", m.storeUri);
 
@@ -130,8 +130,8 @@ static int _main(int argc, char * * argv)
                         m.mandatoryMet(requiredFeatures)) {
                         rightType = true;
                         AutoCloseFD free;
-                        unsigned long long load = 0;
-                        for (unsigned long long slot = 0; slot < m.maxJobs; ++slot) {
+                        uint64_t load = 0;
+                        for (uint64_t slot = 0; slot < m.maxJobs; ++slot) {
                             auto slotLock = openSlotLock(m, slot);
                             if (lockFile(slotLock.get(), ltWrite, false)) {
                                 if (!free) {

src/libexpr/common-eval-args.cc

@@ -4,6 +4,8 @@
 #include "util.hh"
 #include "eval.hh"
 #include "fetchers.hh"
+#include "registry.hh"
+#include "flake/flakeref.hh"
 #include "store-api.hh"
 
 namespace nix {
@@ -31,6 +33,27 @@ MixEvalArgs::MixEvalArgs()
         .labels = {"path"},
         .handler = {[&](std::string s) { searchPath.push_back(s); }}
     });
+
+    addFlag({
+        .longName = "impure",
+        .description = "allow access to mutable paths and repositories",
+        .handler = {[&]() {
+            evalSettings.pureEval = false;
+        }},
+    });
+
+    addFlag({
+        .longName = "override-flake",
+        .description = "override a flake registry value",
+        .labels = {"original-ref", "resolved-ref"},
+        .handler = {[&](std::string _from, std::string _to) {
+            auto from = parseFlakeRef(_from, absPath("."));
+            auto to = parseFlakeRef(_to, absPath("."));
+            fetchers::Attrs extraAttrs;
+            if (to.subdir != "") extraAttrs["dir"] = to.subdir;
+            fetchers::overrideRegistry(from.input, to.input, extraAttrs);
+        }}
+    });
 }
 
 Bindings * MixEvalArgs::getAutoArgs(EvalState & state)
@@ -53,7 +76,7 @@ Path lookupFileArg(EvalState & state, string s)
     if (isUri(s)) {
         return state.store->toRealPath(
             fetchers::downloadTarball(
-                state.store, resolveUri(s), "source", false).storePath);
+                state.store, resolveUri(s), "source", false).first.storePath);
     } else if (s.size() > 2 && s.at(0) == '<' && s.at(s.size() - 1) == '>') {
         Path p = s.substr(1, s.size() - 2);
         return state.findFile(p);

src/libexpr/eval-cache.cc

@@ -0,0 +1,616 @@
+#include "eval-cache.hh"
+#include "sqlite.hh"
+#include "eval.hh"
+#include "eval-inline.hh"
+#include "store-api.hh"
+
+namespace nix::eval_cache {
+
+static const char * schema = R"sql(
+create table if not exists Attributes (
+    parent      integer not null,
+    name        text,
+    type        integer not null,
+    value       text,
+    context     text,
+    primary key (parent, name)
+);
+)sql";
+
+struct AttrDb
+{
+    std::atomic_bool failed{false};
+
+    struct State
+    {
+        SQLite db;
+        SQLiteStmt insertAttribute;
+        SQLiteStmt insertAttributeWithContext;
+        SQLiteStmt queryAttribute;
+        SQLiteStmt queryAttributes;
+        std::unique_ptr<SQLiteTxn> txn;
+    };
+
+    std::unique_ptr<Sync<State>> _state;
+
+    AttrDb(const Hash & fingerprint)
+        : _state(std::make_unique<Sync<State>>())
+    {
+        auto state(_state->lock());
+
+        Path cacheDir = getCacheDir() + "/nix/eval-cache-v2";
+        createDirs(cacheDir);
+
+        Path dbPath = cacheDir + "/" + fingerprint.to_string(Base16, false) + ".sqlite";
+
+        state->db = SQLite(dbPath);
+        state->db.isCache();
+        state->db.exec(schema);
+
+        state->insertAttribute.create(state->db,
+            "insert or replace into Attributes(parent, name, type, value) values (?, ?, ?, ?)");
+
+        state->insertAttributeWithContext.create(state->db,
+            "insert or replace into Attributes(parent, name, type, value, context) values (?, ?, ?, ?, ?)");
+
+        state->queryAttribute.create(state->db,
+            "select rowid, type, value, context from Attributes where parent = ? and name = ?");
+
+        state->queryAttributes.create(state->db,
+            "select name from Attributes where parent = ?");
+
+        state->txn = std::make_unique<SQLiteTxn>(state->db);
+    }
+
+    ~AttrDb()
+    {
+        try {
+            auto state(_state->lock());
+            if (!failed)
+                state->txn->commit();
+            state->txn.reset();
+        } catch (...) {
+            ignoreException();
+        }
+    }
+
+    template<typename F>
+    AttrId doSQLite(F && fun)
+    {
+        if (failed) return 0;
+        try {
+            return fun();
+        } catch (SQLiteError &) {
+            ignoreException();
+            failed = true;
+            return 0;
+        }
+    }
+
+    AttrId setAttrs(
+        AttrKey key,
+        const std::vector<Symbol> & attrs)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::FullAttrs)
+                (0, false).exec();
+
+            AttrId rowId = state->db.getLastInsertedRowId();
+            assert(rowId);
+
+            for (auto & attr : attrs)
+                state->insertAttribute.use()
+                    (rowId)
+                    (attr)
+                    (AttrType::Placeholder)
+                    (0, false).exec();
+
+            return rowId;
+        });
+    }
+
+    AttrId setString(
+        AttrKey key,
+        std::string_view s,
+        const char * * context = nullptr)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            if (context) {
+                std::string ctx;
+                for (const char * * p = context; *p; ++p) {
+                    if (p != context) ctx.push_back(' ');
+                    ctx.append(*p);
+                }
+                state->insertAttributeWithContext.use()
+                    (key.first)
+                    (key.second)
+                    (AttrType::String)
+                    (s)
+                    (ctx).exec();
+            } else {
+                state->insertAttribute.use()
+                    (key.first)
+                    (key.second)
+                    (AttrType::String)
+                (s).exec();
+            }
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    AttrId setBool(
+        AttrKey key,
+        bool b)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::Bool)
+                (b ? 1 : 0).exec();
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    AttrId setPlaceholder(AttrKey key)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::Placeholder)
+                (0, false).exec();
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    AttrId setMissing(AttrKey key)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::Missing)
+                (0, false).exec();
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    AttrId setMisc(AttrKey key)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::Misc)
+                (0, false).exec();
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    AttrId setFailed(AttrKey key)
+    {
+        return doSQLite([&]()
+        {
+            auto state(_state->lock());
+
+            state->insertAttribute.use()
+                (key.first)
+                (key.second)
+                (AttrType::Failed)
+                (0, false).exec();
+
+            return state->db.getLastInsertedRowId();
+        });
+    }
+
+    std::optional<std::pair<AttrId, AttrValue>> getAttr(
+        AttrKey key,
+        SymbolTable & symbols)
+    {
+        auto state(_state->lock());
+
+        auto queryAttribute(state->queryAttribute.use()(key.first)(key.second));
+        if (!queryAttribute.next()) return {};
+
+        auto rowId = (AttrType) queryAttribute.getInt(0);
+        auto type = (AttrType) queryAttribute.getInt(1);
+
+        switch (type) {
+            case AttrType::Placeholder:
+                return {{rowId, placeholder_t()}};
+            case AttrType::FullAttrs: {
+                // FIXME: expensive, should separate this out.
+                std::vector<Symbol> attrs;
+                auto queryAttributes(state->queryAttributes.use()(rowId));
+                while (queryAttributes.next())
+                    attrs.push_back(symbols.create(queryAttributes.getStr(0)));
+                return {{rowId, attrs}};
+            }
+            case AttrType::String: {
+                std::vector<std::pair<Path, std::string>> context;
+                if (!queryAttribute.isNull(3))
+                    for (auto & s : tokenizeString<std::vector<std::string>>(queryAttribute.getStr(3), ";"))
+                        context.push_back(decodeContext(s));
+                return {{rowId, string_t{queryAttribute.getStr(2), context}}};
+            }
+            case AttrType::Bool:
+                return {{rowId, queryAttribute.getInt(2) != 0}};
+            case AttrType::Missing:
+                return {{rowId, missing_t()}};
+            case AttrType::Misc:
+                return {{rowId, misc_t()}};
+            case AttrType::Failed:
+                return {{rowId, failed_t()}};
+            default:
+                throw Error("unexpected type in evaluation cache");
+        }
+    }
+};
+
+static std::shared_ptr<AttrDb> makeAttrDb(const Hash & fingerprint)
+{
+    try {
+        return std::make_shared<AttrDb>(fingerprint);
+    } catch (SQLiteError &) {
+        ignoreException();
+        return nullptr;
+    }
+}
+
+EvalCache::EvalCache(
+    std::optional<std::reference_wrapper<const Hash>> useCache,
+    EvalState & state,
+    RootLoader rootLoader)
+    : db(useCache ? makeAttrDb(*useCache) : nullptr)
+    , state(state)
+    , rootLoader(rootLoader)
+{
+}
+
+Value * EvalCache::getRootValue()
+{
+    if (!value) {
+        debug("getting root value");
+        value = allocRootValue(rootLoader());
+    }
+    return *value;
+}
+
+std::shared_ptr<AttrCursor> EvalCache::getRoot()
+{
+    return std::make_shared<AttrCursor>(ref(shared_from_this()), std::nullopt);
+}
+
+AttrCursor::AttrCursor(
+    ref<EvalCache> root,
+    Parent parent,
+    Value * value,
+    std::optional<std::pair<AttrId, AttrValue>> && cachedValue)
+    : root(root), parent(parent), cachedValue(std::move(cachedValue))
+{
+    if (value)
+        _value = allocRootValue(value);
+}
+
+AttrKey AttrCursor::getKey()
+{
+    if (!parent)
+        return {0, root->state.sEpsilon};
+    if (!parent->first->cachedValue) {
+        parent->first->cachedValue = root->db->getAttr(
+            parent->first->getKey(), root->state.symbols);
+        assert(parent->first->cachedValue);
+    }
+    return {parent->first->cachedValue->first, parent->second};
+}
+
+Value & AttrCursor::getValue()
+{
+    if (!_value) {
+        if (parent) {
+            auto & vParent = parent->first->getValue();
+            root->state.forceAttrs(vParent);
+            auto attr = vParent.attrs->get(parent->second);
+            if (!attr)
+                throw Error("attribute '%s' is unexpectedly missing", getAttrPathStr());
+            _value = allocRootValue(attr->value);
+        } else
+            _value = allocRootValue(root->getRootValue());
+    }
+    return **_value;
+}
+
+std::vector<Symbol> AttrCursor::getAttrPath() const
+{
+    if (parent) {
+        auto attrPath = parent->first->getAttrPath();
+        attrPath.push_back(parent->second);
+        return attrPath;
+    } else
+        return {};
+}
+
+std::vector<Symbol> AttrCursor::getAttrPath(Symbol name) const
+{
+    auto attrPath = getAttrPath();
+    attrPath.push_back(name);
+    return attrPath;
+}
+
+std::string AttrCursor::getAttrPathStr() const
+{
+    return concatStringsSep(".", getAttrPath());
+}
+
+std::string AttrCursor::getAttrPathStr(Symbol name) const
+{
+    return concatStringsSep(".", getAttrPath(name));
+}
+
+Value & AttrCursor::forceValue()
+{
+    debug("evaluating uncached attribute %s", getAttrPathStr());
+
+    auto & v = getValue();
+
+    try {
+        root->state.forceValue(v);
+    } catch (EvalError &) {
+        debug("setting '%s' to failed", getAttrPathStr());
+        if (root->db)
+            cachedValue = {root->db->setFailed(getKey()), failed_t()};
+        throw;
+    }
+
+    if (root->db && (!cachedValue || std::get_if<placeholder_t>(&cachedValue->second))) {
+        if (v.type == tString)
+            cachedValue = {root->db->setString(getKey(), v.string.s, v.string.context), v.string.s};
+        else if (v.type == tPath)
+            cachedValue = {root->db->setString(getKey(), v.path), v.path};
+        else if (v.type == tBool)
+            cachedValue = {root->db->setBool(getKey(), v.boolean), v.boolean};
+        else if (v.type == tAttrs)
+            ; // FIXME: do something?
+        else
+            cachedValue = {root->db->setMisc(getKey()), misc_t()};
+    }
+
+    return v;
+}
+
+std::shared_ptr<AttrCursor> AttrCursor::maybeGetAttr(Symbol name)
+{
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = root->db->getAttr(getKey(), root->state.symbols);
+
+        if (cachedValue) {
+            if (auto attrs = std::get_if<std::vector<Symbol>>(&cachedValue->second)) {
+                for (auto & attr : *attrs)
+                    if (attr == name)
+                        return std::make_shared<AttrCursor>(root, std::make_pair(shared_from_this(), name));
+                return nullptr;
+            } else if (std::get_if<placeholder_t>(&cachedValue->second)) {
+                auto attr = root->db->getAttr({cachedValue->first, name}, root->state.symbols);
+                if (attr) {
+                    if (std::get_if<missing_t>(&attr->second))
+                        return nullptr;
+                    else if (std::get_if<failed_t>(&attr->second))
+                        throw EvalError("cached failure of attribute '%s'", getAttrPathStr(name));
+                    else
+                        return std::make_shared<AttrCursor>(root,
+                            std::make_pair(shared_from_this(), name), nullptr, std::move(attr));
+                }
+                // Incomplete attrset, so need to fall thru and
+                // evaluate to see whether 'name' exists
+            } else
+                return nullptr;
+                //throw TypeError("'%s' is not an attribute set", getAttrPathStr());
+        }
+    }
+
+    auto & v = forceValue();
+
+    if (v.type != tAttrs)
+        return nullptr;
+        //throw TypeError("'%s' is not an attribute set", getAttrPathStr());
+
+    auto attr = v.attrs->get(name);
+
+    if (!attr) {
+        if (root->db) {
+            if (!cachedValue)
+                cachedValue = {root->db->setPlaceholder(getKey()), placeholder_t()};
+            root->db->setMissing({cachedValue->first, name});
+        }
+        return nullptr;
+    }
+
+    std::optional<std::pair<AttrId, AttrValue>> cachedValue2;
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = {root->db->setPlaceholder(getKey()), placeholder_t()};
+        cachedValue2 = {root->db->setPlaceholder({cachedValue->first, name}), placeholder_t()};
+    }
+
+    return std::make_shared<AttrCursor>(
+        root, std::make_pair(shared_from_this(), name), attr->value, std::move(cachedValue2));
+}
+
+std::shared_ptr<AttrCursor> AttrCursor::maybeGetAttr(std::string_view name)
+{
+    return maybeGetAttr(root->state.symbols.create(name));
+}
+
+std::shared_ptr<AttrCursor> AttrCursor::getAttr(Symbol name)
+{
+    auto p = maybeGetAttr(name);
+    if (!p)
+        throw Error("attribute '%s' does not exist", getAttrPathStr(name));
+    return p;
+}
+
+std::shared_ptr<AttrCursor> AttrCursor::getAttr(std::string_view name)
+{
+    return getAttr(root->state.symbols.create(name));
+}
+
+std::shared_ptr<AttrCursor> AttrCursor::findAlongAttrPath(const std::vector<Symbol> & attrPath)
+{
+    auto res = shared_from_this();
+    for (auto & attr : attrPath) {
+        res = res->maybeGetAttr(attr);
+        if (!res) return {};
+    }
+    return res;
+}
+
+std::string AttrCursor::getString()
+{
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = root->db->getAttr(getKey(), root->state.symbols);
+        if (cachedValue && !std::get_if<placeholder_t>(&cachedValue->second)) {
+            if (auto s = std::get_if<string_t>(&cachedValue->second)) {
+                debug("using cached string attribute '%s'", getAttrPathStr());
+                return s->first;
+            } else
+                throw TypeError("'%s' is not a string", getAttrPathStr());
+        }
+    }
+
+    auto & v = forceValue();
+
+    if (v.type != tString && v.type != tPath)
+        throw TypeError("'%s' is not a string but %s", getAttrPathStr(), showType(v.type));
+
+    return v.type == tString ? v.string.s : v.path;
+}
+
+string_t AttrCursor::getStringWithContext()
+{
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = root->db->getAttr(getKey(), root->state.symbols);
+        if (cachedValue && !std::get_if<placeholder_t>(&cachedValue->second)) {
+            if (auto s = std::get_if<string_t>(&cachedValue->second)) {
+                debug("using cached string attribute '%s'", getAttrPathStr());
+                return *s;
+            } else
+                throw TypeError("'%s' is not a string", getAttrPathStr());
+        }
+    }
+
+    auto & v = forceValue();
+
+    if (v.type == tString)
+        return {v.string.s, v.getContext()};
+    else if (v.type == tPath)
+        return {v.path, {}};
+    else
+        throw TypeError("'%s' is not a string but %s", getAttrPathStr(), showType(v.type));
+}
+
+bool AttrCursor::getBool()
+{
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = root->db->getAttr(getKey(), root->state.symbols);
+        if (cachedValue && !std::get_if<placeholder_t>(&cachedValue->second)) {
+            if (auto b = std::get_if<bool>(&cachedValue->second)) {
+                debug("using cached Boolean attribute '%s'", getAttrPathStr());
+                return *b;
+            } else
+                throw TypeError("'%s' is not a Boolean", getAttrPathStr());
+        }
+    }
+
+    auto & v = forceValue();
+
+    if (v.type != tBool)
+        throw TypeError("'%s' is not a Boolean", getAttrPathStr());
+
+    return v.boolean;
+}
+
+std::vector<Symbol> AttrCursor::getAttrs()
+{
+    if (root->db) {
+        if (!cachedValue)
+            cachedValue = root->db->getAttr(getKey(), root->state.symbols);
+        if (cachedValue && !std::get_if<placeholder_t>(&cachedValue->second)) {
+            if (auto attrs = std::get_if<std::vector<Symbol>>(&cachedValue->second)) {
+                debug("using cached attrset attribute '%s'", getAttrPathStr());
+                return *attrs;
+            } else
+                throw TypeError("'%s' is not an attribute set", getAttrPathStr());
+        }
+    }
+
+    auto & v = forceValue();
+
+    if (v.type != tAttrs)
+        throw TypeError("'%s' is not an attribute set", getAttrPathStr());
+
+    std::vector<Symbol> attrs;
+    for (auto & attr : *getValue().attrs)
+        attrs.push_back(attr.name);
+    std::sort(attrs.begin(), attrs.end(), [](const Symbol & a, const Symbol & b) {
+        return (const string &) a < (const string &) b;
+    });
+
+    if (root->db)
+        cachedValue = {root->db->setAttrs(getKey(), attrs), attrs};
+
+    return attrs;
+}
+
+bool AttrCursor::isDerivation()
+{
+    auto aType = maybeGetAttr("type");
+    return aType && aType->getString() == "derivation";
+}
+
+StorePath AttrCursor::forceDerivation()
+{
+    auto aDrvPath = getAttr(root->state.sDrvPath);
+    auto drvPath = root->state.store->parseStorePath(aDrvPath->getString());
+    if (!root->state.store->isValidPath(drvPath) && !settings.readOnlyMode) {
+        /* The eval cache contains 'drvPath', but the actual path has
+           been garbage-collected. So force it to be regenerated. */
+        aDrvPath->forceValue();
+        if (!root->state.store->isValidPath(drvPath))
+            throw Error("don't know how to recreate store derivation '%s'!",
+                root->state.store->printStorePath(drvPath));
+    }
+    return drvPath;
+}
+
+}

src/libexpr/eval-cache.hh

@@ -0,0 +1,121 @@
+#pragma once
+
+#include "sync.hh"
+#include "hash.hh"
+#include "eval.hh"
+
+#include <functional>
+#include <variant>
+
+namespace nix::eval_cache {
+
+class AttrDb;
+class AttrCursor;
+
+class EvalCache : public std::enable_shared_from_this<EvalCache>
+{
+    friend class AttrCursor;
+
+    std::shared_ptr<AttrDb> db;
+    EvalState & state;
+    typedef std::function<Value *()> RootLoader;
+    RootLoader rootLoader;
+    RootValue value;
+
+    Value * getRootValue();
+
+public:
+
+    EvalCache(
+        std::optional<std::reference_wrapper<const Hash>> useCache,
+        EvalState & state,
+        RootLoader rootLoader);
+
+    std::shared_ptr<AttrCursor> getRoot();
+};
+
+enum AttrType {
+    Placeholder = 0,
+    FullAttrs = 1,
+    String = 2,
+    Missing = 3,
+    Misc = 4,
+    Failed = 5,
+    Bool = 6,
+};
+
+struct placeholder_t {};
+struct missing_t {};
+struct misc_t {};
+struct failed_t {};
+typedef uint64_t AttrId;
+typedef std::pair<AttrId, Symbol> AttrKey;
+typedef std::pair<std::string, std::vector<std::pair<Path, std::string>>> string_t;
+
+typedef std::variant<
+    std::vector<Symbol>,
+    string_t,
+    placeholder_t,
+    missing_t,
+    misc_t,
+    failed_t,
+    bool
+    > AttrValue;
+
+class AttrCursor : public std::enable_shared_from_this<AttrCursor>
+{
+    friend class EvalCache;
+
+    ref<EvalCache> root;
+    typedef std::optional<std::pair<std::shared_ptr<AttrCursor>, Symbol>> Parent;
+    Parent parent;
+    RootValue _value;
+    std::optional<std::pair<AttrId, AttrValue>> cachedValue;
+
+    AttrKey getKey();
+
+    Value & getValue();
+
+public:
+
+    AttrCursor(
+        ref<EvalCache> root,
+        Parent parent,
+        Value * value = nullptr,
+        std::optional<std::pair<AttrId, AttrValue>> && cachedValue = {});
+
+    std::vector<Symbol> getAttrPath() const;
+
+    std::vector<Symbol> getAttrPath(Symbol name) const;
+
+    std::string getAttrPathStr() const;
+
+    std::string getAttrPathStr(Symbol name) const;
+
+    std::shared_ptr<AttrCursor> maybeGetAttr(Symbol name);
+
+    std::shared_ptr<AttrCursor> maybeGetAttr(std::string_view name);
+
+    std::shared_ptr<AttrCursor> getAttr(Symbol name);
+
+    std::shared_ptr<AttrCursor> getAttr(std::string_view name);
+
+    std::shared_ptr<AttrCursor> findAlongAttrPath(const std::vector<Symbol> & attrPath);
+
+    std::string getString();
+
+    string_t getStringWithContext();
+
+    bool getBool();
+
+    std::vector<Symbol> getAttrs();
+
+    bool isDerivation();
+
+    Value & forceValue();
+
+    /* Force creation of the .drv file in the Nix store. */
+    StorePath forceDerivation();
+};
+
+}

src/libexpr/eval.cc

@@ -199,6 +199,18 @@ string showType(const Value & v)
 }
 
 
+bool Value::isTrivial() const
+{
+    return
+        type != tApp
+        && type != tPrimOpApp
+        && (type != tThunk
+            || (dynamic_cast<ExprAttrs *>(thunk.expr)
+                && ((ExprAttrs *) thunk.expr)->dynamicAttrs.empty())
+            || dynamic_cast<ExprLambda *>(thunk.expr));
+}
+
+
 #if HAVE_BOEHMGC
 /* Called when the Boehm GC runs out of memory. */
 static void * oomHandler(size_t requested)
@@ -337,6 +349,9 @@ EvalState::EvalState(const Strings & _searchPath, ref<Store> store)
     , sOutputHashAlgo(symbols.create("outputHashAlgo"))
     , sOutputHashMode(symbols.create("outputHashMode"))
     , sRecurseForDerivations(symbols.create("recurseForDerivations"))
+    , sDescription(symbols.create("description"))
+    , sSelf(symbols.create("self"))
+    , sEpsilon(symbols.create(""))
     , repair(NoRepair)
     , store(store)
     , baseEnv(allocEnv(128))
@@ -366,7 +381,7 @@ EvalState::EvalState(const Strings & _searchPath, ref<Store> store)
 
             if (store->isInStore(r.second)) {
                 StorePathSet closure;
-                store->computeFSClosure(store->parseStorePath(store->toStorePath(r.second)), closure);
+                store->computeFSClosure(store->toStorePath(r.second).first, closure);
                 for (auto & path : closure)
                     allowedPaths->insert(store->printStorePath(path));
             } else
@@ -782,7 +797,7 @@ Value * ExprPath::maybeThunk(EvalState & state, Env & env)
 }
 
 
-void EvalState::evalFile(const Path & path_, Value & v)
+void EvalState::evalFile(const Path & path_, Value & v, bool mustBeTrivial)
 {
     auto path = checkSourcePath(path_);
 
@@ -811,6 +826,11 @@ void EvalState::evalFile(const Path & path_, Value & v)
     fileParseCache[path2] = e;
 
     try {
+        // Enforce that 'flake.nix' is a direct attrset, not a
+        // computation.
+        if (mustBeTrivial &&
+            !(dynamic_cast<ExprAttrs *>(e)))
+            throw Error("file '%s' must be an attribute set", path);
         eval(e, v);
     } catch (Error & e) {
         addErrorTrace(e, "while evaluating the file '%1%':", path2);
@@ -1586,6 +1606,18 @@ string EvalState::forceString(Value & v, const Pos & pos)
 }
 
 
+/* Decode a context string ‘!<name>!<path>’ into a pair <path,
+   name>. */
+std::pair<string, string> decodeContext(std::string_view s)
+{
+    if (s.at(0) == '!') {
+        size_t index = s.find("!", 1);
+        return {std::string(s.substr(index + 1)), std::string(s.substr(1, index - 1))};
+    } else
+        return {s.at(0) == '/' ? std::string(s) : std::string(s.substr(1)), ""};
+}
+
+
 void copyContext(const Value & v, PathSet & context)
 {
     if (v.string.context)
@@ -1594,6 +1626,17 @@ void copyContext(const Value & v, PathSet & context)
 }
 
 
+std::vector<std::pair<Path, std::string>> Value::getContext()
+{
+    std::vector<std::pair<Path, std::string>> res;
+    assert(type == tString);
+    if (string.context)
+        for (const char * * p = string.context; *p; ++p)
+            res.push_back(decodeContext(*p));
+    return res;
+}
+
+
 string EvalState::forceString(Value & v, PathSet & context, const Pos & pos)
 {
     string s = forceString(v, pos);

src/libexpr/eval.hh

@@ -4,13 +4,13 @@
 #include "value.hh"
 #include "nixexpr.hh"
 #include "symbol-table.hh"
-#include "hash.hh"
 #include "config.hh"
 
 #include <regex>
 #include <map>
 #include <optional>
 #include <unordered_map>
+#include <mutex>
 
 
 namespace nix {
@@ -75,7 +75,8 @@ public:
         sFile, sLine, sColumn, sFunctor, sToString,
         sRight, sWrong, sStructuredAttrs, sBuilder, sArgs,
         sOutputHash, sOutputHashAlgo, sOutputHashMode,
-        sRecurseForDerivations;
+        sRecurseForDerivations,
+        sDescription, sSelf, sEpsilon;
     Symbol sDerivationNix;
 
     /* If set, force copying files to the Nix store even if they
@@ -90,6 +91,7 @@ public:
 
     const ref<Store> store;
 
+
 private:
     SrcToStore srcToStore;
 
@@ -152,8 +154,9 @@ public:
     Expr * parseStdin();
 
     /* Evaluate an expression read from the given file to normal
-       form. */
-    void evalFile(const Path & path, Value & v);
+       form. Optionally enforce that the top-level expression is
+       trivial (i.e. doesn't require arbitrary computation). */
+    void evalFile(const Path & path, Value & v, bool mustBeTrivial = false);
 
     void resetFileCache();
 
@@ -330,7 +333,7 @@ string showType(const Value & v);
 
 /* Decode a context string ‘!<name>!<path>’ into a pair <path,
    name>. */
-std::pair<string, string> decodeContext(const string & s);
+std::pair<string, string> decodeContext(std::string_view s);
 
 /* If `path' refers to a directory, then append "/default.nix". */
 Path resolveExprPath(Path path);

src/libexpr/flake/call-flake.nix

@@ -0,0 +1,56 @@
+lockFileStr: rootSrc: rootSubdir:
+
+let
+
+  lockFile = builtins.fromJSON lockFileStr;
+
+  allNodes =
+    builtins.mapAttrs
+      (key: node:
+        let
+
+          sourceInfo =
+            if key == lockFile.root
+            then rootSrc
+            else fetchTree (node.info or {} // removeAttrs node.locked ["dir"]);
+
+          subdir = if key == lockFile.root then rootSubdir else node.locked.dir or "";
+
+          flake = import (sourceInfo + (if subdir != "" then "/" else "") + subdir + "/flake.nix");
+
+          inputs = builtins.mapAttrs
+            (inputName: inputSpec: allNodes.${resolveInput inputSpec})
+            (node.inputs or {});
+
+          # Resolve a input spec into a node name. An input spec is
+          # either a node name, or a 'follows' path from the root
+          # node.
+          resolveInput = inputSpec:
+              if builtins.isList inputSpec
+              then getInputByPath lockFile.root inputSpec
+              else inputSpec;
+
+          # Follow an input path (e.g. ["dwarffs" "nixpkgs"]) from the
+          # root node, returning the final node.
+          getInputByPath = nodeName: path:
+            if path == []
+            then nodeName
+            else
+              getInputByPath
+                # Since this could be a 'follows' input, call resolveInput.
+                (resolveInput lockFile.nodes.${nodeName}.inputs.${builtins.head path})
+                (builtins.tail path);
+
+          outputs = flake.outputs (inputs // { self = result; });
+
+          result = outputs // sourceInfo // { inherit inputs; inherit outputs; inherit sourceInfo; };
+        in
+          if node.flake or true then
+            assert builtins.isFunction flake.outputs;
+            result
+          else
+            sourceInfo
+      )
+      lockFile.nodes;
+
+in allNodes.${lockFile.root}

src/libexpr/flake/flake.cc

@@ -0,0 +1,609 @@
+#include "flake.hh"
+#include "lockfile.hh"
+#include "primops.hh"
+#include "eval-inline.hh"
+#include "store-api.hh"
+#include "fetchers.hh"
+#include "finally.hh"
+
+namespace nix {
+
+using namespace flake;
+
+namespace flake {
+
+typedef std::pair<Tree, FlakeRef> FetchedFlake;
+typedef std::vector<std::pair<FlakeRef, FetchedFlake>> FlakeCache;
+
+static std::optional<FetchedFlake> lookupInFlakeCache(
+    const FlakeCache & flakeCache,
+    const FlakeRef & flakeRef)
+{
+    // FIXME: inefficient.
+    for (auto & i : flakeCache) {
+        if (flakeRef == i.first) {
+            debug("mapping '%s' to previously seen input '%s' -> '%s",
+                flakeRef, i.first, i.second.second);
+            return i.second;
+        }
+    }
+
+    return std::nullopt;
+}
+
+static std::tuple<fetchers::Tree, FlakeRef, FlakeRef> fetchOrSubstituteTree(
+    EvalState & state,
+    const FlakeRef & originalRef,
+    bool allowLookup,
+    FlakeCache & flakeCache)
+{
+    auto fetched = lookupInFlakeCache(flakeCache, originalRef);
+    FlakeRef resolvedRef = originalRef;
+
+    if (!fetched) {
+        if (originalRef.input.isDirect()) {
+            fetched.emplace(originalRef.fetchTree(state.store));
+        } else {
+            if (allowLookup) {
+                resolvedRef = originalRef.resolve(state.store);
+                auto fetchedResolved = lookupInFlakeCache(flakeCache, originalRef);
+                if (!fetchedResolved) fetchedResolved.emplace(resolvedRef.fetchTree(state.store));
+                flakeCache.push_back({resolvedRef, fetchedResolved.value()});
+                fetched.emplace(fetchedResolved.value());
+            }
+            else {
+                throw Error("'%s' is an indirect flake reference, but registry lookups are not allowed", originalRef);
+            }
+        }
+        flakeCache.push_back({originalRef, fetched.value()});
+    }
+
+    auto [tree, lockedRef] = fetched.value();
+
+    debug("got tree '%s' from '%s'",
+        state.store->printStorePath(tree.storePath), lockedRef);
+
+    if (state.allowedPaths)
+        state.allowedPaths->insert(tree.actualPath);
+
+    assert(!originalRef.input.getNarHash() || tree.storePath == originalRef.input.computeStorePath(*state.store));
+
+    return {std::move(tree), resolvedRef, lockedRef};
+}
+
+static void expectType(EvalState & state, ValueType type,
+    Value & value, const Pos & pos)
+{
+    if (value.type == tThunk && value.isTrivial())
+        state.forceValue(value, pos);
+    if (value.type != type)
+        throw Error("expected %s but got %s at %s",
+            showType(type), showType(value.type), pos);
+}
+
+static std::map<FlakeId, FlakeInput> parseFlakeInputs(
+    EvalState & state, Value * value, const Pos & pos);
+
+static FlakeInput parseFlakeInput(EvalState & state,
+    const std::string & inputName, Value * value, const Pos & pos)
+{
+    expectType(state, tAttrs, *value, pos);
+
+    FlakeInput input;
+
+    auto sInputs = state.symbols.create("inputs");
+    auto sUrl = state.symbols.create("url");
+    auto sFlake = state.symbols.create("flake");
+    auto sFollows = state.symbols.create("follows");
+
+    fetchers::Attrs attrs;
+    std::optional<std::string> url;
+
+    for (nix::Attr attr : *(value->attrs)) {
+        try {
+            if (attr.name == sUrl) {
+                expectType(state, tString, *attr.value, *attr.pos);
+                url = attr.value->string.s;
+                attrs.emplace("url", *url);
+            } else if (attr.name == sFlake) {
+                expectType(state, tBool, *attr.value, *attr.pos);
+                input.isFlake = attr.value->boolean;
+            } else if (attr.name == sInputs) {
+                input.overrides = parseFlakeInputs(state, attr.value, *attr.pos);
+            } else if (attr.name == sFollows) {
+                expectType(state, tString, *attr.value, *attr.pos);
+                input.follows = parseInputPath(attr.value->string.s);
+            } else {
+                state.forceValue(*attr.value);
+                if (attr.value->type == tString)
+                    attrs.emplace(attr.name, attr.value->string.s);
+                else
+                    throw TypeError("flake input attribute '%s' is %s while a string is expected",
+                        attr.name, showType(*attr.value));
+            }
+        } catch (Error & e) {
+            e.addTrace(*attr.pos, hintfmt("in flake attribute '%s'", attr.name));
+            throw;
+        }
+    }
+
+    if (attrs.count("type"))
+        try {
+            input.ref = FlakeRef::fromAttrs(attrs);
+        } catch (Error & e) {
+            e.addTrace(pos, hintfmt("in flake input"));
+            throw;
+        }
+    else {
+        attrs.erase("url");
+        if (!attrs.empty())
+            throw Error("unexpected flake input attribute '%s', at %s", attrs.begin()->first, pos);
+        if (url)
+            input.ref = parseFlakeRef(*url, {}, true);
+    }
+
+    if (!input.follows && !input.ref)
+        input.ref = FlakeRef::fromAttrs({{"type", "indirect"}, {"id", inputName}});
+
+    return input;
+}
+
+static std::map<FlakeId, FlakeInput> parseFlakeInputs(
+    EvalState & state, Value * value, const Pos & pos)
+{
+    std::map<FlakeId, FlakeInput> inputs;
+
+    expectType(state, tAttrs, *value, pos);
+
+    for (nix::Attr & inputAttr : *(*value).attrs) {
+        inputs.emplace(inputAttr.name,
+            parseFlakeInput(state,
+                inputAttr.name,
+                inputAttr.value,
+                *inputAttr.pos));
+    }
+
+    return inputs;
+}
+
+static Flake getFlake(
+    EvalState & state,
+    const FlakeRef & originalRef,
+    bool allowLookup,
+    FlakeCache & flakeCache)
+{
+    auto [sourceInfo, resolvedRef, lockedRef] = fetchOrSubstituteTree(
+        state, originalRef, allowLookup, flakeCache);
+
+    // Guard against symlink attacks.
+    auto flakeFile = canonPath(sourceInfo.actualPath + "/" + lockedRef.subdir + "/flake.nix");
+    if (!isInDir(flakeFile, sourceInfo.actualPath))
+        throw Error("'flake.nix' file of flake '%s' escapes from '%s'",
+            lockedRef, state.store->printStorePath(sourceInfo.storePath));
+
+    Flake flake {
+        .originalRef = originalRef,
+        .resolvedRef = resolvedRef,
+        .lockedRef = lockedRef,
+        .sourceInfo = std::make_shared<fetchers::Tree>(std::move(sourceInfo))
+    };
+
+    if (!pathExists(flakeFile))
+        throw Error("source tree referenced by '%s' does not contain a '%s/flake.nix' file", lockedRef, lockedRef.subdir);
+
+    Value vInfo;
+    state.evalFile(flakeFile, vInfo, true); // FIXME: symlink attack
+
+    expectType(state, tAttrs, vInfo, Pos(foFile, state.symbols.create(flakeFile), 0, 0));
+
+    auto sEdition = state.symbols.create("edition"); // FIXME: remove soon
+
+    if (vInfo.attrs->get(sEdition))
+        warn("flake '%s' has deprecated attribute 'edition'", lockedRef);
+
+    if (auto description = vInfo.attrs->get(state.sDescription)) {
+        expectType(state, tString, *description->value, *description->pos);
+        flake.description = description->value->string.s;
+    }
+
+    auto sInputs = state.symbols.create("inputs");
+
+    if (auto inputs = vInfo.attrs->get(sInputs))
+        flake.inputs = parseFlakeInputs(state, inputs->value, *inputs->pos);
+
+    auto sOutputs = state.symbols.create("outputs");
+
+    if (auto outputs = vInfo.attrs->get(sOutputs)) {
+        expectType(state, tLambda, *outputs->value, *outputs->pos);
+        flake.vOutputs = allocRootValue(outputs->value);
+
+        if ((*flake.vOutputs)->lambda.fun->matchAttrs) {
+            for (auto & formal : (*flake.vOutputs)->lambda.fun->formals->formals) {
+                if (formal.name != state.sSelf)
+                    flake.inputs.emplace(formal.name, FlakeInput {
+                        .ref = parseFlakeRef(formal.name)
+                    });
+            }
+        }
+
+    } else
+        throw Error("flake '%s' lacks attribute 'outputs'", lockedRef);
+
+    for (auto & attr : *vInfo.attrs) {
+        if (attr.name != sEdition &&
+            attr.name != state.sDescription &&
+            attr.name != sInputs &&
+            attr.name != sOutputs)
+            throw Error("flake '%s' has an unsupported attribute '%s', at %s",
+                lockedRef, attr.name, *attr.pos);
+    }
+
+    return flake;
+}
+
+Flake getFlake(EvalState & state, const FlakeRef & originalRef, bool allowLookup)
+{
+    FlakeCache flakeCache;
+    return getFlake(state, originalRef, allowLookup, flakeCache);
+}
+
+/* Compute an in-memory lock file for the specified top-level flake,
+   and optionally write it to file, it the flake is writable. */
+LockedFlake lockFlake(
+    EvalState & state,
+    const FlakeRef & topRef,
+    const LockFlags & lockFlags)
+{
+    settings.requireExperimentalFeature("flakes");
+
+    FlakeCache flakeCache;
+
+    auto flake = getFlake(state, topRef, lockFlags.useRegistries, flakeCache);
+
+    // FIXME: symlink attack
+    auto oldLockFile = LockFile::read(
+        flake.sourceInfo->actualPath + "/" + flake.lockedRef.subdir + "/flake.lock");
+
+    debug("old lock file: %s", oldLockFile);
+
+    // FIXME: check whether all overrides are used.
+    std::map<InputPath, FlakeInput> overrides;
+    std::set<InputPath> overridesUsed, updatesUsed;
+
+    for (auto & i : lockFlags.inputOverrides)
+        overrides.insert_or_assign(i.first, FlakeInput { .ref = i.second });
+
+    LockFile newLockFile;
+
+    std::vector<FlakeRef> parents;
+
+    std::function<void(
+        const FlakeInputs & flakeInputs,
+        std::shared_ptr<Node> node,
+        const InputPath & inputPathPrefix,
+        std::shared_ptr<const Node> oldNode)>
+        computeLocks;
+
+    computeLocks = [&](
+        const FlakeInputs & flakeInputs,
+        std::shared_ptr<Node> node,
+        const InputPath & inputPathPrefix,
+        std::shared_ptr<const Node> oldNode)
+    {
+        debug("computing lock file node '%s'", printInputPath(inputPathPrefix));
+
+        /* Get the overrides (i.e. attributes of the form
+           'inputs.nixops.inputs.nixpkgs.url = ...'). */
+        // FIXME: check this
+        for (auto & [id, input] : flake.inputs) {
+            for (auto & [idOverride, inputOverride] : input.overrides) {
+                auto inputPath(inputPathPrefix);
+                inputPath.push_back(id);
+                inputPath.push_back(idOverride);
+                overrides.insert_or_assign(inputPath, inputOverride);
+            }
+        }
+
+        /* Go over the flake inputs, resolve/fetch them if
+           necessary (i.e. if they're new or the flakeref changed
+           from what's in the lock file). */
+        for (auto & [id, input2] : flakeInputs) {
+            auto inputPath(inputPathPrefix);
+            inputPath.push_back(id);
+            auto inputPathS = printInputPath(inputPath);
+            debug("computing input '%s'", inputPathS);
+
+            /* Do we have an override for this input from one of the
+               ancestors? */
+            auto i = overrides.find(inputPath);
+            bool hasOverride = i != overrides.end();
+            if (hasOverride) overridesUsed.insert(inputPath);
+            auto & input = hasOverride ? i->second : input2;
+
+            /* Resolve 'follows' later (since it may refer to an input
+               path we haven't processed yet. */
+            if (input.follows) {
+                InputPath target;
+                if (hasOverride || input.absolute)
+                    /* 'follows' from an override is relative to the
+                       root of the graph. */
+                    target = *input.follows;
+                else {
+                    /* Otherwise, it's relative to the current flake. */
+                    target = inputPathPrefix;
+                    for (auto & i : *input.follows) target.push_back(i);
+                }
+                debug("input '%s' follows '%s'", inputPathS, printInputPath(target));
+                node->inputs.insert_or_assign(id, target);
+                continue;
+            }
+
+            assert(input.ref);
+
+            /* Do we have an entry in the existing lock file? And we
+               don't have a --update-input flag for this input? */
+            std::shared_ptr<LockedNode> oldLock;
+
+            updatesUsed.insert(inputPath);
+
+            if (oldNode && !lockFlags.inputUpdates.count(inputPath))
+                if (auto oldLock2 = get(oldNode->inputs, id))
+                    if (auto oldLock3 = std::get_if<0>(&*oldLock2))
+                        oldLock = *oldLock3;
+
+            if (oldLock
+                && oldLock->originalRef == *input.ref
+                && !hasOverride)
+            {
+                debug("keeping existing input '%s'", inputPathS);
+
+                /* Copy the input from the old lock since its flakeref
+                   didn't change and there is no override from a
+                   higher level flake. */
+                auto childNode = std::make_shared<LockedNode>(
+                    oldLock->lockedRef, oldLock->originalRef, oldLock->isFlake);
+
+                node->inputs.insert_or_assign(id, childNode);
+
+                /* If we have an --update-input flag for an input
+                   of this input, then we must fetch the flake to
+                   to update it. */
+                auto lb = lockFlags.inputUpdates.lower_bound(inputPath);
+
+                auto hasChildUpdate =
+                    lb != lockFlags.inputUpdates.end()
+                    && lb->size() > inputPath.size()
+                    && std::equal(inputPath.begin(), inputPath.end(), lb->begin());
+
+                if (hasChildUpdate) {
+                    auto inputFlake = getFlake(
+                        state, oldLock->lockedRef, false, flakeCache);
+                    computeLocks(inputFlake.inputs, childNode, inputPath, oldLock);
+                } else {
+                    /* No need to fetch this flake, we can be
+                       lazy. However there may be new overrides on the
+                       inputs of this flake, so we need to check
+                       those. */
+                    FlakeInputs fakeInputs;
+
+                    for (auto & i : oldLock->inputs) {
+                        if (auto lockedNode = std::get_if<0>(&i.second)) {
+                            fakeInputs.emplace(i.first, FlakeInput {
+                                .ref = (*lockedNode)->originalRef,
+                                .isFlake = (*lockedNode)->isFlake,
+                            });
+                        } else if (auto follows = std::get_if<1>(&i.second)) {
+                            fakeInputs.emplace(i.first, FlakeInput {
+                                .follows = *follows,
+                                .absolute = true
+                            });
+                        }
+                    }
+
+                    computeLocks(fakeInputs, childNode, inputPath, oldLock);
+                }
+
+            } else {
+                /* We need to create a new lock file entry. So fetch
+                   this input. */
+                debug("creating new input '%s'", inputPathS);
+
+                if (!lockFlags.allowMutable && !input.ref->input.isImmutable())
+                    throw Error("cannot update flake input '%s' in pure mode", inputPathS);
+
+                if (input.isFlake) {
+                    auto inputFlake = getFlake(state, *input.ref, lockFlags.useRegistries, flakeCache);
+
+                    /* Note: in case of an --override-input, we use
+                       the *original* ref (input2.ref) for the
+                       "original" field, rather than the
+                       override. This ensures that the override isn't
+                       nuked the next time we update the lock
+                       file. That is, overrides are sticky unless you
+                       use --no-write-lock-file. */
+                    auto childNode = std::make_shared<LockedNode>(
+                        inputFlake.lockedRef, input2.ref ? *input2.ref : *input.ref);
+
+                    node->inputs.insert_or_assign(id, childNode);
+
+                    /* Guard against circular flake imports. */
+                    for (auto & parent : parents)
+                        if (parent == *input.ref)
+                            throw Error("found circular import of flake '%s'", parent);
+                    parents.push_back(*input.ref);
+                    Finally cleanup([&]() { parents.pop_back(); });
+
+                    /* Recursively process the inputs of this
+                       flake. Also, unless we already have this flake
+                       in the top-level lock file, use this flake's
+                       own lock file. */
+                    computeLocks(
+                        inputFlake.inputs, childNode, inputPath,
+                        oldLock
+                        ? std::dynamic_pointer_cast<const Node>(oldLock)
+                        : LockFile::read(
+                            inputFlake.sourceInfo->actualPath + "/" + inputFlake.lockedRef.subdir + "/flake.lock").root);
+                }
+
+                else {
+                    auto [sourceInfo, resolvedRef, lockedRef] = fetchOrSubstituteTree(
+                        state, *input.ref, lockFlags.useRegistries, flakeCache);
+                    node->inputs.insert_or_assign(id,
+                        std::make_shared<LockedNode>(lockedRef, *input.ref, false));
+                }
+            }
+        }
+    };
+
+    computeLocks(
+        flake.inputs, newLockFile.root, {},
+        lockFlags.recreateLockFile ? nullptr : oldLockFile.root);
+
+    for (auto & i : lockFlags.inputOverrides)
+        if (!overridesUsed.count(i.first))
+            warn("the flag '--override-input %s %s' does not match any input",
+                printInputPath(i.first), i.second);
+
+    for (auto & i : lockFlags.inputUpdates)
+        if (!updatesUsed.count(i))
+            warn("the flag '--update-input %s' does not match any input", printInputPath(i));
+
+    /* Check 'follows' inputs. */
+    newLockFile.check();
+
+    debug("new lock file: %s", newLockFile);
+
+    /* Check whether we need to / can write the new lock file. */
+    if (!(newLockFile == oldLockFile)) {
+
+        auto diff = LockFile::diff(oldLockFile, newLockFile);
+
+        if (lockFlags.writeLockFile) {
+            if (auto sourcePath = topRef.input.getSourcePath()) {
+                if (!newLockFile.isImmutable()) {
+                    if (settings.warnDirty)
+                        warn("will not write lock file of flake '%s' because it has a mutable input", topRef);
+                } else {
+                    if (!lockFlags.updateLockFile)
+                        throw Error("flake '%s' requires lock file changes but they're not allowed due to '--no-update-lock-file'", topRef);
+
+                    auto relPath = (topRef.subdir == "" ? "" : topRef.subdir + "/") + "flake.lock";
+
+                    auto path = *sourcePath + "/" + relPath;
+
+                    bool lockFileExists = pathExists(path);
+
+                    if (lockFileExists) {
+                        auto s = chomp(diff);
+                        if (s.empty())
+                            warn("updating lock file '%s'", path);
+                        else
+                            warn("updating lock file '%s':\n%s", path, s);
+                    } else
+                        warn("creating lock file '%s'", path);
+
+                    newLockFile.write(path);
+
+                    topRef.input.markChangedFile(
+                        (topRef.subdir == "" ? "" : topRef.subdir + "/") + "flake.lock",
+                        lockFlags.commitLockFile
+                        ? std::optional<std::string>(fmt("%s: %s\n\nFlake input changes:\n\n%s",
+                                relPath, lockFileExists ? "Update" : "Add", diff))
+                        : std::nullopt);
+
+                    /* Rewriting the lockfile changed the top-level
+                       repo, so we should re-read it. FIXME: we could
+                       also just clear the 'rev' field... */
+                    auto prevLockedRef = flake.lockedRef;
+                    FlakeCache dummyCache;
+                    flake = getFlake(state, topRef, lockFlags.useRegistries, dummyCache);
+
+                    if (lockFlags.commitLockFile &&
+                        flake.lockedRef.input.getRev() &&
+                        prevLockedRef.input.getRev() != flake.lockedRef.input.getRev())
+                        warn("committed new revision '%s'", flake.lockedRef.input.getRev()->gitRev());
+
+                    /* Make sure that we picked up the change,
+                       i.e. the tree should usually be dirty
+                       now. Corner case: we could have reverted from a
+                       dirty to a clean tree! */
+                    if (flake.lockedRef.input == prevLockedRef.input
+                        && !flake.lockedRef.input.isImmutable())
+                        throw Error("'%s' did not change after I updated its 'flake.lock' file; is 'flake.lock' under version control?", flake.originalRef);
+                }
+            } else
+                throw Error("cannot write modified lock file of flake '%s' (use '--no-write-lock-file' to ignore)", topRef);
+        } else
+            warn("not writing modified lock file of flake '%s':\n%s", topRef, chomp(diff));
+    }
+
+    return LockedFlake { .flake = std::move(flake), .lockFile = std::move(newLockFile) };
+}
+
+void callFlake(EvalState & state,
+    const LockedFlake & lockedFlake,
+    Value & vRes)
+{
+    auto vLocks = state.allocValue();
+    auto vRootSrc = state.allocValue();
+    auto vRootSubdir = state.allocValue();
+    auto vTmp1 = state.allocValue();
+    auto vTmp2 = state.allocValue();
+
+    mkString(*vLocks, lockedFlake.lockFile.to_string());
+
+    emitTreeAttrs(state, *lockedFlake.flake.sourceInfo, lockedFlake.flake.lockedRef.input, *vRootSrc);
+
+    mkString(*vRootSubdir, lockedFlake.flake.lockedRef.subdir);
+
+    static RootValue vCallFlake = nullptr;
+
+    if (!vCallFlake) {
+        vCallFlake = allocRootValue(state.allocValue());
+        state.eval(state.parseExprFromString(
+            #include "call-flake.nix.gen.hh"
+            , "/"), **vCallFlake);
+    }
+
+    state.callFunction(**vCallFlake, *vLocks, *vTmp1, noPos);
+    state.callFunction(*vTmp1, *vRootSrc, *vTmp2, noPos);
+    state.callFunction(*vTmp2, *vRootSubdir, vRes, noPos);
+}
+
+static void prim_getFlake(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    auto flakeRefS = state.forceStringNoCtx(*args[0], pos);
+    auto flakeRef = parseFlakeRef(flakeRefS, {}, true);
+    if (evalSettings.pureEval && !flakeRef.input.isImmutable())
+        throw Error("cannot call 'getFlake' on mutable flake reference '%s', at %s (use --impure to override)", flakeRefS, pos);
+
+    callFlake(state,
+        lockFlake(state, flakeRef,
+            LockFlags {
+                .updateLockFile = false,
+                .useRegistries = !evalSettings.pureEval,
+                .allowMutable  = !evalSettings.pureEval,
+            }),
+        v);
+}
+
+static RegisterPrimOp r2("__getFlake", 1, prim_getFlake, "flakes");
+
+}
+
+Fingerprint LockedFlake::getFingerprint() const
+{
+    // FIXME: as an optimization, if the flake contains a lock file
+    // and we haven't changed it, then it's sufficient to use
+    // flake.sourceInfo.storePath for the fingerprint.
+    return hashString(htSHA256,
+        fmt("%s;%d;%d;%s",
+            flake.sourceInfo->storePath.to_string(),
+            flake.lockedRef.input.getRevCount().value_or(0),
+            flake.lockedRef.input.getLastModified().value_or(0),
+            lockFile));
+}
+
+Flake::~Flake() { }
+
+}

src/libexpr/flake/flake.hh

@@ -0,0 +1,111 @@
+#pragma once
+
+#include "types.hh"
+#include "flakeref.hh"
+#include "lockfile.hh"
+#include "value.hh"
+
+namespace nix {
+
+class EvalState;
+
+namespace fetchers { struct Tree; }
+
+namespace flake {
+
+struct FlakeInput;
+
+typedef std::map<FlakeId, FlakeInput> FlakeInputs;
+
+struct FlakeInput
+{
+    std::optional<FlakeRef> ref;
+    bool isFlake = true;
+    std::optional<InputPath> follows;
+    bool absolute = false; // whether 'follows' is relative to the flake root
+    FlakeInputs overrides;
+};
+
+struct Flake
+{
+    FlakeRef originalRef;
+    FlakeRef resolvedRef;
+    FlakeRef lockedRef;
+    std::optional<std::string> description;
+    std::shared_ptr<const fetchers::Tree> sourceInfo;
+    FlakeInputs inputs;
+    RootValue vOutputs;
+    ~Flake();
+};
+
+Flake getFlake(EvalState & state, const FlakeRef & flakeRef, bool allowLookup);
+
+/* Fingerprint of a locked flake; used as a cache key. */
+typedef Hash Fingerprint;
+
+struct LockedFlake
+{
+    Flake flake;
+    LockFile lockFile;
+
+    Fingerprint getFingerprint() const;
+};
+
+struct LockFlags
+{
+    /* Whether to ignore the existing lock file, creating a new one
+       from scratch. */
+    bool recreateLockFile = false;
+
+    /* Whether to update the lock file at all. If set to false, if any
+       change to the lock file is needed (e.g. when an input has been
+       added to flake.nix), you get a fatal error. */
+    bool updateLockFile = true;
+
+    /* Whether to write the lock file to disk. If set to true, if the
+       any changes to the lock file are needed and the flake is not
+       writable (i.e. is not a local Git working tree or similar), you
+       get a fatal error. If set to false, Nix will use the modified
+       lock file in memory only, without writing it to disk. */
+    bool writeLockFile = true;
+
+    /* Whether to use the registries to lookup indirect flake
+       references like 'nixpkgs'. */
+    bool useRegistries = true;
+
+    /* Whether mutable flake references (i.e. those without a Git
+       revision or similar) without a corresponding lock are
+       allowed. Mutable flake references with a lock are always
+       allowed. */
+    bool allowMutable = true;
+
+    /* Whether to commit changes to flake.lock. */
+    bool commitLockFile = false;
+
+    /* Flake inputs to be overriden. */
+    std::map<InputPath, FlakeRef> inputOverrides;
+
+    /* Flake inputs to be updated. This means that any existing lock
+       for those inputs will be ignored. */
+    std::set<InputPath> inputUpdates;
+};
+
+LockedFlake lockFlake(
+    EvalState & state,
+    const FlakeRef & flakeRef,
+    const LockFlags & lockFlags);
+
+void callFlake(
+    EvalState & state,
+    const LockedFlake & lockedFlake,
+    Value & v);
+
+}
+
+void emitTreeAttrs(
+    EvalState & state,
+    const fetchers::Tree & tree,
+    const fetchers::Input & input,
+    Value & v, bool emptyRevFallback = false);
+
+}

src/libexpr/flake/flakeref.cc

@@ -0,0 +1,204 @@
+#include "flakeref.hh"
+#include "store-api.hh"
+#include "url.hh"
+#include "fetchers.hh"
+#include "registry.hh"
+
+namespace nix {
+
+#if 0
+// 'dir' path elements cannot start with a '.'. We also reject
+// potentially dangerous characters like ';'.
+const static std::string subDirElemRegex = "(?:[a-zA-Z0-9_-]+[a-zA-Z0-9._-]*)";
+const static std::string subDirRegex = subDirElemRegex + "(?:/" + subDirElemRegex + ")*";
+#endif
+
+std::string FlakeRef::to_string() const
+{
+    auto url = input.toURL();
+    if (subdir != "")
+        url.query.insert_or_assign("dir", subdir);
+    return url.to_string();
+}
+
+fetchers::Attrs FlakeRef::toAttrs() const
+{
+    auto attrs = input.toAttrs();
+    if (subdir != "")
+        attrs.emplace("dir", subdir);
+    return attrs;
+}
+
+std::ostream & operator << (std::ostream & str, const FlakeRef & flakeRef)
+{
+    str << flakeRef.to_string();
+    return str;
+}
+
+bool FlakeRef::operator ==(const FlakeRef & other) const
+{
+    return input == other.input && subdir == other.subdir;
+}
+
+FlakeRef FlakeRef::resolve(ref<Store> store) const
+{
+    auto [input2, extraAttrs] = lookupInRegistries(store, input);
+    return FlakeRef(std::move(input2), fetchers::maybeGetStrAttr(extraAttrs, "dir").value_or(subdir));
+}
+
+FlakeRef parseFlakeRef(
+    const std::string & url, const std::optional<Path> & baseDir, bool allowMissing)
+{
+    auto [flakeRef, fragment] = parseFlakeRefWithFragment(url, baseDir, allowMissing);
+    if (fragment != "")
+        throw Error("unexpected fragment '%s' in flake reference '%s'", fragment, url);
+    return flakeRef;
+}
+
+std::optional<FlakeRef> maybeParseFlakeRef(
+    const std::string & url, const std::optional<Path> & baseDir)
+{
+    try {
+        return parseFlakeRef(url, baseDir);
+    } catch (Error &) {
+        return {};
+    }
+}
+
+std::pair<FlakeRef, std::string> parseFlakeRefWithFragment(
+    const std::string & url, const std::optional<Path> & baseDir, bool allowMissing)
+{
+    using namespace fetchers;
+
+    static std::string fnRegex = "[0-9a-zA-Z-._~!$&'\"()*+,;=]+";
+
+    static std::regex pathUrlRegex(
+        "(/?" + fnRegex + "(?:/" + fnRegex + ")*/?)"
+        + "(?:\\?(" + queryRegex + "))?"
+        + "(?:#(" + queryRegex + "))?",
+        std::regex::ECMAScript);
+
+    static std::regex flakeRegex(
+        "((" + flakeIdRegexS + ")(?:/(?:" + refAndOrRevRegex + "))?)"
+        + "(?:#(" + queryRegex + "))?",
+        std::regex::ECMAScript);
+
+    std::smatch match;
+
+    /* Check if 'url' is a flake ID. This is an abbreviated syntax for
+       'flake:<flake-id>?ref=<ref>&rev=<rev>'. */
+
+    if (std::regex_match(url, match, flakeRegex)) {
+        auto parsedURL = ParsedURL{
+            .url = url,
+            .base = "flake:" + std::string(match[1]),
+            .scheme = "flake",
+            .authority = "",
+            .path = match[1],
+        };
+
+        return std::make_pair(
+            FlakeRef(Input::fromURL(parsedURL), ""),
+            percentDecode(std::string(match[6])));
+    }
+
+    else if (std::regex_match(url, match, pathUrlRegex)) {
+        std::string path = match[1];
+        std::string fragment = percentDecode(std::string(match[3]));
+
+        if (baseDir) {
+            /* Check if 'url' is a path (either absolute or relative
+               to 'baseDir'). If so, search upward to the root of the
+               repo (i.e. the directory containing .git). */
+
+            path = absPath(path, baseDir, true);
+
+            if (!S_ISDIR(lstat(path).st_mode))
+                throw BadURL("path '%s' is not a flake (because it's not a directory)", path);
+
+            if (!allowMissing && !pathExists(path + "/flake.nix"))
+                throw BadURL("path '%s' is not a flake (because it doesn't contain a 'flake.nix' file)", path);
+
+            auto flakeRoot = path;
+            std::string subdir;
+
+            while (flakeRoot != "/") {
+                if (pathExists(flakeRoot + "/.git")) {
+                    auto base = std::string("git+file://") + flakeRoot;
+
+                    auto parsedURL = ParsedURL{
+                        .url = base, // FIXME
+                        .base = base,
+                        .scheme = "git+file",
+                        .authority = "",
+                        .path = flakeRoot,
+                        .query = decodeQuery(match[2]),
+                    };
+
+                    if (subdir != "") {
+                        if (parsedURL.query.count("dir"))
+                            throw Error("flake URL '%s' has an inconsistent 'dir' parameter", url);
+                        parsedURL.query.insert_or_assign("dir", subdir);
+                    }
+
+                    if (pathExists(flakeRoot + "/.git/shallow"))
+                        parsedURL.query.insert_or_assign("shallow", "1");
+
+                    return std::make_pair(
+                        FlakeRef(Input::fromURL(parsedURL), get(parsedURL.query, "dir").value_or("")),
+                        fragment);
+                }
+
+                subdir = std::string(baseNameOf(flakeRoot)) + (subdir.empty() ? "" : "/" + subdir);
+                flakeRoot = dirOf(flakeRoot);
+            }
+
+        } else {
+            if (!hasPrefix(path, "/"))
+                throw BadURL("flake reference '%s' is not an absolute path", url);
+            path = canonPath(path);
+        }
+
+        fetchers::Attrs attrs;
+        attrs.insert_or_assign("type", "path");
+        attrs.insert_or_assign("path", path);
+
+        return std::make_pair(FlakeRef(Input::fromAttrs(std::move(attrs)), ""), fragment);
+    }
+
+    else {
+        auto parsedURL = parseURL(url);
+        std::string fragment;
+        std::swap(fragment, parsedURL.fragment);
+        return std::make_pair(
+            FlakeRef(Input::fromURL(parsedURL), get(parsedURL.query, "dir").value_or("")),
+            fragment);
+    }
+}
+
+std::optional<std::pair<FlakeRef, std::string>> maybeParseFlakeRefWithFragment(
+    const std::string & url, const std::optional<Path> & baseDir)
+{
+    try {
+        return parseFlakeRefWithFragment(url, baseDir);
+    } catch (Error & e) {
+        return {};
+    }
+}
+
+FlakeRef FlakeRef::fromAttrs(const fetchers::Attrs & attrs)
+{
+    auto attrs2(attrs);
+    attrs2.erase("dir");
+    return FlakeRef(
+        fetchers::Input::fromAttrs(std::move(attrs2)),
+        fetchers::maybeGetStrAttr(attrs, "dir").value_or(""));
+}
+
+std::pair<fetchers::Tree, FlakeRef> FlakeRef::fetchTree(ref<Store> store) const
+{
+    auto [tree, lockedInput] = input.fetch(store);
+    return {std::move(tree), FlakeRef(std::move(lockedInput), subdir)};
+}
+
+}

src/libexpr/flake/flakeref.hh

@@ -0,0 +1,53 @@
+#pragma once
+
+#include "types.hh"
+#include "hash.hh"
+#include "fetchers.hh"
+
+#include <variant>
+
+namespace nix {
+
+class Store;
+
+typedef std::string FlakeId;
+
+struct FlakeRef
+{
+    fetchers::Input input;
+
+    Path subdir;
+
+    bool operator==(const FlakeRef & other) const;
+
+    FlakeRef(fetchers::Input && input, const Path & subdir)
+        : input(std::move(input)), subdir(subdir)
+    { }
+
+    // FIXME: change to operator <<.
+    std::string to_string() const;
+
+    fetchers::Attrs toAttrs() const;
+
+    FlakeRef resolve(ref<Store> store) const;
+
+    static FlakeRef fromAttrs(const fetchers::Attrs & attrs);
+
+    std::pair<fetchers::Tree, FlakeRef> fetchTree(ref<Store> store) const;
+};
+
+std::ostream & operator << (std::ostream & str, const FlakeRef & flakeRef);
+
+FlakeRef parseFlakeRef(
+    const std::string & url, const std::optional<Path> & baseDir = {}, bool allowMissing = false);
+
+std::optional<FlakeRef> maybeParseFlake(
+    const std::string & url, const std::optional<Path> & baseDir = {});
+
+std::pair<FlakeRef, std::string> parseFlakeRefWithFragment(
+    const std::string & url, const std::optional<Path> & baseDir = {}, bool allowMissing = false);
+
+std::optional<std::pair<FlakeRef, std::string>> maybeParseFlakeRefWithFragment(
+    const std::string & url, const std::optional<Path> & baseDir = {});
+
+}

src/libexpr/flake/lockfile.cc

@@ -0,0 +1,338 @@
+#include "lockfile.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::flake {
+
+FlakeRef getFlakeRef(
+    const nlohmann::json & json,
+    const char * attr,
+    const char * info)
+{
+    auto i = json.find(attr);
+    if (i != json.end()) {
+        auto attrs = jsonToAttrs(*i);
+        // FIXME: remove when we drop support for version 5.
+        if (info) {
+            auto j = json.find(info);
+            if (j != json.end()) {
+                for (auto k : jsonToAttrs(*j))
+                    attrs.insert_or_assign(k.first, k.second);
+            }
+        }
+        return FlakeRef::fromAttrs(attrs);
+    }
+
+    throw Error("attribute '%s' missing in lock file", attr);
+}
+
+LockedNode::LockedNode(const nlohmann::json & json)
+    : lockedRef(getFlakeRef(json, "locked", "info"))
+    , originalRef(getFlakeRef(json, "original", nullptr))
+    , isFlake(json.find("flake") != json.end() ? (bool) json["flake"] : true)
+{
+    if (!lockedRef.input.isImmutable())
+        throw Error("lockfile contains mutable lock '%s'", attrsToJson(lockedRef.input.toAttrs()));
+}
+
+StorePath LockedNode::computeStorePath(Store & store) const
+{
+    return lockedRef.input.computeStorePath(store);
+}
+
+std::shared_ptr<Node> LockFile::findInput(const InputPath & path)
+{
+    auto pos = root;
+
+    if (!pos) return {};
+
+    for (auto & elem : path) {
+        if (auto i = get(pos->inputs, elem)) {
+            if (auto node = std::get_if<0>(&*i))
+                pos = *node;
+            else if (auto follows = std::get_if<1>(&*i)) {
+                pos = findInput(*follows);
+                if (!pos) return {};
+            }
+        } else
+            return {};
+    }
+
+    return pos;
+}
+
+LockFile::LockFile(const nlohmann::json & json, const Path & path)
+{
+    auto version = json.value("version", 0);
+    if (version < 5 || version > 7)
+        throw Error("lock file '%s' has unsupported version %d", path, version);
+
+    std::unordered_map<std::string, std::shared_ptr<Node>> nodeMap;
+
+    std::function<void(Node & node, const nlohmann::json & jsonNode)> getInputs;
+
+    getInputs = [&](Node & node, const nlohmann::json & jsonNode)
+    {
+        if (jsonNode.find("inputs") == jsonNode.end()) return;
+        for (auto & i : jsonNode["inputs"].items()) {
+            if (i.value().is_array()) {
+                InputPath path;
+                for (auto & j : i.value())
+                    path.push_back(j);
+                node.inputs.insert_or_assign(i.key(), path);
+            } else {
+                std::string inputKey = i.value();
+                auto k = nodeMap.find(inputKey);
+                if (k == nodeMap.end()) {
+                    auto jsonNode2 = json["nodes"][inputKey];
+                    auto input = std::make_shared<LockedNode>(jsonNode2);
+                    k = nodeMap.insert_or_assign(inputKey, input).first;
+                    getInputs(*input, jsonNode2);
+                }
+                if (auto child = std::dynamic_pointer_cast<LockedNode>(k->second))
+                    node.inputs.insert_or_assign(i.key(), child);
+                else
+                    // FIXME: replace by follows node
+                    throw Error("lock file contains cycle to root node");
+            }
+        }
+    };
+
+    std::string rootKey = json["root"];
+    nodeMap.insert_or_assign(rootKey, root);
+    getInputs(*root, json["nodes"][rootKey]);
+
+    // FIXME: check that there are no cycles in version >= 7. Cycles
+    // between inputs are only possible using 'follows' indirections.
+    // Once we drop support for version <= 6, we can simplify the code
+    // a bit since we don't need to worry about cycles.
+}
+
+nlohmann::json LockFile::toJson() const
+{
+    nlohmann::json nodes;
+    std::unordered_map<std::shared_ptr<const Node>, std::string> nodeKeys;
+    std::unordered_set<std::string> keys;
+
+    std::function<std::string(const std::string & key, std::shared_ptr<const Node> node)> dumpNode;
+
+    dumpNode = [&](std::string key, std::shared_ptr<const Node> node) -> std::string
+    {
+        auto k = nodeKeys.find(node);
+        if (k != nodeKeys.end())
+            return k->second;
+
+        if (!keys.insert(key).second) {
+            for (int n = 2; ; ++n) {
+                auto k = fmt("%s_%d", key, n);
+                if (keys.insert(k).second) {
+                    key = k;
+                    break;
+                }
+            }
+        }
+
+        nodeKeys.insert_or_assign(node, key);
+
+        auto n = nlohmann::json::object();
+
+        if (!node->inputs.empty()) {
+            auto inputs = nlohmann::json::object();
+            for (auto & i : node->inputs) {
+                if (auto child = std::get_if<0>(&i.second)) {
+                    inputs[i.first] = dumpNode(i.first, *child);
+                } else if (auto follows = std::get_if<1>(&i.second)) {
+                    auto arr = nlohmann::json::array();
+                    for (auto & x : *follows)
+                        arr.push_back(x);
+                    inputs[i.first] = std::move(arr);
+                }
+            }
+            n["inputs"] = std::move(inputs);
+        }
+
+        if (auto lockedNode = std::dynamic_pointer_cast<const LockedNode>(node)) {
+            n["original"] = fetchers::attrsToJson(lockedNode->originalRef.toAttrs());
+            n["locked"] = fetchers::attrsToJson(lockedNode->lockedRef.toAttrs());
+            if (!lockedNode->isFlake) n["flake"] = false;
+        }
+
+        nodes[key] = std::move(n);
+
+        return key;
+    };
+
+    nlohmann::json json;
+    json["version"] = 7;
+    json["root"] = dumpNode("root", root);
+    json["nodes"] = std::move(nodes);
+
+    return json;
+}
+
+std::string LockFile::to_string() const
+{
+    return toJson().dump(2);
+}
+
+LockFile LockFile::read(const Path & path)
+{
+    if (!pathExists(path)) return LockFile();
+    return LockFile(nlohmann::json::parse(readFile(path)), path);
+}
+
+std::ostream & operator <<(std::ostream & stream, const LockFile & lockFile)
+{
+    stream << lockFile.toJson().dump(2);
+    return stream;
+}
+
+void LockFile::write(const Path & path) const
+{
+    createDirs(dirOf(path));
+    writeFile(path, fmt("%s\n", *this));
+}
+
+bool LockFile::isImmutable() const
+{
+    std::unordered_set<std::shared_ptr<const Node>> nodes;
+
+    std::function<void(std::shared_ptr<const Node> node)> visit;
+
+    visit = [&](std::shared_ptr<const Node> node)
+    {
+        if (!nodes.insert(node).second) return;
+        for (auto & i : node->inputs)
+            if (auto child = std::get_if<0>(&i.second))
+                visit(*child);
+    };
+
+    visit(root);
+
+    for (auto & i : nodes) {
+        if (i == root) continue;
+        auto lockedNode = std::dynamic_pointer_cast<const LockedNode>(i);
+        if (lockedNode && !lockedNode->lockedRef.input.isImmutable()) return false;
+    }
+
+    return true;
+}
+
+bool LockFile::operator ==(const LockFile & other) const
+{
+    // FIXME: slow
+    return toJson() == other.toJson();
+}
+
+InputPath parseInputPath(std::string_view s)
+{
+    InputPath path;
+
+    for (auto & elem : tokenizeString<std::vector<std::string>>(s, "/")) {
+        if (!std::regex_match(elem, flakeIdRegex))
+            throw UsageError("invalid flake input path element '%s'", elem);
+        path.push_back(elem);
+    }
+
+    return path;
+}
+
+std::map<InputPath, Node::Edge> LockFile::getAllInputs() const
+{
+    std::unordered_set<std::shared_ptr<Node>> done;
+    std::map<InputPath, Node::Edge> res;
+
+    std::function<void(const InputPath & prefix, std::shared_ptr<Node> node)> recurse;
+
+    recurse = [&](const InputPath & prefix, std::shared_ptr<Node> node)
+    {
+        if (!done.insert(node).second) return;
+
+        for (auto &[id, input] : node->inputs) {
+            auto inputPath(prefix);
+            inputPath.push_back(id);
+            res.emplace(inputPath, input);
+            if (auto child = std::get_if<0>(&input))
+                recurse(inputPath, *child);
+        }
+    };
+
+    recurse({}, root);
+
+    return res;
+}
+
+std::ostream & operator <<(std::ostream & stream, const Node::Edge & edge)
+{
+    if (auto node = std::get_if<0>(&edge))
+        stream << "'" << (*node)->lockedRef << "'";
+    else if (auto follows = std::get_if<1>(&edge))
+        stream << fmt("follows '%s'", printInputPath(*follows));
+    return stream;
+}
+
+static bool equals(const Node::Edge & e1, const Node::Edge & e2)
+{
+    if (auto n1 = std::get_if<0>(&e1))
+        if (auto n2 = std::get_if<0>(&e2))
+            return (*n1)->lockedRef == (*n2)->lockedRef;
+    if (auto f1 = std::get_if<1>(&e1))
+        if (auto f2 = std::get_if<1>(&e2))
+            return *f1 == *f2;
+    return false;
+}
+
+std::string LockFile::diff(const LockFile & oldLocks, const LockFile & newLocks)
+{
+    auto oldFlat = oldLocks.getAllInputs();
+    auto newFlat = newLocks.getAllInputs();
+
+    auto i = oldFlat.begin();
+    auto j = newFlat.begin();
+    std::string res;
+
+    while (i != oldFlat.end() || j != newFlat.end()) {
+        if (j != newFlat.end() && (i == oldFlat.end() || i->first > j->first)) {
+            res += fmt("* Added '%s': %s\n", printInputPath(j->first), j->second);
+            ++j;
+        } else if (i != oldFlat.end() && (j == newFlat.end() || i->first < j->first)) {
+            res += fmt("* Removed '%s'\n", printInputPath(i->first));
+            ++i;
+        } else {
+            if (!equals(i->second, j->second)) {
+                res += fmt("* Updated '%s': %s -> %s\n",
+                    printInputPath(i->first),
+                    i->second,
+                    j->second);
+            }
+            ++i;
+            ++j;
+        }
+    }
+
+    return res;
+}
+
+void LockFile::check()
+{
+    auto inputs = getAllInputs();
+
+    for (auto & [inputPath, input] : inputs) {
+        if (auto follows = std::get_if<1>(&input)) {
+            if (!follows->empty() && !get(inputs, *follows))
+                throw Error("input '%s' follows a non-existent input '%s'",
+                    printInputPath(inputPath),
+                    printInputPath(*follows));
+        }
+    }
+}
+
+void check();
+
+std::string printInputPath(const InputPath & path)
+{
+    return concatStringsSep("/", path);
+}
+
+}

src/libexpr/flake/lockfile.hh

@@ -0,0 +1,85 @@
+#pragma once
+
+#include "flakeref.hh"
+
+#include <nlohmann/json_fwd.hpp>
+
+namespace nix {
+class Store;
+struct StorePath;
+}
+
+namespace nix::flake {
+
+using namespace fetchers;
+
+typedef std::vector<FlakeId> InputPath;
+
+struct LockedNode;
+
+/* A node in the lock file. It has outgoing edges to other nodes (its
+   inputs). Only the root node has this type; all other nodes have
+   type LockedNode. */
+struct Node : std::enable_shared_from_this<Node>
+{
+    typedef std::variant<std::shared_ptr<LockedNode>, InputPath> Edge;
+
+    std::map<FlakeId, Edge> inputs;
+
+    virtual ~Node() { }
+};
+
+/* A non-root node in the lock file. */
+struct LockedNode : Node
+{
+    FlakeRef lockedRef, originalRef;
+    bool isFlake = true;
+
+    LockedNode(
+        const FlakeRef & lockedRef,
+        const FlakeRef & originalRef,
+        bool isFlake = true)
+        : lockedRef(lockedRef), originalRef(originalRef), isFlake(isFlake)
+    { }
+
+    LockedNode(const nlohmann::json & json);
+
+    StorePath computeStorePath(Store & store) const;
+};
+
+struct LockFile
+{
+    std::shared_ptr<Node> root = std::make_shared<Node>();
+
+    LockFile() {};
+    LockFile(const nlohmann::json & json, const Path & path);
+
+    nlohmann::json toJson() const;
+
+    std::string to_string() const;
+
+    static LockFile read(const Path & path);
+
+    void write(const Path & path) const;
+
+    bool isImmutable() const;
+
+    bool operator ==(const LockFile & other) const;
+
+    std::shared_ptr<Node> findInput(const InputPath & path);
+
+    std::map<InputPath, Node::Edge> getAllInputs() const;
+
+    static std::string diff(const LockFile & oldLocks, const LockFile & newLocks);
+
+    /* Check that every 'follows' input target exists. */
+    void check();
+};
+
+std::ostream & operator <<(std::ostream & stream, const LockFile & lockFile);
+
+InputPath parseInputPath(std::string_view s);
+
+std::string printInputPath(const InputPath & path);
+
+}

src/libexpr/get-drvs.cc

@@ -39,7 +39,7 @@ DrvInfo::DrvInfo(EvalState & state, ref<Store> store, const std::string & drvPat
     if (i == drv.outputs.end())
         throw Error("derivation '%s' does not have output '%s'", store->printStorePath(drvPath), outputName);
 
-    outPath = store->printStorePath(i->second.path);
+    outPath = store->printStorePath(i->second.path(*store, drv.name));
 }
 
 

src/libexpr/local.mk

@@ -4,7 +4,12 @@ libexpr_NAME = libnixexpr
 
 libexpr_DIR := $(d)
 
-libexpr_SOURCES := $(wildcard $(d)/*.cc) $(wildcard $(d)/primops/*.cc) $(d)/lexer-tab.cc $(d)/parser-tab.cc
+libexpr_SOURCES := \
+  $(wildcard $(d)/*.cc) \
+  $(wildcard $(d)/primops/*.cc) \
+  $(wildcard $(d)/flake/*.cc) \
+  $(d)/lexer-tab.cc \
+  $(d)/parser-tab.cc
 
 libexpr_CXXFLAGS += -I src/libutil -I src/libstore -I src/libfetchers -I src/libmain -I src/libexpr
 
@@ -34,4 +39,9 @@ dist-files += $(d)/parser-tab.cc $(d)/parser-tab.hh $(d)/lexer-tab.cc $(d)/lexer
 
 $(eval $(call install-file-in, $(d)/nix-expr.pc, $(prefix)/lib/pkgconfig, 0644))
 
+$(foreach i, $(wildcard src/libexpr/flake/*.hh), \
+  $(eval $(call install-file-in, $(i), $(includedir)/nix/flake, 0644)))
+
 $(d)/primops.cc: $(d)/imported-drv-to-derivation.nix.gen.hh
+
+$(d)/flake/flake.cc: $(d)/flake/call-flake.nix.gen.hh

src/libexpr/parser.y

@@ -719,7 +719,7 @@ std::pair<bool, std::string> EvalState::resolveSearchPathElem(const SearchPathEl
     if (isUri(elem.second)) {
         try {
             res = { true, store->toRealPath(fetchers::downloadTarball(
-                        store, resolveUri(elem.second), "source", false).storePath) };
+                        store, resolveUri(elem.second), "source", false).first.storePath) };
         } catch (FileTransferError & e) {
             logWarning({
                 .name = "Entry download",

src/libexpr/primops.cc

@@ -30,18 +30,6 @@ namespace nix {
  *************************************************************/
 
 
-/* Decode a context string ‘!<name>!<path>’ into a pair <path,
-   name>. */
-std::pair<string, string> decodeContext(const string & s)
-{
-    if (s.at(0) == '!') {
-        size_t index = s.find("!", 1);
-        return std::pair<string, string>(string(s, index + 1), string(s, 1, index - 1));
-    } else
-        return std::pair<string, string>(s.at(0) == '/' ? s : string(s, 1), "");
-}
-
-
 InvalidPathError::InvalidPathError(const Path & path) :
     EvalError("path '%s' is not valid", path), path(path) {}
 
@@ -64,7 +52,7 @@ void EvalState::realiseContext(const PathSet & context)
                 DerivationOutputs::iterator i = drv.outputs.find(outputName);
                 if (i == drv.outputs.end())
                     throw Error("derivation '%s' does not have an output named '%s'", ctxS, outputName);
-                allowedPaths->insert(store->printStorePath(i->second.path));
+                allowedPaths->insert(store->printStorePath(i->second.path(*store, drv.name)));
             }
         }
     }
@@ -77,7 +65,7 @@ void EvalState::realiseContext(const PathSet & context)
 
     /* For performance, prefetch all substitute info. */
     StorePathSet willBuild, willSubstitute, unknown;
-    unsigned long long downloadSize, narSize;
+    uint64_t downloadSize, narSize;
     store->queryMissing(drvs, willBuild, willSubstitute, unknown, downloadSize, narSize);
 
     store->buildPaths(drvs);
@@ -103,8 +91,17 @@ static void prim_scopedImport(EvalState & state, const Pos & pos, Value * * args
     Path realPath = state.checkSourcePath(state.toRealPath(path, context));
 
     // FIXME
-    if (state.store->isStorePath(path) && state.store->isValidPath(state.store->parseStorePath(path)) && isDerivation(path)) {
-        Derivation drv = readDerivation(*state.store, realPath);
+    auto isValidDerivationInStore = [&]() -> std::optional<StorePath> {
+        if (!state.store->isStorePath(path))
+            return std::nullopt;
+        auto storePath = state.store->parseStorePath(path);
+        if (!(state.store->isValidPath(storePath) && isDerivation(path)))
+            return std::nullopt;
+        return storePath;
+    };
+    if (auto optStorePath = isValidDerivationInStore()) {
+        auto storePath = *optStorePath;
+        Derivation drv = readDerivation(*state.store, realPath, Derivation::nameFromPath(storePath));
         Value & w = *state.allocValue();
         state.mkAttrs(w, 3 + drv.outputs.size());
         Value * v2 = state.allocAttr(w, state.sDrvPath);
@@ -118,7 +115,7 @@ static void prim_scopedImport(EvalState & state, const Pos & pos, Value * * args
 
         for (const auto & o : drv.outputs) {
             v2 = state.allocAttr(w, state.symbols.create(o.first));
-            mkString(*v2, state.store->printStorePath(o.second.path), {"!" + o.first + "!" + path});
+            mkString(*v2, state.store->printStorePath(o.second.path(*state.store, drv.name)), {"!" + o.first + "!" + path});
             outputsVal->listElems()[outputs_index] = state.allocValue();
             mkString(*(outputsVal->listElems()[outputs_index++]), o.first);
         }
@@ -582,6 +579,7 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
 
     /* Build the derivation expression by processing the attributes. */
     Derivation drv;
+    drv.name = drvName;
 
     PathSet context;
 
@@ -776,11 +774,12 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
         auto outPath = state.store->makeFixedOutputPath(ingestionMethod, h, drvName);
         if (!jsonObject) drv.env["out"] = state.store->printStorePath(outPath);
         drv.outputs.insert_or_assign("out", DerivationOutput {
-            .path = std::move(outPath),
-            .hash = FixedOutputHash {
-                .method = ingestionMethod,
-                .hash = std::move(h),
-            },
+                .output = DerivationOutputFixed {
+                    .hash = FixedOutputHash {
+                        .method = ingestionMethod,
+                        .hash = std::move(h),
+                    },
+                },
         });
     }
 
@@ -795,8 +794,9 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
             if (!jsonObject) drv.env[i] = "";
             drv.outputs.insert_or_assign(i,
                 DerivationOutput {
-                    .path = StorePath::dummy,
-                    .hash = std::optional<FixedOutputHash> {},
+                    .output = DerivationOutputInputAddressed {
+                        .path = StorePath::dummy,
+                    },
                 });
         }
 
@@ -807,8 +807,9 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
             if (!jsonObject) drv.env[i] = state.store->printStorePath(outPath);
             drv.outputs.insert_or_assign(i,
                 DerivationOutput {
-                    .path = std::move(outPath),
-                    .hash = std::optional<FixedOutputHash>(),
+                    .output = DerivationOutputInputAddressed {
+                        .path = std::move(outPath),
+                    },
                 });
         }
     }
@@ -829,7 +830,7 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
     mkString(*state.allocAttr(v, state.sDrvPath), drvPathS, {"=" + drvPathS});
     for (auto & i : drv.outputs) {
         mkString(*state.allocAttr(v, state.symbols.create(i.first)),
-            state.store->printStorePath(i.second.path), {"!" + i.first + "!" + drvPathS});
+            state.store->printStorePath(i.second.path(*state.store, drv.name)), {"!" + i.first + "!" + drvPathS});
     }
     v.attrs->sort();
 }
@@ -883,10 +884,10 @@ static void prim_storePath(EvalState & state, const Pos & pos, Value * * args, V
             .hint = hintfmt("path '%1%' is not in the Nix store", path),
             .errPos = pos
         });
-    Path path2 = state.store->toStorePath(path);
+    auto path2 = state.store->toStorePath(path).first;
     if (!settings.readOnlyMode)
-        state.store->ensurePath(state.store->parseStorePath(path2));
-    context.insert(path2);
+        state.store->ensurePath(path2);
+    context.insert(state.store->printStorePath(path2));
     mkString(v, path, context);
 }
 
@@ -1123,7 +1124,7 @@ static void prim_toFile(EvalState & state, const Pos & pos, Value * * args, Valu
 
 
 static void addPath(EvalState & state, const Pos & pos, const string & name, const Path & path_,
-    Value * filterFun, FileIngestionMethod method, const Hash & expectedHash, Value & v)
+    Value * filterFun, FileIngestionMethod method, const std::optional<Hash> expectedHash, Value & v)
 {
     const auto path = evalSettings.pureEval && expectedHash ?
         path_ :
@@ -1154,7 +1155,7 @@ static void addPath(EvalState & state, const Pos & pos, const string & name, con
 
     std::optional<StorePath> expectedStorePath;
     if (expectedHash)
-        expectedStorePath = state.store->makeFixedOutputPath(method, expectedHash, name);
+        expectedStorePath = state.store->makeFixedOutputPath(method, *expectedHash, name);
     Path dstPath;
     if (!expectedHash || !state.store->isValidPath(*expectedStorePath)) {
         dstPath = state.store->printStorePath(settings.readOnlyMode
@@ -1188,7 +1189,7 @@ static void prim_filterSource(EvalState & state, const Pos & pos, Value * * args
             .errPos = pos
         });
 
-    addPath(state, pos, std::string(baseNameOf(path)), path, args[0], FileIngestionMethod::Recursive, Hash(), v);
+    addPath(state, pos, std::string(baseNameOf(path)), path, args[0], FileIngestionMethod::Recursive, std::nullopt, v);
 }
 
 static void prim_path(EvalState & state, const Pos & pos, Value * * args, Value & v)
@@ -1198,7 +1199,7 @@ static void prim_path(EvalState & state, const Pos & pos, Value * * args, Value
     string name;
     Value * filterFun = nullptr;
     auto method = FileIngestionMethod::Recursive;
-    Hash expectedHash;
+    Hash expectedHash(htSHA256);
 
     for (auto & attr : *args[0]->attrs) {
         const string & n(attr.name);

src/libexpr/primops/fetchGit.cc

@@ -1,91 +0,0 @@
-#include "primops.hh"
-#include "eval-inline.hh"
-#include "store-api.hh"
-#include "hash.hh"
-#include "fetchers.hh"
-#include "url.hh"
-
-namespace nix {
-
-static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Value & v)
-{
-    std::string url;
-    std::optional<std::string> ref;
-    std::optional<Hash> rev;
-    std::string name = "source";
-    bool fetchSubmodules = false;
-    PathSet context;
-
-    state.forceValue(*args[0]);
-
-    if (args[0]->type == tAttrs) {
-
-        state.forceAttrs(*args[0], pos);
-
-        for (auto & attr : *args[0]->attrs) {
-            string n(attr.name);
-            if (n == "url")
-                url = state.coerceToString(*attr.pos, *attr.value, context, false, false);
-            else if (n == "ref")
-                ref = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else if (n == "rev")
-                rev = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA1);
-            else if (n == "name")
-                name = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else if (n == "submodules")
-                fetchSubmodules = state.forceBool(*attr.value, *attr.pos);
-            else
-                throw EvalError({
-                    .hint = hintfmt("unsupported argument '%s' to 'fetchGit'", attr.name),
-                    .errPos = *attr.pos
-                });
-        }
-
-        if (url.empty())
-            throw EvalError({
-                .hint = hintfmt("'url' argument required"),
-                .errPos = pos
-            });
-
-    } else
-        url = state.coerceToString(pos, *args[0], context, false, false);
-
-    // FIXME: git externals probably can be used to bypass the URI
-    // whitelist. Ah well.
-    state.checkURI(url);
-
-    if (evalSettings.pureEval && !rev)
-        throw Error("in pure evaluation mode, 'fetchGit' requires a Git revision");
-
-    fetchers::Attrs attrs;
-    attrs.insert_or_assign("type", "git");
-    attrs.insert_or_assign("url", url.find("://") != std::string::npos ? url : "file://" + url);
-    if (ref) attrs.insert_or_assign("ref", *ref);
-    if (rev) attrs.insert_or_assign("rev", rev->gitRev());
-    if (fetchSubmodules) attrs.insert_or_assign("submodules", true);
-    auto input = fetchers::inputFromAttrs(attrs);
-
-    // FIXME: use name?
-    auto [tree, input2] = input->fetchTree(state.store);
-
-    state.mkAttrs(v, 8);
-    auto storePath = state.store->printStorePath(tree.storePath);
-    mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
-    // Backward compatibility: set 'rev' to
-    // 0000000000000000000000000000000000000000 for a dirty tree.
-    auto rev2 = input2->getRev().value_or(Hash(htSHA1));
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), rev2.gitRev());
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), rev2.gitShortRev());
-    // Backward compatibility: set 'revCount' to 0 for a dirty tree.
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")),
-        tree.info.revCount.value_or(0));
-    mkBool(*state.allocAttr(v, state.symbols.create("submodules")), fetchSubmodules);
-    v.attrs->sort();
-
-    if (state.allowedPaths)
-        state.allowedPaths->insert(tree.actualPath);
-}
-
-static RegisterPrimOp r("fetchGit", 1, prim_fetchGit);
-
-}

src/libexpr/primops/fetchMercurial.cc

@@ -65,23 +65,23 @@ static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * ar
     attrs.insert_or_assign("url", url.find("://") != std::string::npos ? url : "file://" + url);
     if (ref) attrs.insert_or_assign("ref", *ref);
     if (rev) attrs.insert_or_assign("rev", rev->gitRev());
-    auto input = fetchers::inputFromAttrs(attrs);
+    auto input = fetchers::Input::fromAttrs(std::move(attrs));
 
     // FIXME: use name
-    auto [tree, input2] = input->fetchTree(state.store);
+    auto [tree, input2] = input.fetch(state.store);
 
     state.mkAttrs(v, 8);
     auto storePath = state.store->printStorePath(tree.storePath);
     mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
-    if (input2->getRef())
-        mkString(*state.allocAttr(v, state.symbols.create("branch")), *input2->getRef());
+    if (input2.getRef())
+        mkString(*state.allocAttr(v, state.symbols.create("branch")), *input2.getRef());
     // Backward compatibility: set 'rev' to
     // 0000000000000000000000000000000000000000 for a dirty tree.
-    auto rev2 = input2->getRev().value_or(Hash(htSHA1));
+    auto rev2 = input2.getRev().value_or(Hash(htSHA1));
     mkString(*state.allocAttr(v, state.symbols.create("rev")), rev2.gitRev());
     mkString(*state.allocAttr(v, state.symbols.create("shortRev")), std::string(rev2.gitRev(), 0, 12));
-    if (tree.info.revCount)
-        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
+    if (auto revCount = input2.getRevCount())
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *revCount);
     v.attrs->sort();
 
     if (state.allowedPaths)

src/libexpr/primops/fetchTree.cc

@@ -3,6 +3,7 @@
 #include "store-api.hh"
 #include "fetchers.hh"
 #include "filetransfer.hh"
+#include "registry.hh"
 
 #include <ctime>
 #include <iomanip>
@@ -12,39 +13,73 @@ namespace nix {
 void emitTreeAttrs(
     EvalState & state,
     const fetchers::Tree & tree,
-    std::shared_ptr<const fetchers::Input> input,
-    Value & v)
+    const fetchers::Input & input,
+    Value & v,
+    bool emptyRevFallback)
 {
+    assert(input.isImmutable());
+
     state.mkAttrs(v, 8);
 
     auto storePath = state.store->printStorePath(tree.storePath);
 
     mkString(*state.allocAttr(v, state.sOutPath), storePath, PathSet({storePath}));
 
-    assert(tree.info.narHash);
-    mkString(*state.allocAttr(v, state.symbols.create("narHash")),
-        tree.info.narHash.to_string(SRI, true));
+    // FIXME: support arbitrary input attributes.
 
-    if (input->getRev()) {
-        mkString(*state.allocAttr(v, state.symbols.create("rev")), input->getRev()->gitRev());
-        mkString(*state.allocAttr(v, state.symbols.create("shortRev")), input->getRev()->gitShortRev());
+    auto narHash = input.getNarHash();
+    assert(narHash);
+    mkString(*state.allocAttr(v, state.symbols.create("narHash")),
+        narHash->to_string(SRI, true));
+
+    if (auto rev = input.getRev()) {
+        mkString(*state.allocAttr(v, state.symbols.create("rev")), rev->gitRev());
+        mkString(*state.allocAttr(v, state.symbols.create("shortRev")), rev->gitShortRev());
+    } else if (emptyRevFallback) {
+        // Backwards compat for `builtins.fetchGit`: dirty repos return an empty sha1 as rev
+        auto emptyHash = Hash(htSHA1);
+        mkString(*state.allocAttr(v, state.symbols.create("rev")), emptyHash.gitRev());
+        mkString(*state.allocAttr(v, state.symbols.create("shortRev")), emptyHash.gitRev());
     }
 
-    if (tree.info.revCount)
-        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
+    if (input.getType() == "git")
+        mkBool(*state.allocAttr(v, state.symbols.create("submodules")), maybeGetBoolAttr(input.attrs, "submodules").value_or(false));
 
-    if (tree.info.lastModified)
-        mkString(*state.allocAttr(v, state.symbols.create("lastModified")),
-            fmt("%s", std::put_time(std::gmtime(&*tree.info.lastModified), "%Y%m%d%H%M%S")));
+    if (auto revCount = input.getRevCount())
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *revCount);
+    else if (emptyRevFallback)
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), 0);
+
+    if (auto lastModified = input.getLastModified()) {
+        mkInt(*state.allocAttr(v, state.symbols.create("lastModified")), *lastModified);
+        mkString(*state.allocAttr(v, state.symbols.create("lastModifiedDate")),
+            fmt("%s", std::put_time(std::gmtime(&*lastModified), "%Y%m%d%H%M%S")));
+    }
 
     v.attrs->sort();
 }
 
-static void prim_fetchTree(EvalState & state, const Pos & pos, Value * * args, Value & v)
+std::string fixURI(std::string uri, EvalState &state)
 {
-    settings.requireExperimentalFeature("flakes");
+    state.checkURI(uri);
+    return uri.find("://") != std::string::npos ? uri : "file://" + uri;
+}
 
-    std::shared_ptr<const fetchers::Input> input;
+void addURI(EvalState &state, fetchers::Attrs &attrs, Symbol name, std::string v)
+{
+    string n(name);
+    attrs.emplace(name, n == "url" ? fixURI(v, state) : v);
+}
+
+static void fetchTree(
+    EvalState &state,
+    const Pos &pos,
+    Value **args,
+    Value &v,
+    const std::optional<std::string> type,
+    bool emptyRevFallback = false
+) {
+    fetchers::Input input;
     PathSet context;
 
     state.forceValue(*args[0]);
@@ -56,35 +91,65 @@ static void prim_fetchTree(EvalState & state, const Pos & pos, Value * * args, V
 
         for (auto & attr : *args[0]->attrs) {
             state.forceValue(*attr.value);
-            if (attr.value->type == tString)
-                attrs.emplace(attr.name, attr.value->string.s);
+            if (attr.value->type == tPath || attr.value->type == tString)
+                addURI(
+                    state,
+                    attrs,
+                    attr.name,
+                    state.coerceToString(*attr.pos, *attr.value, context, false, false)
+                );
+            else if (attr.value->type == tString)
+                addURI(state, attrs, attr.name, attr.value->string.s);
             else if (attr.value->type == tBool)
-                attrs.emplace(attr.name, attr.value->boolean);
+                attrs.emplace(attr.name, fetchers::Explicit<bool>{attr.value->boolean});
+            else if (attr.value->type == tInt)
+                attrs.emplace(attr.name, attr.value->integer);
             else
-                throw TypeError("fetchTree argument '%s' is %s while a string or Boolean is expected",
+                throw TypeError("fetchTree argument '%s' is %s while a string, Boolean or integer is expected",
                     attr.name, showType(*attr.value));
         }
 
+        if (type)
+            attrs.emplace("type", type.value());
+
         if (!attrs.count("type"))
             throw Error({
                 .hint = hintfmt("attribute 'type' is missing in call to 'fetchTree'"),
                 .errPos = pos
             });
 
-        input = fetchers::inputFromAttrs(attrs);
-    } else
-        input = fetchers::inputFromURL(state.coerceToString(pos, *args[0], context, false, false));
+        input = fetchers::Input::fromAttrs(std::move(attrs));
+    } else {
+        auto url = fixURI(state.coerceToString(pos, *args[0], context, false, false), state);
 
-    if (evalSettings.pureEval && !input->isImmutable())
-        throw Error("in pure evaluation mode, 'fetchTree' requires an immutable input");
+        if (type == "git") {
+            fetchers::Attrs attrs;
+            attrs.emplace("type", "git");
+            attrs.emplace("url", url);
+            input = fetchers::Input::fromAttrs(std::move(attrs));
+        } else {
+            input = fetchers::Input::fromURL(url);
+        }
+    }
 
-    // FIXME: use fetchOrSubstituteTree
-    auto [tree, input2] = input->fetchTree(state.store);
+    if (!evalSettings.pureEval && !input.isDirect())
+        input = lookupInRegistries(state.store, input).first;
+
+    if (evalSettings.pureEval && !input.isImmutable())
+        throw Error("in pure evaluation mode, 'fetchTree' requires an immutable input, at %s", pos);
+
+    auto [tree, input2] = input.fetch(state.store);
 
     if (state.allowedPaths)
         state.allowedPaths->insert(tree.actualPath);
 
-    emitTreeAttrs(state, tree, input2, v);
+    emitTreeAttrs(state, tree, input2, v, emptyRevFallback);
+}
+
+static void prim_fetchTree(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    settings.requireExperimentalFeature("flakes");
+    fetchTree(state, pos, args, v, std::nullopt);
 }
 
 static RegisterPrimOp r("fetchTree", 1, prim_fetchTree);
@@ -136,7 +201,7 @@ static void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
 
     auto storePath =
         unpack
-        ? fetchers::downloadTarball(state.store, *url, name, (bool) expectedHash).storePath
+        ? fetchers::downloadTarball(state.store, *url, name, (bool) expectedHash).first.storePath
         : fetchers::downloadFile(state.store, *url, name, (bool) expectedHash).storePath;
 
     auto path = state.store->toRealPath(storePath);
@@ -147,7 +212,7 @@ static void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
             : hashFile(htSHA256, path);
         if (hash != *expectedHash)
             throw Error((unsigned int) 102, "hash mismatch in file downloaded from '%s':\n  wanted: %s\n  got:    %s",
-                *url, expectedHash->to_string(Base32, true), hash.to_string(Base32, true));
+                *url, expectedHash->to_string(Base32, true), hash->to_string(Base32, true));
     }
 
     if (state.allowedPaths)
@@ -166,7 +231,13 @@ static void prim_fetchTarball(EvalState & state, const Pos & pos, Value * * args
     fetch(state, pos, args, v, "fetchTarball", true, "source");
 }
 
+static void prim_fetchGit(EvalState &state, const Pos &pos, Value **args, Value &v)
+{
+    fetchTree(state, pos, args, v, "git", true);
+}
+
 static RegisterPrimOp r2("__fetchurl", 1, prim_fetchurl);
 static RegisterPrimOp r3("fetchTarball", 1, prim_fetchTarball);
+static RegisterPrimOp r4("fetchGit", 1, prim_fetchGit);
 
 }

src/libexpr/symbol-table.hh

@@ -28,6 +28,12 @@ public:
         return s == s2.s;
     }
 
+    // FIXME: remove
+    bool operator == (std::string_view s2) const
+    {
+        return s->compare(s2) == 0;
+    }
+
     bool operator != (const Symbol & s2) const
     {
         return s != s2.s;
@@ -68,9 +74,10 @@ private:
     Symbols symbols;
 
 public:
-    Symbol create(const string & s)
+    Symbol create(std::string_view s)
     {
-        std::pair<Symbols::iterator, bool> res = symbols.insert(s);
+        // FIXME: avoid allocation if 's' already exists in the symbol table.
+        std::pair<Symbols::iterator, bool> res = symbols.emplace(std::string(s));
         return Symbol(&*res.first);
     }
 

src/libexpr/value.hh

@@ -166,6 +166,13 @@ struct Value
     {
         return type == tList1 ? 1 : type == tList2 ? 2 : bigList.size;
     }
+
+    /* Check whether forcing this value requires a trivial amount of
+       computation. In particular, function applications are
+       non-trivial. */
+    bool isTrivial() const;
+
+    std::vector<std::pair<Path, std::string>> getContext();
 };
 
 

src/libfetchers/attrs.cc

@@ -27,7 +27,7 @@ nlohmann::json attrsToJson(const Attrs & attrs)
 {
     nlohmann::json json;
     for (auto & attr : attrs) {
-        if (auto v = std::get_if<int64_t>(&attr.second)) {
+        if (auto v = std::get_if<uint64_t>(&attr.second)) {
             json[attr.first] = *v;
         } else if (auto v = std::get_if<std::string>(&attr.second)) {
             json[attr.first] = *v;
@@ -55,16 +55,16 @@ std::string getStrAttr(const Attrs & attrs, const std::string & name)
     return *s;
 }
 
-std::optional<int64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name)
+std::optional<uint64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name)
 {
     auto i = attrs.find(name);
     if (i == attrs.end()) return {};
-    if (auto v = std::get_if<int64_t>(&i->second))
+    if (auto v = std::get_if<uint64_t>(&i->second))
         return *v;
     throw Error("input attribute '%s' is not an integer", name);
 }
 
-int64_t getIntAttr(const Attrs & attrs, const std::string & name)
+uint64_t getIntAttr(const Attrs & attrs, const std::string & name)
 {
     auto s = maybeGetIntAttr(attrs, name);
     if (!s)
@@ -76,8 +76,8 @@ std::optional<bool> maybeGetBoolAttr(const Attrs & attrs, const std::string & na
 {
     auto i = attrs.find(name);
     if (i == attrs.end()) return {};
-    if (auto v = std::get_if<int64_t>(&i->second))
-        return *v;
+    if (auto v = std::get_if<Explicit<bool>>(&i->second))
+        return v->t;
     throw Error("input attribute '%s' is not a Boolean", name);
 }
 
@@ -93,7 +93,7 @@ std::map<std::string, std::string> attrsToQuery(const Attrs & attrs)
 {
     std::map<std::string, std::string> query;
     for (auto & attr : attrs) {
-        if (auto v = std::get_if<int64_t>(&attr.second)) {
+        if (auto v = std::get_if<uint64_t>(&attr.second)) {
             query.insert_or_assign(attr.first, fmt("%d", *v));
         } else if (auto v = std::get_if<std::string>(&attr.second)) {
             query.insert_or_assign(attr.first, *v);

src/libfetchers/attrs.hh

@@ -13,9 +13,14 @@ namespace nix::fetchers {
 template<typename T>
 struct Explicit {
     T t;
+
+    bool operator ==(const Explicit<T> & other) const
+    {
+        return t == other.t;
+    }
 };
 
-typedef std::variant<std::string, int64_t, Explicit<bool>> Attr;
+typedef std::variant<std::string, uint64_t, Explicit<bool>> Attr;
 typedef std::map<std::string, Attr> Attrs;
 
 Attrs jsonToAttrs(const nlohmann::json & json);
@@ -26,9 +31,9 @@ std::optional<std::string> maybeGetStrAttr(const Attrs & attrs, const std::strin
 
 std::string getStrAttr(const Attrs & attrs, const std::string & name);
 
-std::optional<int64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name);
+std::optional<uint64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name);
 
-int64_t getIntAttr(const Attrs & attrs, const std::string & name);
+uint64_t getIntAttr(const Attrs & attrs, const std::string & name);
 
 std::optional<bool> maybeGetBoolAttr(const Attrs & attrs, const std::string & name);
 

src/libfetchers/fetchers.cc

@@ -5,71 +5,265 @@
 
 namespace nix::fetchers {
 
-std::unique_ptr<std::vector<std::unique_ptr<InputScheme>>> inputSchemes = nullptr;
+std::unique_ptr<std::vector<std::shared_ptr<InputScheme>>> inputSchemes = nullptr;
 
-void registerInputScheme(std::unique_ptr<InputScheme> && inputScheme)
+void registerInputScheme(std::shared_ptr<InputScheme> && inputScheme)
 {
-    if (!inputSchemes) inputSchemes = std::make_unique<std::vector<std::unique_ptr<InputScheme>>>();
+    if (!inputSchemes) inputSchemes = std::make_unique<std::vector<std::shared_ptr<InputScheme>>>();
     inputSchemes->push_back(std::move(inputScheme));
 }
 
-std::unique_ptr<Input> inputFromURL(const ParsedURL & url)
+Input Input::fromURL(const std::string & url)
+{
+    return fromURL(parseURL(url));
+}
+
+static void fixupInput(Input & input)
+{
+    // Check common attributes.
+    input.getType();
+    input.getRef();
+    if (input.getRev())
+        input.immutable = true;
+    input.getRevCount();
+    input.getLastModified();
+    if (input.getNarHash())
+        input.immutable = true;
+}
+
+Input Input::fromURL(const ParsedURL & url)
 {
     for (auto & inputScheme : *inputSchemes) {
         auto res = inputScheme->inputFromURL(url);
-        if (res) return res;
+        if (res) {
+            res->scheme = inputScheme;
+            fixupInput(*res);
+            return std::move(*res);
+        }
     }
+
     throw Error("input '%s' is unsupported", url.url);
 }
 
-std::unique_ptr<Input> inputFromURL(const std::string & url)
+Input Input::fromAttrs(Attrs && attrs)
 {
-    return inputFromURL(parseURL(url));
-}
-
-std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs)
-{
-    auto attrs2(attrs);
-    attrs2.erase("narHash");
     for (auto & inputScheme : *inputSchemes) {
-        auto res = inputScheme->inputFromAttrs(attrs2);
+        auto res = inputScheme->inputFromAttrs(attrs);
         if (res) {
-            if (auto narHash = maybeGetStrAttr(attrs, "narHash"))
-                // FIXME: require SRI hash.
-                res->narHash = newHashAllowEmpty(*narHash, {});
-            return res;
+            res->scheme = inputScheme;
+            fixupInput(*res);
+            return std::move(*res);
         }
     }
-    throw Error("input '%s' is unsupported", attrsToJson(attrs));
+
+    Input input;
+    input.attrs = attrs;
+    fixupInput(input);
+    return input;
+}
+
+ParsedURL Input::toURL() const
+{
+    if (!scheme)
+        throw Error("cannot show unsupported input '%s'", attrsToJson(attrs));
+    return scheme->toURL(*this);
+}
+
+std::string Input::to_string() const
+{
+    return toURL().to_string();
 }
 
 Attrs Input::toAttrs() const
 {
-    auto attrs = toAttrsInternal();
-    if (narHash)
-        attrs.emplace("narHash", narHash->to_string(SRI, true));
-    attrs.emplace("type", type());
     return attrs;
 }
 
-std::pair<Tree, std::shared_ptr<const Input>> Input::fetchTree(ref<Store> store) const
+bool Input::hasAllInfo() const
 {
-    auto [tree, input] = fetchTreeInternal(store);
+    return getNarHash() && scheme && scheme->hasAllInfo(*this);
+}
+
+bool Input::operator ==(const Input & other) const
+{
+    return attrs == other.attrs;
+}
+
+bool Input::contains(const Input & other) const
+{
+    if (*this == other) return true;
+    auto other2(other);
+    other2.attrs.erase("ref");
+    other2.attrs.erase("rev");
+    if (*this == other2) return true;
+    return false;
+}
+
+std::pair<Tree, Input> Input::fetch(ref<Store> store) const
+{
+    if (!scheme)
+        throw Error("cannot fetch unsupported input '%s'", attrsToJson(toAttrs()));
+
+    /* The tree may already be in the Nix store, or it could be
+       substituted (which is often faster than fetching from the
+       original source). So check that. */
+    if (hasAllInfo()) {
+        try {
+            auto storePath = computeStorePath(*store);
+
+            store->ensurePath(storePath);
+
+            debug("using substituted/cached input '%s' in '%s'",
+                to_string(), store->printStorePath(storePath));
+
+            auto actualPath = store->toRealPath(storePath);
+
+            return {fetchers::Tree(std::move(actualPath), std::move(storePath)), *this};
+        } catch (Error & e) {
+            debug("substitution of input '%s' failed: %s", to_string(), e.what());
+        }
+    }
+
+    auto [tree, input] = scheme->fetch(store, *this);
 
     if (tree.actualPath == "")
         tree.actualPath = store->toRealPath(tree.storePath);
 
-    if (!tree.info.narHash)
-        tree.info.narHash = store->queryPathInfo(tree.storePath)->narHash;
+    auto narHash = store->queryPathInfo(tree.storePath)->narHash;
+    input.attrs.insert_or_assign("narHash", narHash->to_string(SRI, true));
 
-    if (input->narHash)
-        assert(input->narHash == tree.info.narHash);
+    if (auto prevNarHash = getNarHash()) {
+        if (narHash != *prevNarHash)
+            throw Error((unsigned int) 102, "NAR hash mismatch in input '%s' (%s), expected '%s', got '%s'",
+                to_string(), tree.actualPath, prevNarHash->to_string(SRI, true), narHash->to_string(SRI, true));
+    }
 
-    if (narHash && narHash != input->narHash)
-        throw Error("NAR hash mismatch in input '%s' (%s), expected '%s', got '%s'",
-            to_string(), tree.actualPath, narHash->to_string(SRI, true), input->narHash->to_string(SRI, true));
+    if (auto prevLastModified = getLastModified()) {
+        if (input.getLastModified() != prevLastModified)
+            throw Error("'lastModified' attribute mismatch in input '%s', expected %d",
+                input.to_string(), *prevLastModified);
+    }
+
+    if (auto prevRevCount = getRevCount()) {
+        if (input.getRevCount() != prevRevCount)
+            throw Error("'revCount' attribute mismatch in input '%s', expected %d",
+                input.to_string(), *prevRevCount);
+    }
+
+    input.immutable = true;
+
+    assert(input.hasAllInfo());
 
     return {std::move(tree), input};
 }
 
+Input Input::applyOverrides(
+    std::optional<std::string> ref,
+    std::optional<Hash> rev) const
+{
+    if (!scheme) return *this;
+    return scheme->applyOverrides(*this, ref, rev);
+}
+
+void Input::clone(const Path & destDir) const
+{
+    assert(scheme);
+    scheme->clone(*this, destDir);
+}
+
+std::optional<Path> Input::getSourcePath() const
+{
+    assert(scheme);
+    return scheme->getSourcePath(*this);
+}
+
+void Input::markChangedFile(
+    std::string_view file,
+    std::optional<std::string> commitMsg) const
+{
+    assert(scheme);
+    return scheme->markChangedFile(*this, file, commitMsg);
+}
+
+StorePath Input::computeStorePath(Store & store) const
+{
+    auto narHash = getNarHash();
+    if (!narHash)
+        throw Error("cannot compute store path for mutable input '%s'", to_string());
+    return store.makeFixedOutputPath(FileIngestionMethod::Recursive, *narHash, "source");
+}
+
+std::string Input::getType() const
+{
+    return getStrAttr(attrs, "type");
+}
+
+std::optional<Hash> Input::getNarHash() const
+{
+    if (auto s = maybeGetStrAttr(attrs, "narHash"))
+        // FIXME: require SRI hash.
+        return newHashAllowEmpty(*s, htSHA256);
+    return {};
+}
+
+std::optional<std::string> Input::getRef() const
+{
+    if (auto s = maybeGetStrAttr(attrs, "ref"))
+        return *s;
+    return {};
+}
+
+std::optional<Hash> Input::getRev() const
+{
+    if (auto s = maybeGetStrAttr(attrs, "rev"))
+        return Hash(*s, htSHA1);
+    return {};
+}
+
+std::optional<uint64_t> Input::getRevCount() const
+{
+    if (auto n = maybeGetIntAttr(attrs, "revCount"))
+        return *n;
+    return {};
+}
+
+std::optional<time_t> Input::getLastModified() const
+{
+    if (auto n = maybeGetIntAttr(attrs, "lastModified"))
+        return *n;
+    return {};
+}
+
+ParsedURL InputScheme::toURL(const Input & input)
+{
+    throw Error("don't know how to convert input '%s' to a URL", attrsToJson(input.attrs));
+}
+
+Input InputScheme::applyOverrides(
+    const Input & input,
+    std::optional<std::string> ref,
+    std::optional<Hash> rev)
+{
+    if (ref)
+        throw Error("don't know how to set branch/tag name of input '%s' to '%s'", input.to_string(), *ref);
+    if (rev)
+        throw Error("don't know how to set revision of input '%s' to '%s'", input.to_string(), rev->gitRev());
+    return input;
+}
+
+std::optional<Path> InputScheme::getSourcePath(const Input & input)
+{
+    return {};
+}
+
+void InputScheme::markChangedFile(const Input & input, std::string_view file, std::optional<std::string> commitMsg)
+{
+    assert(false);
+}
+
+void InputScheme::clone(const Input & input, const Path & destDir)
+{
+    throw Error("do not know how to clone input '%s'", input.to_string());
+}
+
 }

src/libfetchers/fetchers.hh

@@ -3,7 +3,6 @@
 #include "types.hh"
 #include "hash.hh"
 #include "path.hh"
-#include "tree-info.hh"
 #include "attrs.hh"
 #include "url.hh"
 
@@ -13,73 +12,101 @@ namespace nix { class Store; }
 
 namespace nix::fetchers {
 
-struct Input;
-
 struct Tree
 {
     Path actualPath;
     StorePath storePath;
-    TreeInfo info;
+    Tree(Path && actualPath, StorePath && storePath) : actualPath(actualPath), storePath(std::move(storePath)) {}
 };
 
-struct Input : std::enable_shared_from_this<Input>
+struct InputScheme;
+
+struct Input
 {
-    std::optional<Hash> narHash; // FIXME: implement
+    friend class InputScheme;
 
-    virtual std::string type() const = 0;
+    std::shared_ptr<InputScheme> scheme; // note: can be null
+    Attrs attrs;
+    bool immutable = false;
+    bool direct = true;
 
-    virtual ~Input() { }
+public:
+    static Input fromURL(const std::string & url);
 
-    virtual bool operator ==(const Input & other) const { return false; }
+    static Input fromURL(const ParsedURL & url);
 
-    /* Check whether this is a "direct" input, that is, not
-       one that goes through a registry. */
-    virtual bool isDirect() const { return true; }
+    static Input fromAttrs(Attrs && attrs);
 
-    /* Check whether this is an "immutable" input, that is,
-       one that contains a commit hash or content hash. */
-    virtual bool isImmutable() const { return (bool) narHash; }
+    ParsedURL toURL() const;
 
-    virtual bool contains(const Input & other) const { return false; }
-
-    virtual std::optional<std::string> getRef() const { return {}; }
-
-    virtual std::optional<Hash> getRev() const { return {}; }
-
-    virtual ParsedURL toURL() const = 0;
-
-    std::string to_string() const
-    {
-        return toURL().to_string();
-    }
+    std::string to_string() const;
 
     Attrs toAttrs() const;
 
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTree(ref<Store> store) const;
+    /* Check whether this is a "direct" input, that is, not
+       one that goes through a registry. */
+    bool isDirect() const { return direct; }
 
-private:
+    /* Check whether this is an "immutable" input, that is,
+       one that contains a commit hash or content hash. */
+    bool isImmutable() const { return immutable; }
 
-    virtual std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(ref<Store> store) const = 0;
+    bool hasAllInfo() const;
 
-    virtual Attrs toAttrsInternal() const = 0;
+    bool operator ==(const Input & other) const;
+
+    bool contains(const Input & other) const;
+
+    std::pair<Tree, Input> fetch(ref<Store> store) const;
+
+    Input applyOverrides(
+        std::optional<std::string> ref,
+        std::optional<Hash> rev) const;
+
+    void clone(const Path & destDir) const;
+
+    std::optional<Path> getSourcePath() const;
+
+    void markChangedFile(
+        std::string_view file,
+        std::optional<std::string> commitMsg) const;
+
+    StorePath computeStorePath(Store & store) const;
+
+    // Convience functions for common attributes.
+    std::string getType() const;
+    std::optional<Hash> getNarHash() const;
+    std::optional<std::string> getRef() const;
+    std::optional<Hash> getRev() const;
+    std::optional<uint64_t> getRevCount() const;
+    std::optional<time_t> getLastModified() const;
 };
 
 struct InputScheme
 {
-    virtual ~InputScheme() { }
+    virtual std::optional<Input> inputFromURL(const ParsedURL & url) = 0;
 
-    virtual std::unique_ptr<Input> inputFromURL(const ParsedURL & url) = 0;
+    virtual std::optional<Input> inputFromAttrs(const Attrs & attrs) = 0;
 
-    virtual std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) = 0;
+    virtual ParsedURL toURL(const Input & input);
+
+    virtual bool hasAllInfo(const Input & input) = 0;
+
+    virtual Input applyOverrides(
+        const Input & input,
+        std::optional<std::string> ref,
+        std::optional<Hash> rev);
+
+    virtual void clone(const Input & input, const Path & destDir);
+
+    virtual std::optional<Path> getSourcePath(const Input & input);
+
+    virtual void markChangedFile(const Input & input, std::string_view file, std::optional<std::string> commitMsg);
+
+    virtual std::pair<Tree, Input> fetch(ref<Store> store, const Input & input) = 0;
 };
 
-std::unique_ptr<Input> inputFromURL(const ParsedURL & url);
-
-std::unique_ptr<Input> inputFromURL(const std::string & url);
-
-std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs);
-
-void registerInputScheme(std::unique_ptr<InputScheme> && fetcher);
+void registerInputScheme(std::shared_ptr<InputScheme> && fetcher);
 
 struct DownloadFileResult
 {
@@ -94,7 +121,7 @@ DownloadFileResult downloadFile(
     const std::string & name,
     bool immutable);
 
-Tree downloadTarball(
+std::pair<Tree, time_t> downloadTarball(
     ref<Store> store,
     const std::string & url,
     const std::string & name,

src/libfetchers/git.cc

@@ -22,80 +22,152 @@ static bool isNotDotGitDirectory(const Path & path)
     return not std::regex_match(path, gitDirRegex);
 }
 
-struct GitInput : Input
+struct GitInputScheme : InputScheme
 {
-    ParsedURL url;
-    std::optional<std::string> ref;
-    std::optional<Hash> rev;
-    bool shallow = false;
-    bool submodules = false;
-
-    GitInput(const ParsedURL & url) : url(url)
-    { }
-
-    std::string type() const override { return "git"; }
-
-    bool operator ==(const Input & other) const override
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
     {
-        auto other2 = dynamic_cast<const GitInput *>(&other);
-        return
-            other2
-            && url == other2->url
-            && rev == other2->rev
-            && ref == other2->ref;
-    }
+        if (url.scheme != "git" &&
+            url.scheme != "git+http" &&
+            url.scheme != "git+https" &&
+            url.scheme != "git+ssh" &&
+            url.scheme != "git+file") return {};
 
-    bool isImmutable() const override
-    {
-        return (bool) rev || narHash;
-    }
+        auto url2(url);
+        if (hasPrefix(url2.scheme, "git+")) url2.scheme = std::string(url2.scheme, 4);
+        url2.query.clear();
 
-    std::optional<std::string> getRef() const override { return ref; }
-
-    std::optional<Hash> getRev() const override { return rev; }
-
-    ParsedURL toURL() const override
-    {
-        ParsedURL url2(url);
-        if (url2.scheme != "git") url2.scheme = "git+" + url2.scheme;
-        if (rev) url2.query.insert_or_assign("rev", rev->gitRev());
-        if (ref) url2.query.insert_or_assign("ref", *ref);
-        if (shallow) url2.query.insert_or_assign("shallow", "1");
-        return url2;
-    }
-
-    Attrs toAttrsInternal() const override
-    {
         Attrs attrs;
-        attrs.emplace("url", url.to_string());
-        if (ref)
-            attrs.emplace("ref", *ref);
-        if (rev)
-            attrs.emplace("rev", rev->gitRev());
-        if (shallow)
-            attrs.emplace("shallow", true);
-        if (submodules)
-            attrs.emplace("submodules", true);
-        return attrs;
+        attrs.emplace("type", "git");
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev" || name == "ref")
+                attrs.emplace(name, value);
+            else if (name == "shallow")
+                attrs.emplace(name, Explicit<bool> { value == "1" });
+            else
+                url2.query.emplace(name, value);
+        }
+
+        attrs.emplace("url", url2.to_string());
+
+        return inputFromAttrs(attrs);
     }
 
-    std::pair<bool, std::string> getActualUrl() const
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "git") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "ref" && name != "rev" && name != "shallow" && name != "submodules" && name != "lastModified" && name != "revCount" && name != "narHash")
+                throw Error("unsupported Git input attribute '%s'", name);
+
+        parseURL(getStrAttr(attrs, "url"));
+        maybeGetBoolAttr(attrs, "shallow");
+        maybeGetBoolAttr(attrs, "submodules");
+
+        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
+            if (std::regex_search(*ref, badGitRefRegex))
+                throw BadURL("invalid Git branch/tag name '%s'", *ref);
+        }
+
+        Input input;
+        input.attrs = attrs;
+        return input;
+    }
+
+    ParsedURL toURL(const Input & input) override
+    {
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
+        if (url.scheme != "git") url.scheme = "git+" + url.scheme;
+        if (auto rev = input.getRev()) url.query.insert_or_assign("rev", rev->gitRev());
+        if (auto ref = input.getRef()) url.query.insert_or_assign("ref", *ref);
+        if (maybeGetBoolAttr(input.attrs, "shallow").value_or(false))
+            url.query.insert_or_assign("shallow", "1");
+        return url;
+    }
+
+    bool hasAllInfo(const Input & input) override
+    {
+        bool maybeDirty = !input.getRef();
+        bool shallow = maybeGetBoolAttr(input.attrs, "shallow").value_or(false);
+        return
+            maybeGetIntAttr(input.attrs, "lastModified")
+            && (shallow || maybeDirty || maybeGetIntAttr(input.attrs, "revCount"));
+    }
+
+    Input applyOverrides(
+        const Input & input,
+        std::optional<std::string> ref,
+        std::optional<Hash> rev) override
+    {
+        auto res(input);
+        if (rev) res.attrs.insert_or_assign("rev", rev->gitRev());
+        if (ref) res.attrs.insert_or_assign("ref", *ref);
+        if (!res.getRef() && res.getRev())
+            throw Error("Git input '%s' has a commit hash but no branch/tag name", res.to_string());
+        return res;
+    }
+
+    void clone(const Input & input, const Path & destDir) override
+    {
+        auto [isLocal, actualUrl] = getActualUrl(input);
+
+        Strings args = {"clone"};
+
+        args.push_back(actualUrl);
+
+        if (auto ref = input.getRef()) {
+            args.push_back("--branch");
+            args.push_back(*ref);
+        }
+
+        if (input.getRev()) throw Error("cloning a specific revision is not implemented");
+
+        args.push_back(destDir);
+
+        runProgram("git", true, args);
+    }
+
+    std::optional<Path> getSourcePath(const Input & input) override
+    {
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
+        if (url.scheme == "file" && !input.getRef() && !input.getRev())
+            return url.path;
+        return {};
+    }
+
+    void markChangedFile(const Input & input, std::string_view file, std::optional<std::string> commitMsg) override
+    {
+        auto sourcePath = getSourcePath(input);
+        assert(sourcePath);
+
+        runProgram("git", true,
+            { "-C", *sourcePath, "add", "--force", "--intent-to-add", "--", std::string(file) });
+
+        if (commitMsg)
+            runProgram("git", true,
+                { "-C", *sourcePath, "commit", std::string(file), "-m", *commitMsg });
+    }
+
+    std::pair<bool, std::string> getActualUrl(const Input & input) const
     {
         // Don't clone file:// URIs (but otherwise treat them the
         // same as remote URIs, i.e. don't use the working tree or
         // HEAD).
         static bool forceHttp = getEnv("_NIX_FORCE_HTTP") == "1"; // for testing
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
         bool isLocal = url.scheme == "file" && !forceHttp;
         return {isLocal, isLocal ? url.path : url.base};
     }
 
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & _input) override
     {
         auto name = "source";
 
-        auto input = std::make_shared<GitInput>(*this);
+        Input input(_input);
 
-        assert(!rev || rev->type == htSHA1);
+        bool shallow = maybeGetBoolAttr(input.attrs, "shallow").value_or(false);
+        bool submodules = maybeGetBoolAttr(input.attrs, "submodules").value_or(false);
 
         std::string cacheType = "git";
         if (shallow) cacheType += "-shallow";
@@ -106,39 +178,35 @@ struct GitInput : Input
             return Attrs({
                 {"type", cacheType},
                 {"name", name},
-                {"rev", input->rev->gitRev()},
+                {"rev", input.getRev()->gitRev()},
             });
         };
 
         auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
-            -> std::pair<Tree, std::shared_ptr<const Input>>
+            -> std::pair<Tree, Input>
         {
-            assert(input->rev);
-            assert(!rev || rev == input->rev);
+            assert(input.getRev());
+            assert(!_input.getRev() || _input.getRev() == input.getRev());
+            if (!shallow)
+                input.attrs.insert_or_assign("revCount", getIntAttr(infoAttrs, "revCount"));
+            input.attrs.insert_or_assign("lastModified", getIntAttr(infoAttrs, "lastModified"));
             return {
-                Tree {
-                    .actualPath = store->toRealPath(storePath),
-                    .storePath = std::move(storePath),
-                    .info = TreeInfo {
-                        .revCount = shallow ? std::nullopt : std::optional(getIntAttr(infoAttrs, "revCount")),
-                        .lastModified = getIntAttr(infoAttrs, "lastModified"),
-                    },
-                },
+                Tree(store->toRealPath(storePath), std::move(storePath)),
                 input
             };
         };
 
-        if (rev) {
+        if (input.getRev()) {
             if (auto res = getCache()->lookup(store, getImmutableAttrs()))
                 return makeResult(res->first, std::move(res->second));
         }
 
-        auto [isLocal, actualUrl_] = getActualUrl();
+        auto [isLocal, actualUrl_] = getActualUrl(input);
         auto actualUrl = actualUrl_; // work around clang bug
 
         // If this is a local directory and no ref or revision is
         // given, then allow the use of an unclean working tree.
-        if (!input->ref && !input->rev && isLocal) {
+        if (!input.getRef() && !input.getRev() && isLocal) {
             bool clean = false;
 
             /* Check whether this repo has any commits. There are
@@ -197,35 +265,35 @@ struct GitInput : Input
 
                 auto storePath = store->addToStore("source", actualUrl, FileIngestionMethod::Recursive, htSHA256, filter);
 
-                auto tree = Tree {
-                    .actualPath = store->printStorePath(storePath),
-                    .storePath = std::move(storePath),
-                    .info = TreeInfo {
-                        // FIXME: maybe we should use the timestamp of the last
-                        // modified dirty file?
-                        .lastModified = haveCommits ? std::stoull(runProgram("git", true, { "-C", actualUrl, "log", "-1", "--format=%ct", "HEAD" })) : 0,
-                    }
-                };
+                // FIXME: maybe we should use the timestamp of the last
+                // modified dirty file?
+                input.attrs.insert_or_assign(
+                    "lastModified",
+                    haveCommits ? std::stoull(runProgram("git", true, { "-C", actualUrl, "log", "-1", "--format=%ct", "HEAD" })) : 0);
 
-                return {std::move(tree), input};
+                return {
+                    Tree(store->printStorePath(storePath), std::move(storePath)),
+                    input
+                };
             }
         }
 
-        if (!input->ref) input->ref = isLocal ? readHead(actualUrl) : "master";
+        if (!input.getRef()) input.attrs.insert_or_assign("ref", isLocal ? readHead(actualUrl) : "master");
 
         Attrs mutableAttrs({
             {"type", cacheType},
             {"name", name},
             {"url", actualUrl},
-            {"ref", *input->ref},
+            {"ref", *input.getRef()},
         });
 
         Path repoDir;
 
         if (isLocal) {
 
-            if (!input->rev)
-                input->rev = Hash(chomp(runProgram("git", true, { "-C", actualUrl, "rev-parse", *input->ref })), htSHA1);
+            if (!input.getRev())
+                input.attrs.insert_or_assign("rev",
+                    Hash(chomp(runProgram("git", true, { "-C", actualUrl, "rev-parse", *input.getRef() })), htSHA1).gitRev());
 
             repoDir = actualUrl;
 
@@ -233,8 +301,8 @@ struct GitInput : Input
 
             if (auto res = getCache()->lookup(store, mutableAttrs)) {
                 auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
-                if (!rev || rev == rev2) {
-                    input->rev = rev2;
+                if (!input.getRev() || input.getRev() == rev2) {
+                    input.attrs.insert_or_assign("rev", rev2.gitRev());
                     return makeResult(res->first, std::move(res->second));
                 }
             }
@@ -248,18 +316,18 @@ struct GitInput : Input
             }
 
             Path localRefFile =
-                input->ref->compare(0, 5, "refs/") == 0
-                ? cacheDir + "/" + *input->ref
-                : cacheDir + "/refs/heads/" + *input->ref;
+                input.getRef()->compare(0, 5, "refs/") == 0
+                ? cacheDir + "/" + *input.getRef()
+                : cacheDir + "/refs/heads/" + *input.getRef();
 
             bool doFetch;
             time_t now = time(0);
 
             /* If a rev was specified, we need to fetch if it's not in the
                repo. */
-            if (input->rev) {
+            if (input.getRev()) {
                 try {
-                    runProgram("git", true, { "-C", repoDir, "cat-file", "-e", input->rev->gitRev() });
+                    runProgram("git", true, { "-C", repoDir, "cat-file", "-e", input.getRev()->gitRev() });
                     doFetch = false;
                 } catch (ExecError & e) {
                     if (WIFEXITED(e.status)) {
@@ -282,9 +350,10 @@ struct GitInput : Input
                 // FIXME: git stderr messes up our progress indicator, so
                 // we're using --quiet for now. Should process its stderr.
                 try {
-                    auto fetchRef = input->ref->compare(0, 5, "refs/") == 0
-                        ? *input->ref
-                        : "refs/heads/" + *input->ref;
+                    auto ref = input.getRef();
+                    auto fetchRef = ref->compare(0, 5, "refs/") == 0
+                        ? *ref
+                        : "refs/heads/" + *ref;
                     runProgram("git", true, { "-C", repoDir, "fetch", "--quiet", "--force", "--", actualUrl, fmt("%s:%s", fetchRef, fetchRef) });
                 } catch (Error & e) {
                     if (!pathExists(localRefFile)) throw;
@@ -300,8 +369,8 @@ struct GitInput : Input
                 utimes(localRefFile.c_str(), times);
             }
 
-            if (!input->rev)
-                input->rev = Hash(chomp(readFile(localRefFile)), htSHA1);
+            if (!input.getRev())
+                input.attrs.insert_or_assign("rev", Hash(chomp(readFile(localRefFile)), htSHA1).gitRev());
         }
 
         bool isShallow = chomp(runProgram("git", true, { "-C", repoDir, "rev-parse", "--is-shallow-repository" })) == "true";
@@ -311,7 +380,7 @@ struct GitInput : Input
 
         // FIXME: check whether rev is an ancestor of ref.
 
-        printTalkative("using revision %s of repo '%s'", input->rev->gitRev(), actualUrl);
+        printTalkative("using revision %s of repo '%s'", input.getRev()->gitRev(), actualUrl);
 
         /* Now that we know the ref, check again whether we have it in
            the store. */
@@ -333,7 +402,7 @@ struct GitInput : Input
             runProgram("git", true, { "-C", tmpDir, "fetch", "--quiet", "--force",
                                       "--update-head-ok", "--", repoDir, "refs/*:refs/*" });
 
-            runProgram("git", true, { "-C", tmpDir, "checkout", "--quiet", input->rev->gitRev() });
+            runProgram("git", true, { "-C", tmpDir, "checkout", "--quiet", input.getRev()->gitRev() });
             runProgram("git", true, { "-C", tmpDir, "remote", "add", "origin", actualUrl });
             runProgram("git", true, { "-C", tmpDir, "submodule", "--quiet", "update", "--init", "--recursive" });
 
@@ -342,7 +411,7 @@ struct GitInput : Input
             // FIXME: should pipe this, or find some better way to extract a
             // revision.
             auto source = sinkToSource([&](Sink & sink) {
-                RunOptions gitOptions("git", { "-C", repoDir, "archive", input->rev->gitRev() });
+                RunOptions gitOptions("git", { "-C", repoDir, "archive", input.getRev()->gitRev() });
                 gitOptions.standardOut = &sink;
                 runProgram2(gitOptions);
             });
@@ -352,18 +421,18 @@ struct GitInput : Input
 
         auto storePath = store->addToStore(name, tmpDir, FileIngestionMethod::Recursive, htSHA256, filter);
 
-        auto lastModified = std::stoull(runProgram("git", true, { "-C", repoDir, "log", "-1", "--format=%ct", input->rev->gitRev() }));
+        auto lastModified = std::stoull(runProgram("git", true, { "-C", repoDir, "log", "-1", "--format=%ct", input.getRev()->gitRev() }));
 
         Attrs infoAttrs({
-            {"rev", input->rev->gitRev()},
+            {"rev", input.getRev()->gitRev()},
             {"lastModified", lastModified},
         });
 
         if (!shallow)
             infoAttrs.insert_or_assign("revCount",
-                std::stoull(runProgram("git", true, { "-C", repoDir, "rev-list", "--count", input->rev->gitRev() })));
+                std::stoull(runProgram("git", true, { "-C", repoDir, "rev-list", "--count", input.getRev()->gitRev() })));
 
-        if (!this->rev)
+        if (!_input.getRev())
             getCache()->add(
                 store,
                 mutableAttrs,
@@ -382,60 +451,6 @@ struct GitInput : Input
     }
 };
 
-struct GitInputScheme : InputScheme
-{
-    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
-    {
-        if (url.scheme != "git" &&
-            url.scheme != "git+http" &&
-            url.scheme != "git+https" &&
-            url.scheme != "git+ssh" &&
-            url.scheme != "git+file") return nullptr;
-
-        auto url2(url);
-        if (hasPrefix(url2.scheme, "git+")) url2.scheme = std::string(url2.scheme, 4);
-        url2.query.clear();
-
-        Attrs attrs;
-        attrs.emplace("type", "git");
-
-        for (auto &[name, value] : url.query) {
-            if (name == "rev" || name == "ref")
-                attrs.emplace(name, value);
-            else
-                url2.query.emplace(name, value);
-        }
-
-        attrs.emplace("url", url2.to_string());
-
-        return inputFromAttrs(attrs);
-    }
-
-    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
-    {
-        if (maybeGetStrAttr(attrs, "type") != "git") return {};
-
-        for (auto & [name, value] : attrs)
-            if (name != "type" && name != "url" && name != "ref" && name != "rev" && name != "shallow" && name != "submodules")
-                throw Error("unsupported Git input attribute '%s'", name);
-
-        auto input = std::make_unique<GitInput>(parseURL(getStrAttr(attrs, "url")));
-        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
-            if (std::regex_search(*ref, badGitRefRegex))
-                throw BadURL("invalid Git branch/tag name '%s'", *ref);
-            input->ref = *ref;
-        }
-        if (auto rev = maybeGetStrAttr(attrs, "rev"))
-            input->rev = Hash(*rev, htSHA1);
-
-        input->shallow = maybeGetBoolAttr(attrs, "shallow").value_or(false);
-
-        input->submodules = maybeGetBoolAttr(attrs, "submodules").value_or(false);
-
-        return input;
-    }
-};
-
 static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<GitInputScheme>()); });
 
 }

src/libfetchers/github.cc

@@ -8,81 +8,142 @@
 
 namespace nix::fetchers {
 
-std::regex ownerRegex("[a-zA-Z][a-zA-Z0-9_-]*", std::regex::ECMAScript);
-std::regex repoRegex("[a-zA-Z][a-zA-Z0-9_-]*", std::regex::ECMAScript);
+// A github or gitlab url
+const static std::string urlRegexS = "[a-zA-Z0-9.]*"; // FIXME: check
+std::regex urlRegex(urlRegexS, std::regex::ECMAScript);
 
-struct GitHubInput : Input
+struct GitArchiveInputScheme : InputScheme
 {
-    std::string owner;
-    std::string repo;
-    std::optional<std::string> ref;
-    std::optional<Hash> rev;
+    virtual std::string type() = 0;
 
-    std::string type() const override { return "github"; }
-
-    bool operator ==(const Input & other) const override
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
     {
-        auto other2 = dynamic_cast<const GitHubInput *>(&other);
-        return
-            other2
-            && owner == other2->owner
-            && repo == other2->repo
-            && rev == other2->rev
-            && ref == other2->ref;
+        if (url.scheme != type()) return {};
+
+        auto path = tokenizeString<std::vector<std::string>>(url.path, "/");
+
+        std::optional<Hash> rev;
+        std::optional<std::string> ref;
+        std::optional<std::string> host_url;
+
+        if (path.size() == 2) {
+        } else if (path.size() == 3) {
+            if (std::regex_match(path[2], revRegex))
+                rev = Hash(path[2], htSHA1);
+            else if (std::regex_match(path[2], refRegex))
+                ref = path[2];
+            else
+                throw BadURL("in URL '%s', '%s' is not a commit hash or branch/tag name", url.url, path[2]);
+        } else
+            throw BadURL("URL '%s' is invalid", url.url);
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev") {
+                if (rev)
+                    throw BadURL("URL '%s' contains multiple commit hashes", url.url);
+                rev = Hash(value, htSHA1);
+            }
+            else if (name == "ref") {
+                if (!std::regex_match(value, refRegex))
+                    throw BadURL("URL '%s' contains an invalid branch/tag name", url.url);
+                if (ref)
+                    throw BadURL("URL '%s' contains multiple branch/tag names", url.url);
+                ref = value;
+            }
+            else if (name == "url") {
+                if (!std::regex_match(value, urlRegex))
+                    throw BadURL("URL '%s' contains an invalid instance url", url.url);
+                host_url = value;
+            }
+            // FIXME: barf on unsupported attributes
+        }
+
+        if (ref && rev)
+            throw BadURL("URL '%s' contains both a commit hash and a branch/tag name %s %s", url.url, *ref, rev->gitRev());
+
+        Input input;
+        input.attrs.insert_or_assign("type", type());
+        input.attrs.insert_or_assign("owner", path[0]);
+        input.attrs.insert_or_assign("repo", path[1]);
+        if (rev) input.attrs.insert_or_assign("rev", rev->gitRev());
+        if (ref) input.attrs.insert_or_assign("ref", *ref);
+        if (host_url) input.attrs.insert_or_assign("url", *host_url);
+
+        return input;
     }
 
-    bool isImmutable() const override
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
     {
-        return (bool) rev || narHash;
+        if (maybeGetStrAttr(attrs, "type") != type()) return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "owner" && name != "repo" && name != "ref" && name != "rev" && name != "narHash" && name != "lastModified")
+                throw Error("unsupported input attribute '%s'", name);
+
+        getStrAttr(attrs, "owner");
+        getStrAttr(attrs, "repo");
+
+        Input input;
+        input.attrs = attrs;
+        return input;
     }
 
-    std::optional<std::string> getRef() const override { return ref; }
-
-    std::optional<Hash> getRev() const override { return rev; }
-
-    ParsedURL toURL() const override
+    ParsedURL toURL(const Input & input) override
     {
+        auto owner = getStrAttr(input.attrs, "owner");
+        auto repo = getStrAttr(input.attrs, "repo");
+        auto ref = input.getRef();
+        auto rev = input.getRev();
         auto path = owner + "/" + repo;
         assert(!(ref && rev));
         if (ref) path += "/" + *ref;
         if (rev) path += "/" + rev->to_string(Base16, false);
         return ParsedURL {
-            .scheme = "github",
+            .scheme = type(),
             .path = path,
         };
     }
 
-    Attrs toAttrsInternal() const override
+    bool hasAllInfo(const Input & input) override
     {
-        Attrs attrs;
-        attrs.emplace("owner", owner);
-        attrs.emplace("repo", repo);
-        if (ref)
-            attrs.emplace("ref", *ref);
-        if (rev)
-            attrs.emplace("rev", rev->gitRev());
-        return attrs;
+        return input.getRev() && maybeGetIntAttr(input.attrs, "lastModified");
     }
 
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    Input applyOverrides(
+        const Input & _input,
+        std::optional<std::string> ref,
+        std::optional<Hash> rev) override
     {
-        auto rev = this->rev;
-        auto ref = this->ref.value_or("master");
-
-        if (!rev) {
-            auto url = fmt("https://api.github.com/repos/%s/%s/commits/%s",
-                owner, repo, ref);
-            auto json = nlohmann::json::parse(
-                readFile(
-                    store->toRealPath(
-                        downloadFile(store, url, "source", false).storePath)));
-            rev = Hash(std::string { json["sha"] }, htSHA1);
-            debug("HEAD revision for '%s' is %s", url, rev->gitRev());
+        auto input(_input);
+        if (rev && ref)
+            throw BadURL("cannot apply both a commit hash (%s) and a branch/tag name ('%s') to input '%s'",
+                rev->gitRev(), *ref, input.to_string());
+        if (rev) {
+            input.attrs.insert_or_assign("rev", rev->gitRev());
+            input.attrs.erase("ref");
         }
+        if (ref) {
+            input.attrs.insert_or_assign("ref", *ref);
+            input.attrs.erase("rev");
+        }
+        return input;
+    }
 
-        auto input = std::make_shared<GitHubInput>(*this);
-        input->ref = {};
-        input->rev = *rev;
+    virtual Hash getRevFromRef(nix::ref<Store> store, const Input & input) const = 0;
+
+    virtual std::string getDownloadUrl(const Input & input) const = 0;
+
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & _input) override
+    {
+        Input input(_input);
+
+        if (!maybeGetStrAttr(input.attrs, "ref")) input.attrs.insert_or_assign("ref", "HEAD");
+
+        auto rev = input.getRev();
+        if (!rev) rev = getRevFromRef(store, input);
+
+        input.attrs.erase("ref");
+        input.attrs.insert_or_assign("rev", rev->gitRev());
 
         Attrs immutableAttrs({
             {"type", "git-tarball"},
@@ -90,36 +151,25 @@ struct GitHubInput : Input
         });
 
         if (auto res = getCache()->lookup(store, immutableAttrs)) {
+            input.attrs.insert_or_assign("lastModified", getIntAttr(res->first, "lastModified"));
             return {
-                Tree{
-                    .actualPath = store->toRealPath(res->second),
-                    .storePath = std::move(res->second),
-                    .info = TreeInfo {
-                        .lastModified = getIntAttr(res->first, "lastModified"),
-                    },
-                },
+                Tree(store->toRealPath(res->second), std::move(res->second)),
                 input
             };
         }
 
-        // FIXME: use regular /archive URLs instead? api.github.com
-        // might have stricter rate limits.
+        auto url = getDownloadUrl(input);
 
-        auto url = fmt("https://api.github.com/repos/%s/%s/tarball/%s",
-            owner, repo, rev->to_string(Base16, false));
+        auto [tree, lastModified] = downloadTarball(store, url, "source", true);
 
-        std::string accessToken = settings.githubAccessToken.get();
-        if (accessToken != "")
-            url += "?access_token=" + accessToken;
-
-        auto tree = downloadTarball(store, url, "source", true);
+        input.attrs.insert_or_assign("lastModified", lastModified);
 
         getCache()->add(
             store,
             immutableAttrs,
             {
                 {"rev", rev->gitRev()},
-                {"lastModified", *tree.info.lastModified}
+                {"lastModified", lastModified}
             },
             tree.storePath,
             true);
@@ -128,68 +178,96 @@ struct GitHubInput : Input
     }
 };
 
-struct GitHubInputScheme : InputScheme
+struct GitHubInputScheme : GitArchiveInputScheme
 {
-    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    std::string type() override { return "github"; }
+
+    Hash getRevFromRef(nix::ref<Store> store, const Input & input) const override
     {
-        if (url.scheme != "github") return nullptr;
-
-        auto path = tokenizeString<std::vector<std::string>>(url.path, "/");
-        auto input = std::make_unique<GitHubInput>();
-
-        if (path.size() == 2) {
-        } else if (path.size() == 3) {
-            if (std::regex_match(path[2], revRegex))
-                input->rev = Hash(path[2], htSHA1);
-            else if (std::regex_match(path[2], refRegex))
-                input->ref = path[2];
-            else
-                throw BadURL("in GitHub URL '%s', '%s' is not a commit hash or branch/tag name", url.url, path[2]);
-        } else
-            throw BadURL("GitHub URL '%s' is invalid", url.url);
-
-        for (auto &[name, value] : url.query) {
-            if (name == "rev") {
-                if (input->rev)
-                    throw BadURL("GitHub URL '%s' contains multiple commit hashes", url.url);
-                input->rev = Hash(value, htSHA1);
-            }
-            else if (name == "ref") {
-                if (!std::regex_match(value, refRegex))
-                    throw BadURL("GitHub URL '%s' contains an invalid branch/tag name", url.url);
-                if (input->ref)
-                    throw BadURL("GitHub URL '%s' contains multiple branch/tag names", url.url);
-                input->ref = value;
-            }
-        }
-
-        if (input->ref && input->rev)
-            throw BadURL("GitHub URL '%s' contains both a commit hash and a branch/tag name", url.url);
-
-        input->owner = path[0];
-        input->repo = path[1];
-
-        return input;
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("github.com");
+        auto url = fmt("https://api.%s/repos/%s/%s/commits/%s", // FIXME: check
+            host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo"), *input.getRef());
+        auto json = nlohmann::json::parse(
+            readFile(
+                store->toRealPath(
+                    downloadFile(store, url, "source", false).storePath)));
+        auto rev = Hash(std::string { json["sha"] }, htSHA1);
+        debug("HEAD revision for '%s' is %s", url, rev.gitRev());
+        return rev;
     }
 
-    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    std::string getDownloadUrl(const Input & input) const override
     {
-        if (maybeGetStrAttr(attrs, "type") != "github") return {};
+        // FIXME: use regular /archive URLs instead? api.github.com
+        // might have stricter rate limits.
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("github.com");
+        auto url = fmt("https://api.%s/repos/%s/%s/tarball/%s", // FIXME: check if this is correct for self hosted instances
+            host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo"),
+            input.getRev()->to_string(Base16, false));
 
-        for (auto & [name, value] : attrs)
-            if (name != "type" && name != "owner" && name != "repo" && name != "ref" && name != "rev")
-                throw Error("unsupported GitHub input attribute '%s'", name);
+        std::string accessToken = settings.githubAccessToken.get();
+        if (accessToken != "")
+            url += "?access_token=" + accessToken;
 
-        auto input = std::make_unique<GitHubInput>();
-        input->owner = getStrAttr(attrs, "owner");
-        input->repo = getStrAttr(attrs, "repo");
-        input->ref = maybeGetStrAttr(attrs, "ref");
-        if (auto rev = maybeGetStrAttr(attrs, "rev"))
-            input->rev = Hash(*rev, htSHA1);
-        return input;
+        return url;
+    }
+
+    void clone(const Input & input, const Path & destDir) override
+    {
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("github.com");
+        Input::fromURL(fmt("git+ssh://git@%s/%s/%s.git",
+                host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo")))
+            .applyOverrides(input.getRef().value_or("HEAD"), input.getRev())
+            .clone(destDir);
+    }
+};
+
+struct GitLabInputScheme : GitArchiveInputScheme
+{
+    std::string type() override { return "gitlab"; }
+
+    Hash getRevFromRef(nix::ref<Store> store, const Input & input) const override
+    {
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("gitlab.com");
+        auto url = fmt("https://%s/api/v4/projects/%s%%2F%s/repository/commits?ref_name=%s",
+            host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo"), *input.getRef());
+        auto json = nlohmann::json::parse(
+            readFile(
+                store->toRealPath(
+                    downloadFile(store, url, "source", false).storePath)));
+        auto rev = Hash(std::string(json[0]["id"]), htSHA1);
+        debug("HEAD revision for '%s' is %s", url, rev.gitRev());
+        return rev;
+    }
+
+    std::string getDownloadUrl(const Input & input) const override
+    {
+        // FIXME: This endpoint has a rate limit threshold of 5 requests per minute
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("gitlab.com");
+        auto url = fmt("https://%s/api/v4/projects/%s%%2F%s/repository/archive.tar.gz?sha=%s",
+            host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo"),
+            input.getRev()->to_string(Base16, false));
+
+        /* # FIXME: add privat token auth (`curl --header "PRIVATE-TOKEN: <your_access_token>"`)
+        std::string accessToken = settings.githubAccessToken.get();
+        if (accessToken != "")
+            url += "?access_token=" + accessToken;*/
+
+        return url;
+    }
+
+    void clone(const Input & input, const Path & destDir) override
+    {
+        auto host_url = maybeGetStrAttr(input.attrs, "url").value_or("gitlab.com");
+        // FIXME: get username somewhere
+        Input::fromURL(fmt("git+ssh://git@%s/%s/%s.git",
+                host_url, getStrAttr(input.attrs, "owner"), getStrAttr(input.attrs, "repo")))
+            .applyOverrides(input.getRef().value_or("HEAD"), input.getRev())
+            .clone(destDir);
     }
 };
 
 static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<GitHubInputScheme>()); });
+static auto r2 = OnStartup([] { registerInputScheme(std::make_unique<GitLabInputScheme>()); });
 
 }

src/libfetchers/indirect.cc

@@ -0,0 +1,104 @@
+#include "fetchers.hh"
+
+namespace nix::fetchers {
+
+std::regex flakeRegex("[a-zA-Z][a-zA-Z0-9_-]*", std::regex::ECMAScript);
+
+struct IndirectInputScheme : InputScheme
+{
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
+    {
+        if (url.scheme != "flake") return {};
+
+        auto path = tokenizeString<std::vector<std::string>>(url.path, "/");
+
+        std::optional<Hash> rev;
+        std::optional<std::string> ref;
+
+        if (path.size() == 1) {
+        } else if (path.size() == 2) {
+            if (std::regex_match(path[1], revRegex))
+                rev = Hash(path[1], htSHA1);
+            else if (std::regex_match(path[1], refRegex))
+                ref = path[1];
+            else
+                throw BadURL("in flake URL '%s', '%s' is not a commit hash or branch/tag name", url.url, path[1]);
+        } else if (path.size() == 3) {
+            if (!std::regex_match(path[1], refRegex))
+                throw BadURL("in flake URL '%s', '%s' is not a branch/tag name", url.url, path[1]);
+            ref = path[1];
+            if (!std::regex_match(path[2], revRegex))
+                throw BadURL("in flake URL '%s', '%s' is not a commit hash", url.url, path[2]);
+            rev = Hash(path[2], htSHA1);
+        } else
+            throw BadURL("GitHub URL '%s' is invalid", url.url);
+
+        std::string id = path[0];
+        if (!std::regex_match(id, flakeRegex))
+            throw BadURL("'%s' is not a valid flake ID", id);
+
+        // FIXME: forbid query params?
+
+        Input input;
+        input.direct = false;
+        input.attrs.insert_or_assign("type", "indirect");
+        input.attrs.insert_or_assign("id", id);
+        if (rev) input.attrs.insert_or_assign("rev", rev->gitRev());
+        if (ref) input.attrs.insert_or_assign("ref", *ref);
+
+        return input;
+    }
+
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "indirect") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "id" && name != "ref" && name != "rev" && name != "narHash")
+                throw Error("unsupported indirect input attribute '%s'", name);
+
+        auto id = getStrAttr(attrs, "id");
+        if (!std::regex_match(id, flakeRegex))
+            throw BadURL("'%s' is not a valid flake ID", id);
+
+        Input input;
+        input.direct = false;
+        input.attrs = attrs;
+        return input;
+    }
+
+    ParsedURL toURL(const Input & input) override
+    {
+        ParsedURL url;
+        url.scheme = "flake";
+        url.path = getStrAttr(input.attrs, "id");
+        if (auto ref = input.getRef()) { url.path += '/'; url.path += *ref; };
+        if (auto rev = input.getRev()) { url.path += '/'; url.path += rev->gitRev(); };
+        return url;
+    }
+
+    bool hasAllInfo(const Input & input) override
+    {
+        return false;
+    }
+
+    Input applyOverrides(
+        const Input & _input,
+        std::optional<std::string> ref,
+        std::optional<Hash> rev) override
+    {
+        auto input(_input);
+        if (rev) input.attrs.insert_or_assign("rev", rev->gitRev());
+        if (ref) input.attrs.insert_or_assign("ref", *ref);
+        return input;
+    }
+
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & input) override
+    {
+        throw Error("indirect input '%s' cannot be fetched directly", input.to_string());
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<IndirectInputScheme>()); });
+
+}

src/libfetchers/mercurial.cc

@@ -10,76 +10,124 @@ using namespace std::string_literals;
 
 namespace nix::fetchers {
 
-struct MercurialInput : Input
+struct MercurialInputScheme : InputScheme
 {
-    ParsedURL url;
-    std::optional<std::string> ref;
-    std::optional<Hash> rev;
-
-    MercurialInput(const ParsedURL & url) : url(url)
-    { }
-
-    std::string type() const override { return "hg"; }
-
-    bool operator ==(const Input & other) const override
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
     {
-        auto other2 = dynamic_cast<const MercurialInput *>(&other);
-        return
-            other2
-            && url == other2->url
-            && rev == other2->rev
-            && ref == other2->ref;
+        if (url.scheme != "hg+http" &&
+            url.scheme != "hg+https" &&
+            url.scheme != "hg+ssh" &&
+            url.scheme != "hg+file") return {};
+
+        auto url2(url);
+        url2.scheme = std::string(url2.scheme, 3);
+        url2.query.clear();
+
+        Attrs attrs;
+        attrs.emplace("type", "hg");
+
+        for (auto &[name, value] : url.query) {
+            if (name == "rev" || name == "ref")
+                attrs.emplace(name, value);
+            else
+                url2.query.emplace(name, value);
+        }
+
+        attrs.emplace("url", url2.to_string());
+
+        return inputFromAttrs(attrs);
     }
 
-    bool isImmutable() const override
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
     {
-        return (bool) rev || narHash;
+        if (maybeGetStrAttr(attrs, "type") != "hg") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "ref" && name != "rev" && name != "revCount" && name != "narHash")
+                throw Error("unsupported Mercurial input attribute '%s'", name);
+
+        parseURL(getStrAttr(attrs, "url"));
+
+        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
+            if (!std::regex_match(*ref, refRegex))
+                throw BadURL("invalid Mercurial branch/tag name '%s'", *ref);
+        }
+
+        Input input;
+        input.attrs = attrs;
+        return input;
     }
 
-    std::optional<std::string> getRef() const override { return ref; }
-
-    std::optional<Hash> getRev() const override { return rev; }
-
-    ParsedURL toURL() const override
+    ParsedURL toURL(const Input & input) override
     {
-        ParsedURL url2(url);
-        url2.scheme = "hg+" + url2.scheme;
-        if (rev) url2.query.insert_or_assign("rev", rev->gitRev());
-        if (ref) url2.query.insert_or_assign("ref", *ref);
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
+        url.scheme = "hg+" + url.scheme;
+        if (auto rev = input.getRev()) url.query.insert_or_assign("rev", rev->gitRev());
+        if (auto ref = input.getRef()) url.query.insert_or_assign("ref", *ref);
         return url;
     }
 
-    Attrs toAttrsInternal() const override
+    bool hasAllInfo(const Input & input) override
     {
-        Attrs attrs;
-        attrs.emplace("url", url.to_string());
-        if (ref)
-            attrs.emplace("ref", *ref);
-        if (rev)
-            attrs.emplace("rev", rev->gitRev());
-        return attrs;
+        // FIXME: ugly, need to distinguish between dirty and clean
+        // default trees.
+        return input.getRef() == "default" || maybeGetIntAttr(input.attrs, "revCount");
     }
 
-    std::pair<bool, std::string> getActualUrl() const
+    Input applyOverrides(
+        const Input & input,
+        std::optional<std::string> ref,
+        std::optional<Hash> rev) override
     {
+        auto res(input);
+        if (rev) res.attrs.insert_or_assign("rev", rev->gitRev());
+        if (ref) res.attrs.insert_or_assign("ref", *ref);
+        return res;
+    }
+
+    std::optional<Path> getSourcePath(const Input & input) override
+    {
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
+        if (url.scheme == "file" && !input.getRef() && !input.getRev())
+            return url.path;
+        return {};
+    }
+
+    void markChangedFile(const Input & input, std::string_view file, std::optional<std::string> commitMsg) override
+    {
+        auto sourcePath = getSourcePath(input);
+        assert(sourcePath);
+
+        // FIXME: shut up if file is already tracked.
+        runProgram("hg", true,
+            { "add", *sourcePath + "/" + std::string(file) });
+
+        if (commitMsg)
+            runProgram("hg", true,
+                { "commit", *sourcePath + "/" + std::string(file), "-m", *commitMsg });
+    }
+
+    std::pair<bool, std::string> getActualUrl(const Input & input) const
+    {
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
         bool isLocal = url.scheme == "file";
         return {isLocal, isLocal ? url.path : url.base};
     }
 
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & _input) override
     {
         auto name = "source";
 
-        auto input = std::make_shared<MercurialInput>(*this);
+        Input input(_input);
 
-        auto [isLocal, actualUrl_] = getActualUrl();
+        auto [isLocal, actualUrl_] = getActualUrl(input);
         auto actualUrl = actualUrl_; // work around clang bug
 
         // FIXME: return lastModified.
 
         // FIXME: don't clone local repositories.
 
-        if (!input->ref && !input->rev && isLocal && pathExists(actualUrl + "/.hg")) {
+        if (!input.getRef() && !input.getRev() && isLocal && pathExists(actualUrl + "/.hg")) {
 
             bool clean = runProgram("hg", true, { "status", "-R", actualUrl, "--modified", "--added", "--removed" }) == "";
 
@@ -94,7 +142,7 @@ struct MercurialInput : Input
                 if (settings.warnDirty)
                     warn("Mercurial tree '%s' is unclean", actualUrl);
 
-                input->ref = chomp(runProgram("hg", true, { "branch", "-R", actualUrl }));
+                input.attrs.insert_or_assign("ref", chomp(runProgram("hg", true, { "branch", "-R", actualUrl })));
 
                 auto files = tokenizeString<std::set<std::string>>(
                     runProgram("hg", true, { "status", "-R", actualUrl, "--clean", "--modified", "--added", "--no-status", "--print0" }), "\0"s);
@@ -116,60 +164,54 @@ struct MercurialInput : Input
 
                 auto storePath = store->addToStore("source", actualUrl, FileIngestionMethod::Recursive, htSHA256, filter);
 
-                return {Tree {
-                    .actualPath = store->printStorePath(storePath),
-                    .storePath = std::move(storePath),
-                }, input};
+                return {
+                    Tree(store->printStorePath(storePath), std::move(storePath)),
+                    input
+                };
             }
         }
 
-        if (!input->ref) input->ref = "default";
+        if (!input.getRef()) input.attrs.insert_or_assign("ref", "default");
 
         auto getImmutableAttrs = [&]()
         {
             return Attrs({
                 {"type", "hg"},
                 {"name", name},
-                {"rev", input->rev->gitRev()},
+                {"rev", input.getRev()->gitRev()},
             });
         };
 
         auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
-            -> std::pair<Tree, std::shared_ptr<const Input>>
+            -> std::pair<Tree, Input>
         {
-            assert(input->rev);
-            assert(!rev || rev == input->rev);
+            assert(input.getRev());
+            assert(!_input.getRev() || _input.getRev() == input.getRev());
+            input.attrs.insert_or_assign("revCount", getIntAttr(infoAttrs, "revCount"));
             return {
-                Tree{
-                    .actualPath = store->toRealPath(storePath),
-                    .storePath = std::move(storePath),
-                    .info = TreeInfo {
-                        .revCount = getIntAttr(infoAttrs, "revCount"),
-                    },
-                },
+                Tree(store->toRealPath(storePath), std::move(storePath)),
                 input
             };
         };
 
-        if (input->rev) {
+        if (input.getRev()) {
             if (auto res = getCache()->lookup(store, getImmutableAttrs()))
                 return makeResult(res->first, std::move(res->second));
         }
 
-        assert(input->rev || input->ref);
-        auto revOrRef = input->rev ? input->rev->gitRev() : *input->ref;
+        auto revOrRef = input.getRev() ? input.getRev()->gitRev() : *input.getRef();
 
         Attrs mutableAttrs({
             {"type", "hg"},
             {"name", name},
             {"url", actualUrl},
-            {"ref", *input->ref},
+            {"ref", *input.getRef()},
         });
 
         if (auto res = getCache()->lookup(store, mutableAttrs)) {
             auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
-            if (!rev || rev == rev2) {
-                input->rev = rev2;
+            if (!input.getRev() || input.getRev() == rev2) {
+                input.attrs.insert_or_assign("rev", rev2.gitRev());
                 return makeResult(res->first, std::move(res->second));
             }
         }
@@ -178,10 +220,10 @@ struct MercurialInput : Input
 
         /* If this is a commit hash that we already have, we don't
            have to pull again. */
-        if (!(input->rev
+        if (!(input.getRev()
                 && pathExists(cacheDir)
                 && runProgram(
-                    RunOptions("hg", { "log", "-R", cacheDir, "-r", input->rev->gitRev(), "--template", "1" })
+                    RunOptions("hg", { "log", "-R", cacheDir, "-r", input.getRev()->gitRev(), "--template", "1" })
                     .killStderr(true)).second == "1"))
         {
             Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Mercurial repository '%s'", actualUrl));
@@ -210,9 +252,9 @@ struct MercurialInput : Input
             runProgram("hg", true, { "log", "-R", cacheDir, "-r", revOrRef, "--template", "{node} {rev} {branch}" }));
         assert(tokens.size() == 3);
 
-        input->rev = Hash(tokens[0], htSHA1);
+        input.attrs.insert_or_assign("rev", Hash(tokens[0], htSHA1).gitRev());
         auto revCount = std::stoull(tokens[1]);
-        input->ref = tokens[2];
+        input.attrs.insert_or_assign("ref", tokens[2]);
 
         if (auto res = getCache()->lookup(store, getImmutableAttrs()))
             return makeResult(res->first, std::move(res->second));
@@ -220,18 +262,18 @@ struct MercurialInput : Input
         Path tmpDir = createTempDir();
         AutoDelete delTmpDir(tmpDir, true);
 
-        runProgram("hg", true, { "archive", "-R", cacheDir, "-r", input->rev->gitRev(), tmpDir });
+        runProgram("hg", true, { "archive", "-R", cacheDir, "-r", input.getRev()->gitRev(), tmpDir });
 
         deletePath(tmpDir + "/.hg_archival.txt");
 
         auto storePath = store->addToStore(name, tmpDir);
 
         Attrs infoAttrs({
-            {"rev", input->rev->gitRev()},
+            {"rev", input.getRev()->gitRev()},
             {"revCount", (int64_t) revCount},
         });
 
-        if (!this->rev)
+        if (!_input.getRev())
             getCache()->add(
                 store,
                 mutableAttrs,
@@ -250,54 +292,6 @@ struct MercurialInput : Input
     }
 };
 
-struct MercurialInputScheme : InputScheme
-{
-    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
-    {
-        if (url.scheme != "hg+http" &&
-            url.scheme != "hg+https" &&
-            url.scheme != "hg+ssh" &&
-            url.scheme != "hg+file") return nullptr;
-
-        auto url2(url);
-        url2.scheme = std::string(url2.scheme, 3);
-        url2.query.clear();
-
-        Attrs attrs;
-        attrs.emplace("type", "hg");
-
-        for (auto &[name, value] : url.query) {
-            if (name == "rev" || name == "ref")
-                attrs.emplace(name, value);
-            else
-                url2.query.emplace(name, value);
-        }
-
-        attrs.emplace("url", url2.to_string());
-
-        return inputFromAttrs(attrs);
-    }
-
-    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
-    {
-        if (maybeGetStrAttr(attrs, "type") != "hg") return {};
-
-        for (auto & [name, value] : attrs)
-            if (name != "type" && name != "url" && name != "ref" && name != "rev")
-                throw Error("unsupported Mercurial input attribute '%s'", name);
-
-        auto input = std::make_unique<MercurialInput>(parseURL(getStrAttr(attrs, "url")));
-        if (auto ref = maybeGetStrAttr(attrs, "ref")) {
-            if (!std::regex_match(*ref, refRegex))
-                throw BadURL("invalid Mercurial branch/tag name '%s'", *ref);
-            input->ref = *ref;
-        }
-        if (auto rev = maybeGetStrAttr(attrs, "rev"))
-            input->rev = Hash(*rev, htSHA1);
-        return input;
-    }
-};
-
 static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<MercurialInputScheme>()); });
 
 }

src/libfetchers/path.cc

@@ -3,65 +3,86 @@
 
 namespace nix::fetchers {
 
-struct PathInput : Input
+struct PathInputScheme : InputScheme
 {
-    Path path;
-
-    /* Allow the user to pass in "fake" tree info attributes. This is
-       useful for making a pinned tree work the same as the repository
-       from which is exported
-       (e.g. path:/nix/store/...-source?lastModified=1585388205&rev=b0c285...). */
-    std::optional<Hash> rev;
-    std::optional<uint64_t> revCount;
-    std::optional<time_t> lastModified;
-
-    std::string type() const override { return "path"; }
-
-    std::optional<Hash> getRev() const override { return rev; }
-
-    bool operator ==(const Input & other) const override
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
     {
-        auto other2 = dynamic_cast<const PathInput *>(&other);
-        return
-            other2
-            && path == other2->path
-            && rev == other2->rev
-            && revCount == other2->revCount
-            && lastModified == other2->lastModified;
+        if (url.scheme != "path") return {};
+
+        if (url.authority && *url.authority != "")
+            throw Error("path URL '%s' should not have an authority ('%s')", url.url, *url.authority);
+
+        Input input;
+        input.attrs.insert_or_assign("type", "path");
+        input.attrs.insert_or_assign("path", url.path);
+
+        for (auto & [name, value] : url.query)
+            if (name == "rev" || name == "narHash")
+                input.attrs.insert_or_assign(name, value);
+            else if (name == "revCount" || name == "lastModified") {
+                uint64_t n;
+                if (!string2Int(value, n))
+                    throw Error("path URL '%s' has invalid parameter '%s'", url.to_string(), name);
+                input.attrs.insert_or_assign(name, n);
+            }
+            else
+                throw Error("path URL '%s' has unsupported parameter '%s'", url.to_string(), name);
+
+        return input;
     }
 
-    bool isImmutable() const override
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
     {
-        return (bool) narHash;
+        if (maybeGetStrAttr(attrs, "type") != "path") return {};
+
+        getStrAttr(attrs, "path");
+
+        for (auto & [name, value] : attrs)
+            /* Allow the user to pass in "fake" tree info
+               attributes. This is useful for making a pinned tree
+               work the same as the repository from which is exported
+               (e.g. path:/nix/store/...-source?lastModified=1585388205&rev=b0c285...). */
+            if (name == "type" || name == "rev" || name == "revCount" || name == "lastModified" || name == "narHash" || name == "path")
+                // checked in Input::fromAttrs
+                ;
+            else
+                throw Error("unsupported path input attribute '%s'", name);
+
+        Input input;
+        input.attrs = attrs;
+        return input;
     }
 
-    ParsedURL toURL() const override
+    ParsedURL toURL(const Input & input) override
     {
-        auto query = attrsToQuery(toAttrsInternal());
+        auto query = attrsToQuery(input.attrs);
         query.erase("path");
+        query.erase("type");
         return ParsedURL {
             .scheme = "path",
-            .path = path,
+            .path = getStrAttr(input.attrs, "path"),
             .query = query,
         };
     }
 
-    Attrs toAttrsInternal() const override
+    bool hasAllInfo(const Input & input) override
     {
-        Attrs attrs;
-        attrs.emplace("path", path);
-        if (rev)
-            attrs.emplace("rev", rev->gitRev());
-        if (revCount)
-            attrs.emplace("revCount", *revCount);
-        if (lastModified)
-            attrs.emplace("lastModified", *lastModified);
-        return attrs;
+        return true;
     }
 
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    std::optional<Path> getSourcePath(const Input & input) override
     {
-        auto input = std::make_shared<PathInput>(*this);
+        return getStrAttr(input.attrs, "path");
+    }
+
+    void markChangedFile(const Input & input, std::string_view file, std::optional<std::string> commitMsg) override
+    {
+        // nothing to do
+    }
+
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & input) override
+    {
+        auto path = getStrAttr(input.attrs, "path");
 
         // FIXME: check whether access to 'path' is allowed.
 
@@ -74,72 +95,10 @@ struct PathInput : Input
             // FIXME: try to substitute storePath.
             storePath = store->addToStore("source", path);
 
-        return
-            {
-                Tree {
-                    .actualPath = store->toRealPath(*storePath),
-                    .storePath = std::move(*storePath),
-                    .info = TreeInfo {
-                        .revCount = revCount,
-                        .lastModified = lastModified
-                    }
-                },
-                input
-            };
-    }
-
-};
-
-struct PathInputScheme : InputScheme
-{
-    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
-    {
-        if (url.scheme != "path") return nullptr;
-
-        auto input = std::make_unique<PathInput>();
-        input->path = url.path;
-
-        for (auto & [name, value] : url.query)
-            if (name == "rev")
-                input->rev = Hash(value, htSHA1);
-            else if (name == "revCount") {
-                uint64_t revCount;
-                if (!string2Int(value, revCount))
-                    throw Error("path URL '%s' has invalid parameter '%s'", url.to_string(), name);
-                input->revCount = revCount;
-            }
-            else if (name == "lastModified") {
-                time_t lastModified;
-                if (!string2Int(value, lastModified))
-                    throw Error("path URL '%s' has invalid parameter '%s'", url.to_string(), name);
-                input->lastModified = lastModified;
-            }
-            else
-                throw Error("path URL '%s' has unsupported parameter '%s'", url.to_string(), name);
-
-        return input;
-    }
-
-    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
-    {
-        if (maybeGetStrAttr(attrs, "type") != "path") return {};
-
-        auto input = std::make_unique<PathInput>();
-        input->path = getStrAttr(attrs, "path");
-
-        for (auto & [name, value] : attrs)
-            if (name == "rev")
-                input->rev = Hash(getStrAttr(attrs, "rev"), htSHA1);
-            else if (name == "revCount")
-                input->revCount = getIntAttr(attrs, "revCount");
-            else if (name == "lastModified")
-                input->lastModified = getIntAttr(attrs, "lastModified");
-            else if (name == "type" || name == "path")
-                ;
-            else
-                throw Error("unsupported path input attribute '%s'", name);
-
-        return input;
+        return {
+            Tree(store->toRealPath(*storePath), std::move(*storePath)),
+            input
+        };
     }
 };
 

src/libfetchers/registry.cc

@@ -0,0 +1,212 @@
+#include "registry.hh"
+#include "fetchers.hh"
+#include "util.hh"
+#include "globals.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+std::shared_ptr<Registry> Registry::read(
+    const Path & path, RegistryType type)
+{
+    auto registry = std::make_shared<Registry>(type);
+
+    if (!pathExists(path))
+        return std::make_shared<Registry>(type);
+
+    try {
+
+        auto json = nlohmann::json::parse(readFile(path));
+
+        auto version = json.value("version", 0);
+
+        if (version == 2) {
+            for (auto & i : json["flakes"]) {
+                auto toAttrs = jsonToAttrs(i["to"]);
+                Attrs extraAttrs;
+                auto j = toAttrs.find("dir");
+                if (j != toAttrs.end()) {
+                    extraAttrs.insert(*j);
+                    toAttrs.erase(j);
+                }
+                auto exact = i.find("exact");
+                registry->entries.push_back(
+                    Entry {
+                        .from = Input::fromAttrs(jsonToAttrs(i["from"])),
+                        .to = Input::fromAttrs(std::move(toAttrs)),
+                        .extraAttrs = extraAttrs,
+                        .exact = exact != i.end() && exact.value()
+                    });
+            }
+        }
+
+        else
+            throw Error("flake registry '%s' has unsupported version %d", path, version);
+
+    } catch (nlohmann::json::exception & e) {
+        warn("cannot parse flake registry '%s': %s", path, e.what());
+    } catch (Error & e) {
+        warn("cannot read flake registry '%s': %s", path, e.what());
+    }
+
+    return registry;
+}
+
+void Registry::write(const Path & path)
+{
+    nlohmann::json arr;
+    for (auto & entry : entries) {
+        nlohmann::json obj;
+        obj["from"] = attrsToJson(entry.from.toAttrs());
+        obj["to"] = attrsToJson(entry.to.toAttrs());
+        if (!entry.extraAttrs.empty())
+            obj["to"].update(attrsToJson(entry.extraAttrs));
+        if (entry.exact)
+            obj["exact"] = true;
+        arr.emplace_back(std::move(obj));
+    }
+
+    nlohmann::json json;
+    json["version"] = 2;
+    json["flakes"] = std::move(arr);
+
+    createDirs(dirOf(path));
+    writeFile(path, json.dump(2));
+}
+
+void Registry::add(
+    const Input & from,
+    const Input & to,
+    const Attrs & extraAttrs)
+{
+    entries.emplace_back(
+        Entry {
+            .from = from,
+            .to = to,
+            .extraAttrs = extraAttrs
+        });
+}
+
+void Registry::remove(const Input & input)
+{
+    // FIXME: use C++20 std::erase.
+    for (auto i = entries.begin(); i != entries.end(); )
+        if (i->from == input)
+            i = entries.erase(i);
+        else
+            ++i;
+}
+
+static Path getSystemRegistryPath()
+{
+    return settings.nixConfDir + "/registry.json";
+}
+
+static std::shared_ptr<Registry> getSystemRegistry()
+{
+    static auto systemRegistry =
+        Registry::read(getSystemRegistryPath(), Registry::System);
+    return systemRegistry;
+}
+
+Path getUserRegistryPath()
+{
+    return getHome() + "/.config/nix/registry.json";
+}
+
+std::shared_ptr<Registry> getUserRegistry()
+{
+    static auto userRegistry =
+        Registry::read(getUserRegistryPath(), Registry::User);
+    return userRegistry;
+}
+
+static std::shared_ptr<Registry> flagRegistry =
+    std::make_shared<Registry>(Registry::Flag);
+
+std::shared_ptr<Registry> getFlagRegistry()
+{
+    return flagRegistry;
+}
+
+void overrideRegistry(
+    const Input & from,
+    const Input & to,
+    const Attrs & extraAttrs)
+{
+    flagRegistry->add(from, to, extraAttrs);
+}
+
+static std::shared_ptr<Registry> getGlobalRegistry(ref<Store> store)
+{
+    static auto reg = [&]() {
+        auto path = settings.flakeRegistry.get();
+
+        if (!hasPrefix(path, "/")) {
+            auto storePath = downloadFile(store, path, "flake-registry.json", false).storePath;
+            if (auto store2 = store.dynamic_pointer_cast<LocalFSStore>())
+                store2->addPermRoot(storePath, getCacheDir() + "/nix/flake-registry.json", true);
+            path = store->toRealPath(storePath);
+        }
+
+        return Registry::read(path, Registry::Global);
+    }();
+
+    return reg;
+}
+
+Registries getRegistries(ref<Store> store)
+{
+    Registries registries;
+    registries.push_back(getFlagRegistry());
+    registries.push_back(getUserRegistry());
+    registries.push_back(getSystemRegistry());
+    registries.push_back(getGlobalRegistry(store));
+    return registries;
+}
+
+std::pair<Input, Attrs> lookupInRegistries(
+    ref<Store> store,
+    const Input & _input)
+{
+    Attrs extraAttrs;
+    int n = 0;
+    Input input(_input);
+
+ restart:
+
+    n++;
+    if (n > 100) throw Error("cycle detected in flake registry for '%s'", input.to_string());
+
+    for (auto & registry : getRegistries(store)) {
+        // FIXME: O(n)
+        for (auto & entry : registry->entries) {
+            if (entry.exact) {
+                if (entry.from == input) {
+                    input = entry.to;
+                    extraAttrs = entry.extraAttrs;
+                    goto restart;
+                }
+            } else {
+                if (entry.from.contains(input)) {
+                    input = entry.to.applyOverrides(
+                        !entry.from.getRef() && input.getRef() ? input.getRef() : std::optional<std::string>(),
+                        !entry.from.getRev() && input.getRev() ? input.getRev() : std::optional<Hash>());
+                    extraAttrs = entry.extraAttrs;
+                    goto restart;
+                }
+            }
+        }
+    }
+
+    if (!input.isDirect())
+        throw Error("cannot find flake '%s' in the flake registries", input.to_string());
+
+    debug("looked up '%s' -> '%s'", _input.to_string(), input.to_string());
+
+    return {input, extraAttrs};
+}
+
+}

src/libfetchers/registry.hh

@@ -0,0 +1,64 @@
+#pragma once
+
+#include "types.hh"
+#include "fetchers.hh"
+
+namespace nix { class Store; }
+
+namespace nix::fetchers {
+
+struct Registry
+{
+    enum RegistryType {
+        Flag = 0,
+        User = 1,
+        System = 2,
+        Global = 3,
+    };
+
+    RegistryType type;
+
+    struct Entry
+    {
+        Input from, to;
+        Attrs extraAttrs;
+        bool exact = false;
+    };
+
+    std::vector<Entry> entries;
+
+    Registry(RegistryType type)
+        : type(type)
+    { }
+
+    static std::shared_ptr<Registry> read(
+        const Path & path, RegistryType type);
+
+    void write(const Path & path);
+
+    void add(
+        const Input & from,
+        const Input & to,
+        const Attrs & extraAttrs);
+
+    void remove(const Input & input);
+};
+
+typedef std::vector<std::shared_ptr<Registry>> Registries;
+
+std::shared_ptr<Registry> getUserRegistry();
+
+Path getUserRegistryPath();
+
+Registries getRegistries(ref<Store> store);
+
+void overrideRegistry(
+    const Input & from,
+    const Input & to,
+    const Attrs & extraAttrs);
+
+std::pair<Input, Attrs> lookupInRegistries(
+    ref<Store> store,
+    const Input & input);
+
+}

src/libfetchers/tarball.cc

@@ -105,7 +105,7 @@ DownloadFileResult downloadFile(
     };
 }
 
-Tree downloadTarball(
+std::pair<Tree, time_t> downloadTarball(
     ref<Store> store,
     const std::string & url,
     const std::string & name,
@@ -120,12 +120,9 @@ Tree downloadTarball(
     auto cached = getCache()->lookupExpired(store, inAttrs);
 
     if (cached && !cached->expired)
-        return Tree {
-            .actualPath = store->toRealPath(cached->storePath),
-            .storePath = std::move(cached->storePath),
-            .info = TreeInfo {
-                .lastModified = getIntAttr(cached->infoAttrs, "lastModified"),
-            },
+        return {
+            Tree(store->toRealPath(cached->storePath), std::move(cached->storePath)),
+            getIntAttr(cached->infoAttrs, "lastModified")
         };
 
     auto res = downloadFile(store, url, name, immutable);
@@ -160,117 +157,72 @@ Tree downloadTarball(
         *unpackedStorePath,
         immutable);
 
-    return Tree {
-        .actualPath = store->toRealPath(*unpackedStorePath),
-        .storePath = std::move(*unpackedStorePath),
-        .info = TreeInfo {
-            .lastModified = lastModified,
-        },
+    return {
+        Tree(store->toRealPath(*unpackedStorePath), std::move(*unpackedStorePath)),
+        lastModified,
     };
 }
 
-struct TarballInput : Input
-{
-    ParsedURL url;
-    std::optional<Hash> hash;
-
-    TarballInput(const ParsedURL & url) : url(url)
-    { }
-
-    std::string type() const override { return "tarball"; }
-
-    bool operator ==(const Input & other) const override
-    {
-        auto other2 = dynamic_cast<const TarballInput *>(&other);
-        return
-            other2
-            && to_string() == other2->to_string()
-            && hash == other2->hash;
-    }
-
-    bool isImmutable() const override
-    {
-        return hash || narHash;
-    }
-
-    ParsedURL toURL() const override
-    {
-        auto url2(url);
-        // NAR hashes are preferred over file hashes since tar/zip files
-        // don't have a canonical representation.
-        if (narHash)
-            url2.query.insert_or_assign("narHash", narHash->to_string(SRI, true));
-        else if (hash)
-            url2.query.insert_or_assign("hash", hash->to_string(SRI, true));
-        return url2;
-    }
-
-    Attrs toAttrsInternal() const override
-    {
-        Attrs attrs;
-        attrs.emplace("url", url.to_string());
-        if (hash)
-            attrs.emplace("hash", hash->to_string(SRI, true));
-        return attrs;
-    }
-
-    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
-    {
-        auto tree = downloadTarball(store, url.to_string(), "source", false);
-
-        auto input = std::make_shared<TarballInput>(*this);
-        input->narHash = store->queryPathInfo(tree.storePath)->narHash;
-
-        return {std::move(tree), input};
-    }
-};
-
 struct TarballInputScheme : InputScheme
 {
-    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) override
+    std::optional<Input> inputFromURL(const ParsedURL & url) override
     {
-        if (url.scheme != "file" && url.scheme != "http" && url.scheme != "https") return nullptr;
+        if (url.scheme != "file" && url.scheme != "http" && url.scheme != "https") return {};
 
         if (!hasSuffix(url.path, ".zip")
             && !hasSuffix(url.path, ".tar")
             && !hasSuffix(url.path, ".tar.gz")
             && !hasSuffix(url.path, ".tar.xz")
             && !hasSuffix(url.path, ".tar.bz2"))
-            return nullptr;
-
-        auto input = std::make_unique<TarballInput>(url);
-
-        auto hash = input->url.query.find("hash");
-        if (hash != input->url.query.end()) {
-            // FIXME: require SRI hash.
-            input->hash = Hash(hash->second);
-            input->url.query.erase(hash);
-        }
-
-        auto narHash = input->url.query.find("narHash");
-        if (narHash != input->url.query.end()) {
-            // FIXME: require SRI hash.
-            input->narHash = Hash(narHash->second);
-            input->url.query.erase(narHash);
-        }
+            return {};
 
+        Input input;
+        input.attrs.insert_or_assign("type", "tarball");
+        input.attrs.insert_or_assign("url", url.to_string());
+        auto narHash = url.query.find("narHash");
+        if (narHash != url.query.end())
+            input.attrs.insert_or_assign("narHash", narHash->second);
         return input;
     }
 
-    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    std::optional<Input> inputFromAttrs(const Attrs & attrs) override
     {
         if (maybeGetStrAttr(attrs, "type") != "tarball") return {};
 
         for (auto & [name, value] : attrs)
-            if (name != "type" && name != "url" && name != "hash")
+            if (name != "type" && name != "url" && /* name != "hash" && */ name != "narHash")
                 throw Error("unsupported tarball input attribute '%s'", name);
 
-        auto input = std::make_unique<TarballInput>(parseURL(getStrAttr(attrs, "url")));
-        if (auto hash = maybeGetStrAttr(attrs, "hash"))
-            input->hash = newHashAllowEmpty(*hash, {});
-
+        Input input;
+        input.attrs = attrs;
+        //input.immutable = (bool) maybeGetStrAttr(input.attrs, "hash");
         return input;
     }
+
+    ParsedURL toURL(const Input & input) override
+    {
+        auto url = parseURL(getStrAttr(input.attrs, "url"));
+        // NAR hashes are preferred over file hashes since tar/zip files
+        // don't have a canonical representation.
+        if (auto narHash = input.getNarHash())
+            url.query.insert_or_assign("narHash", narHash->to_string(SRI, true));
+        /*
+        else if (auto hash = maybeGetStrAttr(input.attrs, "hash"))
+            url.query.insert_or_assign("hash", Hash(*hash).to_string(SRI, true));
+        */
+        return url;
+    }
+
+    bool hasAllInfo(const Input & input) override
+    {
+        return true;
+    }
+
+    std::pair<Tree, Input> fetch(ref<Store> store, const Input & input) override
+    {
+        auto tree = downloadTarball(store, getStrAttr(input.attrs, "url"), "source", false).first;
+        return {std::move(tree), input};
+    }
 };
 
 static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<TarballInputScheme>()); });

src/libfetchers/tree-info.cc

@@ -1,14 +0,0 @@
-#include "tree-info.hh"
-#include "store-api.hh"
-
-#include <nlohmann/json.hpp>
-
-namespace nix::fetchers {
-
-StorePath TreeInfo::computeStorePath(Store & store) const
-{
-    assert(narHash);
-    return store.makeFixedOutputPath(FileIngestionMethod::Recursive, narHash, "source");
-}
-
-}

src/libfetchers/tree-info.hh

@@ -1,29 +0,0 @@
-#pragma once
-
-#include "path.hh"
-#include "hash.hh"
-
-#include <nlohmann/json_fwd.hpp>
-
-namespace nix { class Store; }
-
-namespace nix::fetchers {
-
-struct TreeInfo
-{
-    Hash narHash;
-    std::optional<uint64_t> revCount;
-    std::optional<time_t> lastModified;
-
-    bool operator ==(const TreeInfo & other) const
-    {
-        return
-            narHash == other.narHash
-            && revCount == other.revCount
-            && lastModified == other.lastModified;
-    }
-
-    StorePath computeStorePath(Store & store) const;
-};
-
-}

src/libmain/common-args.cc

@@ -34,9 +34,19 @@ MixCommonArgs::MixCommonArgs(const string & programName)
             try {
                 globalConfig.set(name, value);
             } catch (UsageError & e) {
-                warn(e.what());
+                if (!completions)
+                    warn(e.what());
             }
         }},
+        .completer = [](size_t index, std::string_view prefix) {
+            if (index == 0) {
+                std::map<std::string, Config::SettingInfo> settings;
+                globalConfig.getSettings(settings);
+                for (auto & s : settings)
+                    if (hasPrefix(s.first, prefix))
+                        completions->insert(s.first);
+            }
+        }
     });
 
     addFlag({

src/libmain/loggers.cc

@@ -1,12 +1,13 @@
 #include "loggers.hh"
 #include "progress-bar.hh"
+#include "util.hh"
 
 namespace nix {
 
 LogFormat defaultLogFormat = LogFormat::raw;
 
 LogFormat parseLogFormat(const std::string & logFormatStr) {
-    if (logFormatStr == "raw")
+    if (logFormatStr == "raw" || getEnv("NIX_GET_COMPLETIONS"))
         return LogFormat::raw;
     else if (logFormatStr == "raw-with-logs")
         return LogFormat::rawWithLogs;

src/libmain/shared.cc

@@ -36,7 +36,7 @@ void printGCWarning()
 
 void printMissing(ref<Store> store, const std::vector<StorePathWithOutputs> & paths, Verbosity lvl)
 {
-    unsigned long long downloadSize, narSize;
+    uint64_t downloadSize, narSize;
     StorePathSet willBuild, willSubstitute, unknown;
     store->queryMissing(paths, willBuild, willSubstitute, unknown, downloadSize, narSize);
     printMissing(store, willBuild, willSubstitute, unknown, downloadSize, narSize, lvl);
@@ -45,7 +45,7 @@ void printMissing(ref<Store> store, const std::vector<StorePathWithOutputs> & pa
 
 void printMissing(ref<Store> store, const StorePathSet & willBuild,
     const StorePathSet & willSubstitute, const StorePathSet & unknown,
-    unsigned long long downloadSize, unsigned long long narSize, Verbosity lvl)
+    uint64_t downloadSize, uint64_t narSize, Verbosity lvl)
 {
     if (!willBuild.empty()) {
         if (willBuild.size() == 1)
@@ -384,7 +384,7 @@ RunPager::~RunPager()
 }
 
 
-string showBytes(unsigned long long bytes)
+string showBytes(uint64_t bytes)
 {
     return (format("%.2f MiB") % (bytes / (1024.0 * 1024.0))).str();
 }

src/libmain/shared.hh

@@ -47,7 +47,7 @@ void printMissing(
 
 void printMissing(ref<Store> store, const StorePathSet & willBuild,
     const StorePathSet & willSubstitute, const StorePathSet & unknown,
-    unsigned long long downloadSize, unsigned long long narSize, Verbosity lvl = lvlInfo);
+    uint64_t downloadSize, uint64_t narSize, Verbosity lvl = lvlInfo);
 
 string getArg(const string & opt,
     Strings::iterator & i, const Strings::iterator & end);
@@ -110,7 +110,7 @@ extern volatile ::sig_atomic_t blockInt;
 
 /* GC helpers. */
 
-string showBytes(unsigned long long bytes);
+string showBytes(uint64_t bytes);
 
 struct GCResults;
 

src/libstore/binary-cache-store.cc

@@ -15,6 +15,7 @@
 #include <chrono>
 #include <future>
 #include <regex>
+#include <fstream>
 
 #include <nlohmann/json.hpp>
 
@@ -57,6 +58,13 @@ void BinaryCacheStore::init()
     }
 }
 
+void BinaryCacheStore::upsertFile(const std::string & path,
+    std::string && data,
+    const std::string & mimeType)
+{
+    upsertFile(path, std::make_shared<std::stringstream>(std::move(data)), mimeType);
+}
+
 void BinaryCacheStore::getFile(const std::string & path,
     Callback<std::shared_ptr<std::string>> callback) noexcept
 {
@@ -113,13 +121,74 @@ void BinaryCacheStore::writeNarInfo(ref<NarInfo> narInfo)
         diskCache->upsertNarInfo(getUri(), hashPart, std::shared_ptr<NarInfo>(narInfo));
 }
 
-void BinaryCacheStore::addToStore(const ValidPathInfo & info, Source & narSource,
-    RepairFlag repair, CheckSigsFlag checkSigs, std::shared_ptr<FSAccessor> accessor)
+AutoCloseFD openFile(const Path & path)
 {
-    // FIXME: See if we can use the original source to reduce memory usage.
-    auto nar = make_ref<std::string>(narSource.drain());
+    auto fd = open(path.c_str(), O_RDONLY | O_CLOEXEC);
+    if (!fd)
+        throw SysError("opening file '%1%'", path);
+    return fd;
+}
 
-    if (!repair && isValidPath(info.path)) return;
+struct FileSource : FdSource
+{
+    AutoCloseFD fd2;
+
+    FileSource(const Path & path)
+        : fd2(openFile(path))
+    {
+        fd = fd2.get();
+    }
+};
+
+void BinaryCacheStore::addToStore(const ValidPathInfo & info, Source & narSource,
+    RepairFlag repair, CheckSigsFlag checkSigs)
+{
+    assert(info.narHash && info.narSize);
+
+    if (!repair && isValidPath(info.path)) {
+        // FIXME: copyNAR -> null sink
+        narSource.drain();
+        return;
+    }
+
+    auto [fdTemp, fnTemp] = createTempFile();
+
+    auto now1 = std::chrono::steady_clock::now();
+
+    /* Read the NAR simultaneously into a CompressionSink+FileSink (to
+       write the compressed NAR to disk), into a HashSink (to get the
+       NAR hash), and into a NarAccessor (to get the NAR listing). */
+    HashSink fileHashSink(htSHA256);
+    std::shared_ptr<FSAccessor> narAccessor;
+    {
+    FdSink fileSink(fdTemp.get());
+    TeeSink teeSink(fileSink, fileHashSink);
+    auto compressionSink = makeCompressionSink(compression, teeSink);
+    TeeSource teeSource(narSource, *compressionSink);
+    narAccessor = makeNarAccessor(teeSource);
+    compressionSink->finish();
+    }
+
+    auto now2 = std::chrono::steady_clock::now();
+
+    auto narInfo = make_ref<NarInfo>(info);
+    narInfo->narSize = info.narSize;
+    narInfo->narHash = info.narHash;
+    narInfo->compression = compression;
+    auto [fileHash, fileSize] = fileHashSink.finish();
+    narInfo->fileHash = fileHash;
+    narInfo->fileSize = fileSize;
+    narInfo->url = "nar/" + narInfo->fileHash->to_string(Base32, false) + ".nar"
+        + (compression == "xz" ? ".xz" :
+           compression == "bzip2" ? ".bz2" :
+           compression == "br" ? ".br" :
+           "");
+
+    auto duration = std::chrono::duration_cast<std::chrono::milliseconds>(now2 - now1).count();
+    printMsg(lvlTalkative, "copying path '%1%' (%2% bytes, compressed %3$.1f%% in %4% ms) to binary cache",
+        printStorePath(narInfo->path), info.narSize,
+        ((1.0 - (double) fileSize / info.narSize) * 100.0),
+        duration);
 
     /* Verify that all references are valid. This may do some .narinfo
        reads, but typically they'll already be cached. */
@@ -132,23 +201,6 @@ void BinaryCacheStore::addToStore(const ValidPathInfo & info, Source & narSource
                 printStorePath(info.path), printStorePath(ref));
         }
 
-    assert(nar->compare(0, narMagic.size(), narMagic) == 0);
-
-    auto narInfo = make_ref<NarInfo>(info);
-
-    narInfo->narSize = nar->size();
-    narInfo->narHash = hashString(htSHA256, *nar);
-
-    if (info.narHash && info.narHash != narInfo->narHash)
-        throw Error("refusing to copy corrupted path '%1%' to binary cache", printStorePath(info.path));
-
-    auto accessor_ = std::dynamic_pointer_cast<RemoteFSAccessor>(accessor);
-
-    auto narAccessor = makeNarAccessor(nar);
-
-    if (accessor_)
-        accessor_->addToCache(printStorePath(info.path), *nar, narAccessor);
-
     /* Optionally write a JSON file containing a listing of the
        contents of the NAR. */
     if (writeNARListing) {
@@ -160,33 +212,13 @@ void BinaryCacheStore::addToStore(const ValidPathInfo & info, Source & narSource
 
             {
                 auto res = jsonRoot.placeholder("root");
-                listNar(res, narAccessor, "", true);
+                listNar(res, ref<FSAccessor>(narAccessor), "", true);
             }
         }
 
         upsertFile(std::string(info.path.to_string()) + ".ls", jsonOut.str(), "application/json");
     }
 
-    /* Compress the NAR. */
-    narInfo->compression = compression;
-    auto now1 = std::chrono::steady_clock::now();
-    auto narCompressed = compress(compression, *nar, parallelCompression);
-    auto now2 = std::chrono::steady_clock::now();
-    narInfo->fileHash = hashString(htSHA256, *narCompressed);
-    narInfo->fileSize = narCompressed->size();
-
-    auto duration = std::chrono::duration_cast<std::chrono::milliseconds>(now2 - now1).count();
-    printMsg(lvlTalkative, "copying path '%1%' (%2% bytes, compressed %3$.1f%% in %4% ms) to binary cache",
-        printStorePath(narInfo->path), narInfo->narSize,
-        ((1.0 - (double) narCompressed->size() / nar->size()) * 100.0),
-        duration);
-
-    narInfo->url = "nar/" + narInfo->fileHash.to_string(Base32, false) + ".nar"
-        + (compression == "xz" ? ".xz" :
-           compression == "bzip2" ? ".bz2" :
-           compression == "br" ? ".br" :
-           "");
-
     /* Optionally maintain an index of DWARF debug info files
        consisting of JSON files named 'debuginfo/<build-id>' that
        specify the NAR file and member containing the debug info. */
@@ -247,12 +279,14 @@ void BinaryCacheStore::addToStore(const ValidPathInfo & info, Source & narSource
     /* Atomically write the NAR file. */
     if (repair || !fileExists(narInfo->url)) {
         stats.narWrite++;
-        upsertFile(narInfo->url, *narCompressed, "application/x-nix-nar");
+        upsertFile(narInfo->url,
+            std::make_shared<std::fstream>(fnTemp, std::ios_base::in),
+            "application/x-nix-nar");
     } else
         stats.narWriteAverted++;
 
-    stats.narWriteBytes += nar->size();
-    stats.narWriteCompressedBytes += narCompressed->size();
+    stats.narWriteBytes += info.narSize;
+    stats.narWriteCompressedBytes += fileSize;
     stats.narWriteCompressionTimeMs += duration;
 
     /* Atomically write the NAR info file.*/
@@ -338,7 +372,7 @@ StorePath BinaryCacheStore::addToStore(const string & name, const Path & srcPath
        method for very large paths, but `copyPath' is mainly used for
        small files. */
     StringSink sink;
-    Hash h;
+    std::optional<Hash> h;
     if (method == FileIngestionMethod::Recursive) {
         dumpPath(srcPath, sink, filter);
         h = hashString(hashAlgo, *sink.s);
@@ -348,10 +382,10 @@ StorePath BinaryCacheStore::addToStore(const string & name, const Path & srcPath
         h = hashString(hashAlgo, s);
     }
 
-    ValidPathInfo info(makeFixedOutputPath(method, h, name));
+    ValidPathInfo info(makeFixedOutputPath(method, *h, name));
 
     auto source = StringSource { *sink.s };
-    addToStore(info, source, repair, CheckSigs, nullptr);
+    addToStore(info, source, repair, CheckSigs);
 
     return std::move(info.path);
 }
@@ -366,7 +400,7 @@ StorePath BinaryCacheStore::addTextToStore(const string & name, const string & s
         StringSink sink;
         dumpString(s, sink);
         auto source = StringSource { *sink.s };
-        addToStore(info, source, repair, CheckSigs, nullptr);
+        addToStore(info, source, repair, CheckSigs);
     }
 
     return std::move(info.path);

src/libstore/binary-cache-store.hh

@@ -36,9 +36,13 @@ public:
     virtual bool fileExists(const std::string & path) = 0;
 
     virtual void upsertFile(const std::string & path,
-        const std::string & data,
+        std::shared_ptr<std::basic_iostream<char>> istream,
         const std::string & mimeType) = 0;
 
+    void upsertFile(const std::string & path,
+        std::string && data,
+        const std::string & mimeType);
+
     /* Note: subclasses must implement at least one of the two
        following getFile() methods. */
 
@@ -75,8 +79,7 @@ public:
     { unsupported("queryPathFromHashPart"); }
 
     void addToStore(const ValidPathInfo & info, Source & narSource,
-        RepairFlag repair, CheckSigsFlag checkSigs,
-        std::shared_ptr<FSAccessor> accessor) override;
+        RepairFlag repair, CheckSigsFlag checkSigs) override;
 
     StorePath addToStore(const string & name, const Path & srcPath,
         FileIngestionMethod method, HashType hashAlgo,

src/libstore/build.cc

@@ -1047,7 +1047,7 @@ DerivationGoal::DerivationGoal(const StorePath & drvPath, const BasicDerivation
 {
     this->drv = std::make_unique<BasicDerivation>(BasicDerivation(drv));
     state = &DerivationGoal::haveDerivation;
-    name = fmt("building of %s", worker.store.showPaths(drv.outputPaths()));
+    name = fmt("building of %s", worker.store.showPaths(drv.outputPaths(worker.store)));
     trace("created");
 
     mcExpectedBuilds = std::make_unique<MaintainCount<uint64_t>>(worker.expectedBuilds);
@@ -1182,7 +1182,7 @@ void DerivationGoal::haveDerivation()
     retrySubstitution = false;
 
     for (auto & i : drv->outputs)
-        worker.store.addTempRoot(i.second.path);
+        worker.store.addTempRoot(i.second.path(worker.store, drv->name));
 
     /* Check what outputs paths are not already valid. */
     auto invalidOutputs = checkPathValidity(false, buildMode == bmRepair);
@@ -1290,12 +1290,12 @@ void DerivationGoal::repairClosure()
     StorePathSet outputClosure;
     for (auto & i : drv->outputs) {
         if (!wantOutput(i.first, wantedOutputs)) continue;
-        worker.store.computeFSClosure(i.second.path, outputClosure);
+        worker.store.computeFSClosure(i.second.path(worker.store, drv->name), outputClosure);
     }
 
     /* Filter out our own outputs (which we have already checked). */
     for (auto & i : drv->outputs)
-        outputClosure.erase(i.second.path);
+        outputClosure.erase(i.second.path(worker.store, drv->name));
 
     /* Get all dependencies of this derivation so that we know which
        derivation is responsible for which path in the output
@@ -1307,7 +1307,7 @@ void DerivationGoal::repairClosure()
         if (i.isDerivation()) {
             Derivation drv = worker.store.derivationFromPath(i);
             for (auto & j : drv.outputs)
-                outputsToDrv.insert_or_assign(j.second.path, i);
+                outputsToDrv.insert_or_assign(j.second.path(worker.store, drv.name), i);
         }
 
     /* Check each path (slow!). */
@@ -1379,7 +1379,7 @@ void DerivationGoal::inputsRealised()
             for (auto & j : i.second) {
                 auto k = inDrv.outputs.find(j);
                 if (k != inDrv.outputs.end())
-                    worker.store.computeFSClosure(k->second.path, inputPaths);
+                    worker.store.computeFSClosure(k->second.path(worker.store, inDrv.name), inputPaths);
                 else
                     throw Error(
                         "derivation '%s' requires non-existent output '%s' from input derivation '%s'",
@@ -1432,7 +1432,7 @@ void DerivationGoal::tryToBuild()
        goal can start a build, and if not, the main loop will sleep a
        few seconds and then retry this goal. */
     PathSet lockFiles;
-    for (auto & outPath : drv->outputPaths())
+    for (auto & outPath : drv->outputPaths(worker.store))
         lockFiles.insert(worker.store.Store::toRealPath(outPath));
 
     if (!outputLocks.lockPaths(lockFiles, "", false)) {
@@ -1460,16 +1460,16 @@ void DerivationGoal::tryToBuild()
         return;
     }
 
-    missingPaths = drv->outputPaths();
+    missingPaths = drv->outputPaths(worker.store);
     if (buildMode != bmCheck)
         for (auto & i : validPaths) missingPaths.erase(i);
 
     /* If any of the outputs already exist but are not valid, delete
        them. */
     for (auto & i : drv->outputs) {
-        if (worker.store.isValidPath(i.second.path)) continue;
-        debug("removing invalid path '%s'", worker.store.printStorePath(i.second.path));
-        deletePath(worker.store.Store::toRealPath(i.second.path));
+        if (worker.store.isValidPath(i.second.path(worker.store, drv->name))) continue;
+        debug("removing invalid path '%s'", worker.store.printStorePath(i.second.path(worker.store, drv->name)));
+        deletePath(worker.store.Store::toRealPath(i.second.path(worker.store, drv->name)));
     }
 
     /* Don't do a remote build if the derivation has the attribute
@@ -1646,13 +1646,13 @@ void DerivationGoal::buildDone()
                So instead, check if the disk is (nearly) full now.  If
                so, we don't mark this build as a permanent failure. */
 #if HAVE_STATVFS
-            unsigned long long required = 8ULL * 1024 * 1024; // FIXME: make configurable
+            uint64_t required = 8ULL * 1024 * 1024; // FIXME: make configurable
             struct statvfs st;
             if (statvfs(worker.store.realStoreDir.c_str(), &st) == 0 &&
-                (unsigned long long) st.f_bavail * st.f_bsize < required)
+                (uint64_t) st.f_bavail * st.f_bsize < required)
                 diskFull = true;
             if (statvfs(tmpDir.c_str(), &st) == 0 &&
-                (unsigned long long) st.f_bavail * st.f_bsize < required)
+                (uint64_t) st.f_bavail * st.f_bsize < required)
                 diskFull = true;
 #endif
 
@@ -1692,7 +1692,7 @@ void DerivationGoal::buildDone()
                 fmt("running post-build-hook '%s'", settings.postBuildHook),
                 Logger::Fields{worker.store.printStorePath(drvPath)});
             PushActivity pact(act.id);
-            auto outputPaths = drv->outputPaths();
+            auto outputPaths = drv->outputPaths(worker.store);
             std::map<std::string, std::string> hookEnvironment = getEnv();
 
             hookEnvironment.emplace("DRV_PATH", worker.store.printStorePath(drvPath));
@@ -1920,7 +1920,7 @@ StorePathSet DerivationGoal::exportReferences(const StorePathSet & storePaths)
         if (j.isDerivation()) {
             Derivation drv = worker.store.derivationFromPath(j);
             for (auto & k : drv.outputs)
-                worker.store.computeFSClosure(k.second.path, paths);
+                worker.store.computeFSClosure(k.second.path(worker.store, drv.name), paths);
         }
     }
 
@@ -2015,7 +2015,7 @@ void DerivationGoal::startBuilder()
 
     /* Substitute output placeholders with the actual output paths. */
     for (auto & output : drv->outputs)
-        inputRewrites[hashPlaceholder(output.first)] = worker.store.printStorePath(output.second.path);
+        inputRewrites[hashPlaceholder(output.first)] = worker.store.printStorePath(output.second.path(worker.store, drv->name));
 
     /* Construct the environment passed to the builder. */
     initEnv();
@@ -2044,7 +2044,7 @@ void DerivationGoal::startBuilder()
             auto storePathS = *i++;
             if (!worker.store.isInStore(storePathS))
                 throw BuildError("'exportReferencesGraph' contains a non-store path '%1%'", storePathS);
-            auto storePath = worker.store.parseStorePath(worker.store.toStorePath(storePathS));
+            auto storePath = worker.store.toStorePath(storePathS).first;
 
             /* Write closure info to <fileName>. */
             writeFile(tmpDir + "/" + fileName,
@@ -2083,7 +2083,7 @@ void DerivationGoal::startBuilder()
         for (auto & i : dirsInChroot)
             try {
                 if (worker.store.isInStore(i.second.source))
-                    worker.store.computeFSClosure(worker.store.parseStorePath(worker.store.toStorePath(i.second.source)), closure);
+                    worker.store.computeFSClosure(worker.store.toStorePath(i.second.source).first, closure);
             } catch (InvalidPath & e) {
             } catch (Error & e) {
                 throw Error("while processing 'sandbox-paths': %s", e.what());
@@ -2200,7 +2200,7 @@ void DerivationGoal::startBuilder()
            (typically the dependencies of /bin/sh).  Throw them
            out. */
         for (auto & i : drv->outputs)
-            dirsInChroot.erase(worker.store.printStorePath(i.second.path));
+            dirsInChroot.erase(worker.store.printStorePath(i.second.path(worker.store, drv->name)));
 
 #elif __APPLE__
         /* We don't really have any parent prep work to do (yet?)
@@ -2613,7 +2613,7 @@ void DerivationGoal::writeStructuredAttrs()
     /* Add an "outputs" object containing the output paths. */
     nlohmann::json outputs;
     for (auto & i : drv->outputs)
-        outputs[i.first] = rewriteStrings(worker.store.printStorePath(i.second.path), inputRewrites);
+        outputs[i.first] = rewriteStrings(worker.store.printStorePath(i.second.path(worker.store, drv->name)), inputRewrites);
     json["outputs"] = outputs;
 
     /* Handle exportReferencesGraph. */
@@ -2768,14 +2768,13 @@ struct RestrictedStore : public LocalFSStore
     { throw Error("addToStore"); }
 
     void addToStore(const ValidPathInfo & info, Source & narSource,
-        RepairFlag repair = NoRepair, CheckSigsFlag checkSigs = CheckSigs,
-        std::shared_ptr<FSAccessor> accessor = 0) override
+        RepairFlag repair = NoRepair, CheckSigsFlag checkSigs = CheckSigs) override
     {
-        next->addToStore(info, narSource, repair, checkSigs, accessor);
+        next->addToStore(info, narSource, repair, checkSigs);
         goal.addDependency(info.path);
     }
 
-    StorePath addToStoreFromDump(const string & dump, const string & name,
+    StorePath addToStoreFromDump(Source & dump, const string & name,
         FileIngestionMethod method = FileIngestionMethod::Recursive, HashType hashAlgo = htSHA256, RepairFlag repair = NoRepair) override
     {
         auto path = next->addToStoreFromDump(dump, name, method, hashAlgo, repair);
@@ -2818,7 +2817,7 @@ struct RestrictedStore : public LocalFSStore
                 auto drv = derivationFromPath(path.path);
                 for (auto & output : drv.outputs)
                     if (wantOutput(output.first, path.outputs))
-                        newPaths.insert(output.second.path);
+                        newPaths.insert(output.second.path(*this, drv.name));
             } else if (!goal.isAllowed(path.path))
                 throw InvalidPath("cannot build unknown path '%s' in recursive Nix", printStorePath(path.path));
         }
@@ -2852,7 +2851,7 @@ struct RestrictedStore : public LocalFSStore
 
     void queryMissing(const std::vector<StorePathWithOutputs> & targets,
         StorePathSet & willBuild, StorePathSet & willSubstitute, StorePathSet & unknown,
-        unsigned long long & downloadSize, unsigned long long & narSize) override
+        uint64_t & downloadSize, uint64_t & narSize) override
     {
         /* This is slightly impure since it leaks information to the
            client about what paths will be built/substituted or are
@@ -3580,7 +3579,7 @@ StorePathSet parseReferenceSpecifiers(Store & store, const BasicDerivation & drv
         if (store.isStorePath(i))
             result.insert(store.parseStorePath(i));
         else if (drv.outputs.count(i))
-            result.insert(drv.outputs.find(i)->second.path);
+            result.insert(drv.outputs.find(i)->second.path(store, drv.name));
         else throw BuildError("derivation contains an illegal reference specifier '%s'", i);
     }
     return result;
@@ -3618,7 +3617,7 @@ void DerivationGoal::registerOutputs()
     if (hook) {
         bool allValid = true;
         for (auto & i : drv->outputs)
-            if (!worker.store.isValidPath(i.second.path)) allValid = false;
+            if (!worker.store.isValidPath(i.second.path(worker.store, drv->name))) allValid = false;
         if (allValid) return;
     }
 
@@ -3639,23 +3638,23 @@ void DerivationGoal::registerOutputs()
        Nix calls. */
     StorePathSet referenceablePaths;
     for (auto & p : inputPaths) referenceablePaths.insert(p);
-    for (auto & i : drv->outputs) referenceablePaths.insert(i.second.path);
+    for (auto & i : drv->outputs) referenceablePaths.insert(i.second.path(worker.store, drv->name));
     for (auto & p : addedPaths) referenceablePaths.insert(p);
 
     /* Check whether the output paths were created, and grep each
        output path to determine what other paths it references.  Also make all
        output paths read-only. */
     for (auto & i : drv->outputs) {
-        auto path = worker.store.printStorePath(i.second.path);
-        if (!missingPaths.count(i.second.path)) continue;
+        auto path = worker.store.printStorePath(i.second.path(worker.store, drv->name));
+        if (!missingPaths.count(i.second.path(worker.store, drv->name))) continue;
 
         Path actualPath = path;
         if (needsHashRewrite()) {
-            auto r = redirectedOutputs.find(i.second.path);
+            auto r = redirectedOutputs.find(i.second.path(worker.store, drv->name));
             if (r != redirectedOutputs.end()) {
                 auto redirected = worker.store.Store::toRealPath(r->second);
                 if (buildMode == bmRepair
-                    && redirectedBadOutputs.count(i.second.path)
+                    && redirectedBadOutputs.count(i.second.path(worker.store, drv->name))
                     && pathExists(redirected))
                     replaceValidPath(path, redirected);
                 if (buildMode == bmCheck)
@@ -3724,7 +3723,9 @@ void DerivationGoal::registerOutputs()
 
         if (fixedOutput) {
 
-            if (i.second.hash->method == FileIngestionMethod::Flat) {
+            FixedOutputHash outputHash = std::get<DerivationOutputFixed>(i.second.output).hash;
+
+            if (outputHash.method == FileIngestionMethod::Flat) {
                 /* The output path should be a regular file without execute permission. */
                 if (!S_ISREG(st.st_mode) || (st.st_mode & S_IXUSR) != 0)
                     throw BuildError(
@@ -3735,13 +3736,13 @@ void DerivationGoal::registerOutputs()
 
             /* Check the hash. In hash mode, move the path produced by
                the derivation to its content-addressed location. */
-            Hash h2 = i.second.hash->method == FileIngestionMethod::Recursive
-                ? hashPath(*i.second.hash->hash.type, actualPath).first
-                : hashFile(*i.second.hash->hash.type, actualPath);
+            Hash h2 = outputHash.method == FileIngestionMethod::Recursive
+                ? hashPath(outputHash.hash.type, actualPath).first
+                : hashFile(outputHash.hash.type, actualPath);
 
-            auto dest = worker.store.makeFixedOutputPath(i.second.hash->method, h2, i.second.path.name());
+            auto dest = worker.store.makeFixedOutputPath(outputHash.method, h2, i.second.path(worker.store, drv->name).name());
 
-            if (i.second.hash->hash != h2) {
+            if (outputHash.hash != h2) {
 
                 /* Throw an error after registering the path as
                    valid. */
@@ -3749,7 +3750,7 @@ void DerivationGoal::registerOutputs()
                 delayedException = std::make_exception_ptr(
                     BuildError("hash mismatch in fixed-output derivation '%s':\n  wanted: %s\n  got:    %s",
                         worker.store.printStorePath(dest),
-                        i.second.hash->hash.to_string(SRI, true),
+                        outputHash.hash.to_string(SRI, true),
                         h2.to_string(SRI, true)));
 
                 Path actualDest = worker.store.Store::toRealPath(dest);
@@ -3771,7 +3772,7 @@ void DerivationGoal::registerOutputs()
                 assert(worker.store.parseStorePath(path) == dest);
 
             ca = FixedOutputHash {
-                .method = i.second.hash->method,
+                .method = outputHash.method,
                 .hash = h2,
             };
         }
@@ -3786,8 +3787,10 @@ void DerivationGoal::registerOutputs()
            time.  The hash is stored in the database so that we can
            verify later on whether nobody has messed with the store. */
         debug("scanning for references inside '%1%'", path);
-        HashResult hash;
-        auto references = worker.store.parseStorePathSet(scanForReferences(actualPath, worker.store.printStorePathSet(referenceablePaths), hash));
+        // HashResult hash;
+        auto pathSetAndHash = scanForReferences(actualPath, worker.store.printStorePathSet(referenceablePaths));
+        auto references = worker.store.parseStorePathSet(pathSetAndHash.first);
+        HashResult hash = pathSetAndHash.second;
 
         if (buildMode == bmCheck) {
             if (!worker.store.isValidPath(worker.store.parseStorePath(path))) continue;
@@ -3895,7 +3898,7 @@ void DerivationGoal::registerOutputs()
     /* If this is the first round of several, then move the output out of the way. */
     if (nrRounds > 1 && curRound == 1 && curRound < nrRounds && keepPreviousRound) {
         for (auto & i : drv->outputs) {
-            auto path = worker.store.printStorePath(i.second.path);
+            auto path = worker.store.printStorePath(i.second.path(worker.store, drv->name));
             Path prev = path + checkSuffix;
             deletePath(prev);
             Path dst = path + checkSuffix;
@@ -3913,7 +3916,7 @@ void DerivationGoal::registerOutputs()
        if the result was not determistic? */
     if (curRound == nrRounds) {
         for (auto & i : drv->outputs) {
-            Path prev = worker.store.printStorePath(i.second.path) + checkSuffix;
+            Path prev = worker.store.printStorePath(i.second.path(worker.store, drv->name)) + checkSuffix;
             deletePath(prev);
         }
     }
@@ -4214,9 +4217,9 @@ StorePathSet DerivationGoal::checkPathValidity(bool returnValid, bool checkHash)
     for (auto & i : drv->outputs) {
         if (!wantOutput(i.first, wantedOutputs)) continue;
         bool good =
-            worker.store.isValidPath(i.second.path) &&
-            (!checkHash || worker.pathContentsGood(i.second.path));
-        if (good == returnValid) result.insert(i.second.path);
+            worker.store.isValidPath(i.second.path(worker.store, drv->name)) &&
+            (!checkHash || worker.pathContentsGood(i.second.path(worker.store, drv->name)));
+        if (good == returnValid) result.insert(i.second.path(worker.store, drv->name));
     }
     return result;
 }
@@ -5034,7 +5037,7 @@ bool Worker::pathContentsGood(const StorePath & path)
     if (!pathExists(store.printStorePath(path)))
         res = false;
     else {
-        HashResult current = hashPath(*info->narHash.type, store.printStorePath(path));
+        HashResult current = hashPath(info->narHash->type, store.printStorePath(path));
         Hash nullHash(htSHA256);
         res = info->narHash == nullHash || info->narHash == current.first;
     }
@@ -5060,7 +5063,7 @@ void Worker::markContentsGood(const StorePath & path)
 static void primeCache(Store & store, const std::vector<StorePathWithOutputs> & paths)
 {
     StorePathSet willBuild, willSubstitute, unknown;
-    unsigned long long downloadSize, narSize;
+    uint64_t downloadSize, narSize;
     store.queryMissing(paths, willBuild, willSubstitute, unknown, downloadSize, narSize);
 
     if (!willBuild.empty() && 0 == settings.maxBuildJobs && getMachines().empty())

src/libstore/builtins/buildenv.hh

@@ -9,7 +9,7 @@ struct Package {
     Path path;
     bool active;
     int priority;
-    Package(Path path, bool active, int priority) : path{path}, active{active}, priority{priority} {}
+    Package(const Path & path, bool active, int priority) : path{path}, active{active}, priority{priority} {}
 };
 
 typedef std::vector<Package> Packages;

src/libstore/content-address.cc

@@ -3,7 +3,7 @@
 namespace nix {
 
 std::string FixedOutputHash::printMethodAlgo() const {
-    return makeFileIngestionPrefix(method) + printHashType(*hash.type);
+    return makeFileIngestionPrefix(method) + printHashType(hash.type);
 }
 
 std::string makeFileIngestionPrefix(const FileIngestionMethod m) {
@@ -46,7 +46,7 @@ ContentAddress parseContentAddress(std::string_view rawCa) {
         if (prefix == "text") {
             auto hashTypeAndHash = rawCa.substr(prefixSeparator+1, string::npos);
             Hash hash = Hash(string(hashTypeAndHash));
-            if (*hash.type != htSHA256) {
+            if (hash.type != htSHA256) {
                 throw Error("parseContentAddress: the text hash should have type SHA256");
             }
             return TextHash { hash };

src/libstore/daemon.cc

@@ -86,7 +86,7 @@ struct TunnelLogger : public Logger
     }
 
     /* startWork() means that we're starting an operation for which we
-      want to send out stderr to the client. */
+       want to send out stderr to the client. */
     void startWork()
     {
         auto state(state_.lock());
@@ -173,31 +173,6 @@ struct TunnelSource : BufferedSource
     }
 };
 
-/* If the NAR archive contains a single file at top-level, then save
-   the contents of the file to `s'.  Otherwise barf. */
-struct RetrieveRegularNARSink : ParseSink
-{
-    bool regular;
-    string s;
-
-    RetrieveRegularNARSink() : regular(true) { }
-
-    void createDirectory(const Path & path)
-    {
-        regular = false;
-    }
-
-    void receiveContents(unsigned char * data, unsigned int len)
-    {
-        s.append((const char *) data, len);
-    }
-
-    void createSymlink(const Path & path, const string & target)
-    {
-        regular = false;
-    }
-};
-
 struct ClientSettings
 {
     bool keepFailed;
@@ -314,7 +289,7 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
         logger->startWork();
         auto hash = store->queryPathInfo(path)->narHash;
         logger->stopWork();
-        to << hash.to_string(Base16, false);
+        to << hash->to_string(Base16, false);
         break;
     }
 
@@ -375,42 +350,44 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
     }
 
     case wopAddToStore: {
-        std::string s, baseName;
+        HashType hashAlgo;
+        std::string baseName;
         FileIngestionMethod method;
         {
-            bool fixed; uint8_t recursive;
-            from >> baseName >> fixed /* obsolete */ >> recursive >> s;
+            bool fixed;
+            uint8_t recursive;
+            std::string hashAlgoRaw;
+            from >> baseName >> fixed /* obsolete */ >> recursive >> hashAlgoRaw;
             if (recursive > (uint8_t) FileIngestionMethod::Recursive)
                 throw Error("unsupported FileIngestionMethod with value of %i; you may need to upgrade nix-daemon", recursive);
             method = FileIngestionMethod { recursive };
             /* Compatibility hack. */
             if (!fixed) {
-                s = "sha256";
+                hashAlgoRaw = "sha256";
                 method = FileIngestionMethod::Recursive;
             }
+            hashAlgo = parseHashType(hashAlgoRaw);
         }
-        HashType hashAlgo = parseHashType(s);
 
-        TeeSource savedNAR(from);
-        RetrieveRegularNARSink savedRegular;
+        StringSink saved;
+        TeeSource savedNARSource(from, saved);
+        RetrieveRegularNARSink savedRegular { saved };
 
         if (method == FileIngestionMethod::Recursive) {
             /* Get the entire NAR dump from the client and save it to
                a string so that we can pass it to
                addToStoreFromDump(). */
             ParseSink sink; /* null sink; just parse the NAR */
-            parseDump(sink, savedNAR);
+            parseDump(sink, savedNARSource);
         } else
             parseDump(savedRegular, from);
 
         logger->startWork();
         if (!savedRegular.regular) throw Error("regular file expected");
 
-        auto path = store->addToStoreFromDump(
-            method == FileIngestionMethod::Recursive ? *savedNAR.data : savedRegular.s,
-            baseName,
-            method,
-            hashAlgo);
+        // FIXME: try to stream directly from `from`.
+        StringSource dumpSource { *saved.s };
+        auto path = store->addToStoreFromDump(dumpSource, baseName, method, hashAlgo);
         logger->stopWork();
 
         to << store->printStorePath(path);
@@ -442,7 +419,7 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
     case wopImportPaths: {
         logger->startWork();
         TunnelSource source(from, to);
-        auto paths = store->importPaths(source, nullptr,
+        auto paths = store->importPaths(source,
             trusted ? NoCheckSigs : CheckSigs);
         logger->stopWork();
         Strings paths2;
@@ -474,7 +451,7 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
     case wopBuildDerivation: {
         auto drvPath = store->parseStorePath(readString(from));
         BasicDerivation drv;
-        readDerivation(from, *store, drv);
+        readDerivation(from, *store, drv, Derivation::nameFromPath(drvPath));
         BuildMode buildMode = (BuildMode) readInt(from);
         logger->startWork();
         if (!trusted)
@@ -661,7 +638,7 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
             if (GET_PROTOCOL_MINOR(clientVersion) >= 17)
                 to << 1;
             to << (info->deriver ? store->printStorePath(*info->deriver) : "")
-               << info->narHash.to_string(Base16, false);
+               << info->narHash->to_string(Base16, false);
             writeStorePaths(*store, to, info->references);
             to << info->registrationTime << info->narSize;
             if (GET_PROTOCOL_MINOR(clientVersion) >= 16) {
@@ -732,24 +709,84 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
         if (!trusted)
             info.ultimate = false;
 
-        std::string saved;
-        std::unique_ptr<Source> source;
-        if (GET_PROTOCOL_MINOR(clientVersion) >= 21)
-            source = std::make_unique<TunnelSource>(from, to);
-        else {
-            TeeSink tee(from);
-            parseDump(tee, tee.source);
-            saved = std::move(*tee.source.data);
-            source = std::make_unique<StringSource>(saved);
+        if (GET_PROTOCOL_MINOR(clientVersion) >= 23) {
+
+            struct FramedSource : Source
+            {
+                Source & from;
+                bool eof = false;
+                std::vector<unsigned char> pending;
+                size_t pos = 0;
+
+                FramedSource(Source & from) : from(from)
+                { }
+
+                ~FramedSource()
+                {
+                    if (!eof) {
+                        while (true) {
+                            auto n = readInt(from);
+                            if (!n) break;
+                            std::vector<unsigned char> data(n);
+                            from(data.data(), n);
+                        }
+                    }
+                }
+
+                size_t read(unsigned char * data, size_t len) override
+                {
+                    if (eof) throw EndOfFile("reached end of FramedSource");
+
+                    if (pos >= pending.size()) {
+                        size_t len = readInt(from);
+                        if (!len) {
+                            eof = true;
+                            return 0;
+                        }
+                        pending = std::vector<unsigned char>(len);
+                        pos = 0;
+                        from(pending.data(), len);
+                    }
+
+                    auto n = std::min(len, pending.size() - pos);
+                    memcpy(data, pending.data() + pos, n);
+                    pos += n;
+                    return n;
+                }
+            };
+
+            logger->startWork();
+
+            {
+                FramedSource source(from);
+                store->addToStore(info, source, (RepairFlag) repair,
+                    dontCheckSigs ? NoCheckSigs : CheckSigs);
+            }
+
+            logger->stopWork();
         }
 
-        logger->startWork();
+        else {
+            std::unique_ptr<Source> source;
+            if (GET_PROTOCOL_MINOR(clientVersion) >= 21)
+                source = std::make_unique<TunnelSource>(from, to);
+            else {
+                StringSink saved;
+                TeeSource tee { from, saved };
+                ParseSink ether;
+                parseDump(ether, tee);
+                source = std::make_unique<StringSource>(std::move(*saved.s));
+            }
 
-        // FIXME: race if addToStore doesn't read source?
-        store->addToStore(info, *source, (RepairFlag) repair,
-            dontCheckSigs ? NoCheckSigs : CheckSigs, nullptr);
+            logger->startWork();
+
+            // FIXME: race if addToStore doesn't read source?
+            store->addToStore(info, *source, (RepairFlag) repair,
+                dontCheckSigs ? NoCheckSigs : CheckSigs);
+
+            logger->stopWork();
+        }
 
-        logger->stopWork();
         break;
     }
 
@@ -759,7 +796,7 @@ static void performOp(TunnelLogger * logger, ref<Store> store,
             targets.push_back(store->parsePathWithOutputs(s));
         logger->startWork();
         StorePathSet willBuild, willSubstitute, unknown;
-        unsigned long long downloadSize, narSize;
+        uint64_t downloadSize, narSize;
         store->queryMissing(targets, willBuild, willSubstitute, unknown, downloadSize, narSize);
         logger->stopWork();
         writeStorePaths(*store, to, willBuild);

src/libstore/derivations.cc

@@ -4,16 +4,23 @@
 #include "util.hh"
 #include "worker-protocol.hh"
 #include "fs-accessor.hh"
-#include "istringstream_nocopy.hh"
 
 namespace nix {
 
-const StorePath & BasicDerivation::findOutput(const string & id) const
+// FIXME Put this somewhere?
+template<class... Ts> struct overloaded : Ts... { using Ts::operator()...; };
+template<class... Ts> overloaded(Ts...) -> overloaded<Ts...>;
+
+StorePath DerivationOutput::path(const Store & store, std::string_view drvName) const
 {
-    auto i = outputs.find(id);
-    if (i == outputs.end())
-        throw Error("derivation has no output '%s'", id);
-    return i->second.path;
+    return std::visit(overloaded {
+        [](DerivationOutputInputAddressed doi) {
+            return doi.path;
+        },
+        [&](DerivationOutputFixed dof) {
+            return store.makeFixedOutputPath(dof.hash.method, dof.hash.hash, drvName);
+        }
+    }, output);
 }
 
 
@@ -101,14 +108,13 @@ static StringSet parseStrings(std::istream & str, bool arePaths)
 }
 
 
-static DerivationOutput parseDerivationOutput(const Store & store, istringstream_nocopy & str)
+static DerivationOutput parseDerivationOutput(const Store & store, std::istringstream & str)
 {
     expect(str, ","); auto path = store.parseStorePath(parsePath(str));
     expect(str, ","); auto hashAlgo = parseString(str);
     expect(str, ","); const auto hash = parseString(str);
     expect(str, ")");
 
-    std::optional<FixedOutputHash> fsh;
     if (hashAlgo != "") {
         auto method = FileIngestionMethod::Flat;
         if (string(hashAlgo, 0, 2) == "r:") {
@@ -116,23 +122,30 @@ static DerivationOutput parseDerivationOutput(const Store & store, istringstream
             hashAlgo = string(hashAlgo, 2);
         }
         const HashType hashType = parseHashType(hashAlgo);
-        fsh = FixedOutputHash {
-            .method = std::move(method),
-            .hash = Hash(hash, hashType),
-        };
-    }
 
-    return DerivationOutput {
-        .path = std::move(path),
-        .hash = std::move(fsh),
-    };
+        return DerivationOutput {
+            .output = DerivationOutputFixed {
+                .hash = FixedOutputHash {
+                    .method = std::move(method),
+                    .hash = Hash(hash, hashType),
+                },
+            }
+        };
+    } else
+        return DerivationOutput {
+            .output = DerivationOutputInputAddressed {
+                .path = std::move(path),
+            }
+        };
 }
 
 
-static Derivation parseDerivation(const Store & store, const string & s)
+static Derivation parseDerivation(const Store & store, std::string && s, std::string_view name)
 {
     Derivation drv;
-    istringstream_nocopy str(s);
+    drv.name = name;
+
+    std::istringstream str(std::move(s));
     expect(str, "Derive([");
 
     /* Parse the list of outputs. */
@@ -175,10 +188,10 @@ static Derivation parseDerivation(const Store & store, const string & s)
 }
 
 
-Derivation readDerivation(const Store & store, const Path & drvPath)
+Derivation readDerivation(const Store & store, const Path & drvPath, std::string_view name)
 {
     try {
-        return parseDerivation(store, readFile(drvPath));
+        return parseDerivation(store, readFile(drvPath), name);
     } catch (FormatError & e) {
         throw Error("error parsing derivation '%1%': %2%", drvPath, e.msg());
     }
@@ -196,7 +209,7 @@ Derivation Store::readDerivation(const StorePath & drvPath)
 {
     auto accessor = getFSAccessor();
     try {
-        return parseDerivation(*this, accessor->readFile(printStorePath(drvPath)));
+        return parseDerivation(*this, accessor->readFile(printStorePath(drvPath)), Derivation::nameFromPath(drvPath));
     } catch (FormatError & e) {
         throw Error("error parsing derivation '%s': %s", printStorePath(drvPath), e.msg());
     }
@@ -264,10 +277,14 @@ string Derivation::unparse(const Store & store, bool maskOutputs,
     for (auto & i : outputs) {
         if (first) first = false; else s += ',';
         s += '('; printUnquotedString(s, i.first);
-        s += ','; printUnquotedString(s, maskOutputs ? "" : store.printStorePath(i.second.path));
-        s += ','; printUnquotedString(s, i.second.hash ? i.second.hash->printMethodAlgo() : "");
-        s += ','; printUnquotedString(s,
-            i.second.hash ? i.second.hash->hash.to_string(Base16, false) : "");
+        s += ','; printUnquotedString(s, maskOutputs ? "" : store.printStorePath(i.second.path(store, name)));
+        if (auto hash = std::get_if<DerivationOutputFixed>(&i.second.output)) {
+            s += ','; printUnquotedString(s, hash->hash.printMethodAlgo());
+            s += ','; printUnquotedString(s, hash->hash.hash.to_string(Base16, false));
+        } else {
+            s += ','; printUnquotedString(s, "");
+            s += ','; printUnquotedString(s, "");
+        }
         s += ')';
     }
 
@@ -323,7 +340,7 @@ bool BasicDerivation::isFixedOutput() const
 {
     return outputs.size() == 1 &&
         outputs.begin()->first == "out" &&
-        outputs.begin()->second.hash;
+        std::holds_alternative<DerivationOutputFixed>(outputs.begin()->second.output);
 }
 
 
@@ -355,10 +372,11 @@ Hash hashDerivationModulo(Store & store, const Derivation & drv, bool maskOutput
     /* Return a fixed hash for fixed-output derivations. */
     if (drv.isFixedOutput()) {
         DerivationOutputs::const_iterator i = drv.outputs.begin();
+        auto hash = std::get<DerivationOutputFixed>(i->second.output);
         return hashString(htSHA256, "fixed:out:"
-            + i->second.hash->printMethodAlgo() + ":"
-            + i->second.hash->hash.to_string(Base16, false) + ":"
-            + store.printStorePath(i->second.path));
+            + hash.hash.printMethodAlgo() + ":"
+            + hash.hash.hash.to_string(Base16, false) + ":"
+            + store.printStorePath(i->second.path(store, drv.name)));
     }
 
     /* For other derivations, replace the inputs paths with recursive
@@ -392,11 +410,11 @@ bool wantOutput(const string & output, const std::set<string> & wanted)
 }
 
 
-StorePathSet BasicDerivation::outputPaths() const
+StorePathSet BasicDerivation::outputPaths(const Store & store) const
 {
     StorePathSet paths;
     for (auto & i : outputs)
-        paths.insert(i.second.path);
+        paths.insert(i.second.path(store, name));
     return paths;
 }
 
@@ -406,7 +424,6 @@ static DerivationOutput readDerivationOutput(Source & in, const Store & store)
     auto hashAlgo = readString(in);
     auto hash = readString(in);
 
-    std::optional<FixedOutputHash> fsh;
     if (hashAlgo != "") {
         auto method = FileIngestionMethod::Flat;
         if (string(hashAlgo, 0, 2) == "r:") {
@@ -414,16 +431,20 @@ static DerivationOutput readDerivationOutput(Source & in, const Store & store)
             hashAlgo = string(hashAlgo, 2);
         }
         auto hashType = parseHashType(hashAlgo);
-        fsh = FixedOutputHash {
-            .method = std::move(method),
-            .hash = Hash(hash, hashType),
+        return DerivationOutput {
+            .output = DerivationOutputFixed {
+                .hash = FixedOutputHash {
+                    .method = std::move(method),
+                    .hash = Hash(hash, hashType),
+                },
+            }
+        };
+    } else
+        return DerivationOutput {
+            .output = DerivationOutputInputAddressed {
+                .path = std::move(path),
+            }
         };
-    }
-
-    return DerivationOutput {
-        .path = std::move(path),
-        .hash = std::move(fsh),
-    };
 }
 
 StringSet BasicDerivation::outputNames() const
@@ -435,8 +456,19 @@ StringSet BasicDerivation::outputNames() const
 }
 
 
-Source & readDerivation(Source & in, const Store & store, BasicDerivation & drv)
+std::string_view BasicDerivation::nameFromPath(const StorePath & drvPath) {
+    auto nameWithSuffix = drvPath.name();
+    constexpr std::string_view extension = ".drv";
+    assert(hasSuffix(nameWithSuffix, extension));
+    nameWithSuffix.remove_suffix(extension.size());
+    return nameWithSuffix;
+}
+
+
+Source & readDerivation(Source & in, const Store & store, BasicDerivation & drv, std::string_view name)
 {
+    drv.name = name;
+
     drv.outputs.clear();
     auto nr = readNum<size_t>(in);
     for (size_t n = 0; n < nr; n++) {
@@ -465,10 +497,10 @@ void writeDerivation(Sink & out, const Store & store, const BasicDerivation & dr
     out << drv.outputs.size();
     for (auto & i : drv.outputs) {
         out << i.first
-            << store.printStorePath(i.second.path);
-        if (i.second.hash) {
-            out << i.second.hash->printMethodAlgo()
-                << i.second.hash->hash.to_string(Base16, false);
+            << store.printStorePath(i.second.path(store, drv.name));
+        if (auto hash = std::get_if<DerivationOutputFixed>(&i.second.output)) {
+            out << hash->hash.printMethodAlgo()
+                << hash->hash.hash.to_string(Base16, false);
         } else {
             out << "" << "";
         }

src/libstore/derivations.hh

@@ -13,10 +13,20 @@ namespace nix {
 
 /* Abstract syntax of derivations. */
 
-struct DerivationOutput
+struct DerivationOutputInputAddressed
 {
     StorePath path;
-    std::optional<FixedOutputHash> hash; /* hash used for expected hash computation */
+};
+
+struct DerivationOutputFixed
+{
+    FixedOutputHash hash; /* hash used for expected hash computation */
+};
+
+struct DerivationOutput
+{
+    std::variant<DerivationOutputInputAddressed, DerivationOutputFixed> output;
+    StorePath path(const Store & store, std::string_view drvName) const;
 };
 
 typedef std::map<string, DerivationOutput> DerivationOutputs;
@@ -35,24 +45,23 @@ struct BasicDerivation
     Path builder;
     Strings args;
     StringPairs env;
+    std::string name;
 
     BasicDerivation() { }
     virtual ~BasicDerivation() { };
 
-    /* Return the path corresponding to the output identifier `id' in
-       the given derivation. */
-    const StorePath & findOutput(const std::string & id) const;
-
     bool isBuiltin() const;
 
     /* Return true iff this is a fixed-output derivation. */
     bool isFixedOutput() const;
 
     /* Return the output paths of a derivation. */
-    StorePathSet outputPaths() const;
+    StorePathSet outputPaths(const Store & store) const;
 
     /* Return the output names of a derivation. */
     StringSet outputNames() const;
+
+    static std::string_view nameFromPath(const StorePath & storePath);
 };
 
 struct Derivation : BasicDerivation
@@ -76,7 +85,7 @@ StorePath writeDerivation(ref<Store> store,
     const Derivation & drv, std::string_view name, RepairFlag repair = NoRepair);
 
 /* Read a derivation from a file. */
-Derivation readDerivation(const Store & store, const Path & drvPath);
+Derivation readDerivation(const Store & store, const Path & drvPath, std::string_view name);
 
 // FIXME: remove
 bool isDerivation(const string & fileName);
@@ -93,7 +102,7 @@ bool wantOutput(const string & output, const std::set<string> & wanted);
 struct Source;
 struct Sink;
 
-Source & readDerivation(Source & in, const Store & store, BasicDerivation & drv);
+Source & readDerivation(Source & in, const Store & store, BasicDerivation & drv, std::string_view name);
 void writeDerivation(Sink & out, const Store & store, const BasicDerivation & drv);
 
 std::string hashPlaceholder(const std::string & outputName);

src/libstore/export-import.cc

@@ -7,24 +7,6 @@
 
 namespace nix {
 
-struct HashAndWriteSink : Sink
-{
-    Sink & writeSink;
-    HashSink hashSink;
-    HashAndWriteSink(Sink & writeSink) : writeSink(writeSink), hashSink(htSHA256)
-    {
-    }
-    virtual void operator () (const unsigned char * data, size_t len)
-    {
-        writeSink(data, len);
-        hashSink(data, len);
-    }
-    Hash currentHash()
-    {
-        return hashSink.currentHash().first;
-    }
-};
-
 void Store::exportPaths(const StorePathSet & paths, Sink & sink)
 {
     auto sorted = topoSortPaths(paths);
@@ -47,28 +29,29 @@ void Store::exportPath(const StorePath & path, Sink & sink)
 {
     auto info = queryPathInfo(path);
 
-    HashAndWriteSink hashAndWriteSink(sink);
+    HashSink hashSink(htSHA256);
+    TeeSink teeSink(sink, hashSink);
 
-    narFromPath(path, hashAndWriteSink);
+    narFromPath(path, teeSink);
 
     /* Refuse to export paths that have changed.  This prevents
        filesystem corruption from spreading to other machines.
        Don't complain if the stored hash is zero (unknown). */
-    Hash hash = hashAndWriteSink.currentHash();
-    if (hash != info->narHash && info->narHash != Hash(*info->narHash.type))
+    Hash hash = hashSink.currentHash().first;
+    if (hash != info->narHash && info->narHash != Hash(info->narHash->type))
         throw Error("hash of path '%s' has changed from '%s' to '%s'!",
-            printStorePath(path), info->narHash.to_string(Base32, true), hash.to_string(Base32, true));
+            printStorePath(path), info->narHash->to_string(Base32, true), hash.to_string(Base32, true));
 
-    hashAndWriteSink
+    teeSink
         << exportMagic
         << printStorePath(path);
-    writeStorePaths(*this, hashAndWriteSink, info->references);
-    hashAndWriteSink
+    writeStorePaths(*this, teeSink, info->references);
+    teeSink
         << (info->deriver ? printStorePath(*info->deriver) : "")
         << 0;
 }
 
-StorePaths Store::importPaths(Source & source, std::shared_ptr<FSAccessor> accessor, CheckSigsFlag checkSigs)
+StorePaths Store::importPaths(Source & source, CheckSigsFlag checkSigs)
 {
     StorePaths res;
     while (true) {
@@ -77,8 +60,10 @@ StorePaths Store::importPaths(Source & source, std::shared_ptr<FSAccessor> acces
         if (n != 1) throw Error("input doesn't look like something created by 'nix-store --export'");
 
         /* Extract the NAR from the source. */
-        TeeSink tee(source);
-        parseDump(tee, tee.source);
+        StringSink saved;
+        TeeSource tee { source, saved };
+        ParseSink ether;
+        parseDump(ether, tee);
 
         uint32_t magic = readInt(source);
         if (magic != exportMagic)
@@ -94,16 +79,16 @@ StorePaths Store::importPaths(Source & source, std::shared_ptr<FSAccessor> acces
         if (deriver != "")
             info.deriver = parseStorePath(deriver);
 
-        info.narHash = hashString(htSHA256, *tee.source.data);
-        info.narSize = tee.source.data->size();
+        info.narHash = hashString(htSHA256, *saved.s);
+        info.narSize = saved.s->size();
 
         // Ignore optional legacy signature.
         if (readInt(source) == 1)
             readString(source);
 
         // Can't use underlying source, which would have been exhausted
-        auto source = StringSource { *tee.source.data };
-        addToStore(info, source, NoRepair, checkSigs, accessor);
+        auto source = StringSource { *saved.s };
+        addToStore(info, source, NoRepair, checkSigs);
 
         res.push_back(info.path);
     }

src/libstore/filetransfer.cc

@@ -22,6 +22,7 @@
 #include <queue>
 #include <random>
 #include <thread>
+#include <regex>
 
 using namespace std::string_literals;
 
@@ -56,7 +57,7 @@ struct curlFileTransfer : public FileTransfer
         Callback<FileTransferResult> callback;
         CURL * req = 0;
         bool active = false; // whether the handle has been added to the multi object
-        std::string status;
+        std::string statusMsg;
 
         unsigned int attempt = 0;
 
@@ -123,7 +124,7 @@ struct curlFileTransfer : public FileTransfer
             if (requestHeaders) curl_slist_free_all(requestHeaders);
             try {
                 if (!done)
-                    fail(FileTransferError(Interrupted, "download of '%s' was interrupted", request.uri));
+                    fail(FileTransferError(Interrupted, nullptr, "download of '%s' was interrupted", request.uri));
             } catch (...) {
                 ignoreException();
             }
@@ -144,6 +145,7 @@ struct curlFileTransfer : public FileTransfer
 
         LambdaSink finalSink;
         std::shared_ptr<CompressionSink> decompressionSink;
+        std::optional<StringSink> errorSink;
 
         std::exception_ptr writeException;
 
@@ -153,9 +155,19 @@ struct curlFileTransfer : public FileTransfer
                 size_t realSize = size * nmemb;
                 result.bodySize += realSize;
 
-                if (!decompressionSink)
+                if (!decompressionSink) {
                     decompressionSink = makeDecompressionSink(encoding, finalSink);
+                    if (! successfulStatuses.count(getHTTPStatus())) {
+                        // In this case we want to construct a TeeSink, to keep
+                        // the response around (which we figure won't be big
+                        // like an actual download should be) to improve error
+                        // messages.
+                        errorSink = StringSink { };
+                    }
+                }
 
+                if (errorSink)
+                    (*errorSink)((unsigned char *) contents, realSize);
                 (*decompressionSink)((unsigned char *) contents, realSize);
 
                 return realSize;
@@ -175,12 +187,13 @@ struct curlFileTransfer : public FileTransfer
             size_t realSize = size * nmemb;
             std::string line((char *) contents, realSize);
             printMsg(lvlVomit, format("got header for '%s': %s") % request.uri % trim(line));
-            if (line.compare(0, 5, "HTTP/") == 0) { // new response starts
+            static std::regex statusLine("HTTP/[^ ]+ +[0-9]+(.*)", std::regex::extended | std::regex::icase);
+            std::smatch match;
+            if (std::regex_match(line, match, statusLine)) {
                 result.etag = "";
-                auto ss = tokenizeString<vector<string>>(line, " ");
-                status = ss.size() >= 2 ? ss[1] : "";
                 result.data = std::make_shared<std::string>();
                 result.bodySize = 0;
+                statusMsg = trim(match[1]);
                 acceptRanges = false;
                 encoding = "";
             } else {
@@ -194,7 +207,9 @@ struct curlFileTransfer : public FileTransfer
                            the expected ETag on a 200 response, then shut
                            down the connection because we already have the
                            data. */
-                        if (result.etag == request.expectedETag && status == "200") {
+                        long httpStatus = 0;
+                        curl_easy_getinfo(req, CURLINFO_RESPONSE_CODE, &httpStatus);
+                        if (result.etag == request.expectedETag && httpStatus == 200) {
                             debug(format("shutting down on 200 HTTP response with expected ETag"));
                             return 0;
                         }
@@ -408,16 +423,21 @@ struct curlFileTransfer : public FileTransfer
 
                 attempt++;
 
+                std::shared_ptr<std::string> response;
+                if (errorSink)
+                    response = errorSink->s;
                 auto exc =
                     code == CURLE_ABORTED_BY_CALLBACK && _isInterrupted
-                    ? FileTransferError(Interrupted, fmt("%s of '%s' was interrupted", request.verb(), request.uri))
+                    ? FileTransferError(Interrupted, response, "%s of '%s' was interrupted", request.verb(), request.uri)
                     : httpStatus != 0
                     ? FileTransferError(err,
-                        fmt("unable to %s '%s': HTTP error %d",
-                            request.verb(), request.uri, httpStatus)
+                        response,
+                        fmt("unable to %s '%s': HTTP error %d ('%s')",
+                            request.verb(), request.uri, httpStatus, statusMsg)
                         + (code == CURLE_OK ? "" : fmt(" (curl error: %s)", curl_easy_strerror(code)))
                         )
                     : FileTransferError(err,
+                        response,
                         fmt("unable to %s '%s': %s (%d)",
                             request.verb(), request.uri, curl_easy_strerror(code), code));
 
@@ -675,7 +695,7 @@ struct curlFileTransfer : public FileTransfer
                 auto s3Res = s3Helper.getObject(bucketName, key);
                 FileTransferResult res;
                 if (!s3Res.data)
-                    throw FileTransferError(NotFound, fmt("S3 object '%s' does not exist", request.uri));
+                    throw FileTransferError(NotFound, nullptr, "S3 object '%s' does not exist", request.uri);
                 res.data = s3Res.data;
                 callback(std::move(res));
 #else
@@ -820,6 +840,21 @@ void FileTransfer::download(FileTransferRequest && request, Sink & sink)
     }
 }
 
+template<typename... Args>
+FileTransferError::FileTransferError(FileTransfer::Error error, std::shared_ptr<string> response, const Args & ... args)
+    : Error(args...), error(error), response(response)
+{
+    const auto hf = hintfmt(args...);
+    // FIXME: Due to https://github.com/NixOS/nix/issues/3841 we don't know how
+    // to print different messages for different verbosity levels. For now
+    // we add some heuristics for detecting when we want to show the response.
+    if (response && (response->size() < 1024 || response->find("<html>") != string::npos)) {
+            err.hint = hintfmt("%1%\n\nresponse body:\n\n%2%", normaltxt(hf.str()), *response);
+    } else {
+        err.hint = hf;
+    }
+}
+
 bool isUri(const string & s)
 {
     if (s.compare(0, 8, "channel:") == 0) return true;

src/libstore/filetransfer.hh

@@ -103,10 +103,12 @@ class FileTransferError : public Error
 {
 public:
     FileTransfer::Error error;
+    std::shared_ptr<string> response; // intentionally optional
+
     template<typename... Args>
-    FileTransferError(FileTransfer::Error error, const Args & ... args)
-        : Error(args...), error(error)
-    { }
+    FileTransferError(FileTransfer::Error error, std::shared_ptr<string> response, const Args & ... args);
+
+    virtual const char* sname() const override { return "FileTransferError"; }
 };
 
 bool isUri(const string & s);

src/libstore/gc.cc

@@ -262,11 +262,13 @@ void LocalStore::findTempRoots(FDs & fds, Roots & tempRoots, bool censor)
 void LocalStore::findRoots(const Path & path, unsigned char type, Roots & roots)
 {
     auto foundRoot = [&](const Path & path, const Path & target) {
-        auto storePath = maybeParseStorePath(toStorePath(target));
-        if (storePath && isValidPath(*storePath))
-            roots[std::move(*storePath)].emplace(path);
-        else
-            printInfo("skipping invalid root from '%1%' to '%2%'", path, target);
+        try {
+            auto storePath = toStorePath(target).first;
+            if (isValidPath(storePath))
+                roots[std::move(storePath)].emplace(path);
+            else
+                printInfo("skipping invalid root from '%1%' to '%2%'", path, target);
+        } catch (BadStorePath &) { }
     };
 
     try {
@@ -472,15 +474,15 @@ void LocalStore::findRuntimeRoots(Roots & roots, bool censor)
 
     for (auto & [target, links] : unchecked) {
         if (!isInStore(target)) continue;
-        Path pathS = toStorePath(target);
-        if (!isStorePath(pathS)) continue;
-        auto path = parseStorePath(pathS);
-        if (!isValidPath(path)) continue;
-        debug("got additional root '%1%'", pathS);
-        if (censor)
-            roots[path].insert(censored);
-        else
-            roots[path].insert(links.begin(), links.end());
+        try {
+            auto path = toStorePath(target).first;
+            if (!isValidPath(path)) continue;
+            debug("got additional root '%1%'", printStorePath(path));
+            if (censor)
+                roots[path].insert(censored);
+            else
+                roots[path].insert(links.begin(), links.end());
+        } catch (BadStorePath &) { }
     }
 }
 
@@ -498,7 +500,7 @@ struct LocalStore::GCState
     StorePathSet alive;
     bool gcKeepOutputs;
     bool gcKeepDerivations;
-    unsigned long long bytesInvalidated;
+    uint64_t bytesInvalidated;
     bool moveToTrash = true;
     bool shouldDelete;
     GCState(const GCOptions & options, GCResults & results)
@@ -516,7 +518,7 @@ bool LocalStore::isActiveTempFile(const GCState & state,
 
 void LocalStore::deleteGarbage(GCState & state, const Path & path)
 {
-    unsigned long long bytesFreed;
+    uint64_t bytesFreed;
     deletePath(path, bytesFreed);
     state.results.bytesFreed += bytesFreed;
 }
@@ -526,7 +528,7 @@ void LocalStore::deletePathRecursive(GCState & state, const Path & path)
 {
     checkInterrupt();
 
-    unsigned long long size = 0;
+    uint64_t size = 0;
 
     auto storePath = maybeParseStorePath(path);
     if (storePath && isValidPath(*storePath)) {
@@ -685,7 +687,7 @@ void LocalStore::removeUnusedLinks(const GCState & state)
     AutoCloseDir dir(opendir(linksDir.c_str()));
     if (!dir) throw SysError("opening directory '%1%'", linksDir);
 
-    long long actualSize = 0, unsharedSize = 0;
+    int64_t actualSize = 0, unsharedSize = 0;
 
     struct dirent * dirent;
     while (errno = 0, dirent = readdir(dir.get())) {
@@ -715,10 +717,10 @@ void LocalStore::removeUnusedLinks(const GCState & state)
     struct stat st;
     if (stat(linksDir.c_str(), &st) == -1)
         throw SysError("statting '%1%'", linksDir);
-    long long overhead = st.st_blocks * 512ULL;
+    auto overhead = st.st_blocks * 512ULL;
 
-    printInfo(format("note: currently hard linking saves %.2f MiB")
-        % ((unsharedSize - actualSize - overhead) / (1024.0 * 1024.0)));
+    printInfo("note: currently hard linking saves %.2f MiB",
+        ((unsharedSize - actualSize - overhead) / (1024.0 * 1024.0)));
 }
 
 

src/libstore/globals.hh

@@ -365,6 +365,9 @@ public:
 
     Setting<size_t> narBufferSize{this, 32 * 1024 * 1024, "nar-buffer-size",
         "Maximum size of NARs before spilling them to disk."};
+
+    Setting<std::string> flakeRegistry{this, "https://github.com/NixOS/flake-registry/raw/master/flake-registry.json", "flake-registry",
+        "Path or URI of the global flake registry."};
 };
 
 

src/libstore/http-binary-cache-store.cc

@@ -100,11 +100,11 @@ protected:
     }
 
     void upsertFile(const std::string & path,
-        const std::string & data,
+        std::shared_ptr<std::basic_iostream<char>> istream,
         const std::string & mimeType) override
     {
         auto req = FileTransferRequest(cacheUri + "/" + path);
-        req.data = std::make_shared<string>(data); // FIXME: inefficient
+        req.data = std::make_shared<string>(StreamToSourceAdapter(istream).drain());
         req.mimeType = mimeType;
         try {
             getFileTransfer()->upload(req);

src/libstore/legacy-ssh-store.cc

@@ -113,7 +113,7 @@ struct LegacySSHStore : public Store
 
             if (GET_PROTOCOL_MINOR(conn->remoteVersion) >= 4) {
                 auto s = readString(conn->from);
-                info->narHash = s.empty() ? Hash() : Hash(s);
+                info->narHash = s.empty() ? std::optional<Hash>{} : Hash{s};
                 info->ca = parseContentAddressOpt(readString(conn->from));
                 info->sigs = readStrings<StringSet>(conn->from);
             }
@@ -126,8 +126,7 @@ struct LegacySSHStore : public Store
     }
 
     void addToStore(const ValidPathInfo & info, Source & source,
-        RepairFlag repair, CheckSigsFlag checkSigs,
-        std::shared_ptr<FSAccessor> accessor) override
+        RepairFlag repair, CheckSigsFlag checkSigs) override
     {
         debug("adding path '%s' to remote host '%s'", printStorePath(info.path), host);
 
@@ -139,7 +138,7 @@ struct LegacySSHStore : public Store
                 << cmdAddToStoreNar
                 << printStorePath(info.path)
                 << (info.deriver ? printStorePath(*info.deriver) : "")
-                << info.narHash.to_string(Base16, false);
+                << info.narHash->to_string(Base16, false);
             writeStorePaths(*this, conn->to, info.references);
             conn->to
                 << info.registrationTime

src/libstore/local-binary-cache-store.cc

@@ -31,8 +31,18 @@ protected:
     bool fileExists(const std::string & path) override;
 
     void upsertFile(const std::string & path,
-        const std::string & data,
-        const std::string & mimeType) override;
+        std::shared_ptr<std::basic_iostream<char>> istream,
+        const std::string & mimeType) override
+    {
+        auto path2 = binaryCacheDir + "/" + path;
+        Path tmp = path2 + ".tmp." + std::to_string(getpid());
+        AutoDelete del(tmp, false);
+        StreamToSourceAdapter source(istream);
+        writeFile(tmp, source);
+        if (rename(tmp.c_str(), path2.c_str()))
+            throw SysError("renaming '%1%' to '%2%'", tmp, path2);
+        del.cancel();
+    }
 
     void getFile(const std::string & path, Sink & sink) override
     {
@@ -52,7 +62,9 @@ protected:
             if (entry.name.size() != 40 ||
                 !hasSuffix(entry.name, ".narinfo"))
                 continue;
-            paths.insert(parseStorePath(storeDir + "/" + entry.name.substr(0, entry.name.size() - 8)));
+            paths.insert(parseStorePath(
+                    storeDir + "/" + entry.name.substr(0, entry.name.size() - 8)
+                    + "-" + MissingName));
         }
 
         return paths;
@@ -68,28 +80,11 @@ void LocalBinaryCacheStore::init()
     BinaryCacheStore::init();
 }
 
-static void atomicWrite(const Path & path, const std::string & s)
-{
-    Path tmp = path + ".tmp." + std::to_string(getpid());
-    AutoDelete del(tmp, false);
-    writeFile(tmp, s);
-    if (rename(tmp.c_str(), path.c_str()))
-        throw SysError("renaming '%1%' to '%2%'", tmp, path);
-    del.cancel();
-}
-
 bool LocalBinaryCacheStore::fileExists(const std::string & path)
 {
     return pathExists(binaryCacheDir + "/" + path);
 }
 
-void LocalBinaryCacheStore::upsertFile(const std::string & path,
-    const std::string & data,
-    const std::string & mimeType)
-{
-    atomicWrite(binaryCacheDir + "/" + path, data);
-}
-
 static RegisterStoreImplementation regStore([](
     const std::string & uri, const Store::Params & params)
     -> std::shared_ptr<Store>

src/libstore/local-fs-store.cc

@@ -20,9 +20,9 @@ struct LocalStoreAccessor : public FSAccessor
 
     Path toRealPath(const Path & path)
     {
-        Path storePath = store->toStorePath(path);
-        if (!store->isValidPath(store->parseStorePath(storePath)))
-            throw InvalidPath("path '%1%' is not a valid store path", storePath);
+        auto storePath = store->toStorePath(path).first;
+        if (!store->isValidPath(storePath))
+            throw InvalidPath("path '%1%' is not a valid store path", store->printStorePath(storePath));
         return store->getRealStoreDir() + std::string(path, store->storeDir.size());
     }
 

src/libstore/local-store.cc

@@ -560,19 +560,12 @@ void LocalStore::checkDerivationOutputs(const StorePath & drvPath, const Derivat
         DerivationOutputs::const_iterator out = drv.outputs.find("out");
         if (out == drv.outputs.end())
             throw Error("derivation '%s' does not have an output named 'out'", printStorePath(drvPath));
-
-        check(
-            makeFixedOutputPath(
-                out->second.hash->method,
-                out->second.hash->hash,
-                drvName),
-            out->second.path, "out");
     }
 
     else {
         Hash h = hashDerivationModulo(*this, drv, true);
         for (auto & i : drv.outputs)
-            check(makeOutputPath(i.first, h, drvName), i.second.path, i.first);
+            check(makeOutputPath(i.first, h, drvName), i.second.path(*this, drv.name), i.first);
     }
 }
 
@@ -586,7 +579,7 @@ uint64_t LocalStore::addValidPath(State & state,
 
     state.stmtRegisterValidPath.use()
         (printStorePath(info.path))
-        (info.narHash.to_string(Base16, true))
+        (info.narHash->to_string(Base16, true))
         (info.registrationTime == 0 ? time(0) : info.registrationTime)
         (info.deriver ? printStorePath(*info.deriver) : "", (bool) info.deriver)
         (info.narSize, info.narSize != 0)
@@ -594,7 +587,7 @@ uint64_t LocalStore::addValidPath(State & state,
         (concatStringsSep(" ", info.sigs), !info.sigs.empty())
         (renderContentAddress(info.ca), (bool) info.ca)
         .exec();
-    uint64_t id = sqlite3_last_insert_rowid(state.db);
+    uint64_t id = state.db.getLastInsertedRowId();
 
     /* If this is a derivation, then store the derivation outputs in
        the database.  This is useful for the garbage collector: it can
@@ -614,7 +607,7 @@ uint64_t LocalStore::addValidPath(State & state,
             state.stmtAddDerivationOutput.use()
                 (id)
                 (i.first)
-                (printStorePath(i.second.path))
+                (printStorePath(i.second.path(*this, drv.name)))
                 .exec();
         }
     }
@@ -686,7 +679,7 @@ void LocalStore::updatePathInfo(State & state, const ValidPathInfo & info)
 {
     state.stmtUpdatePathInfo.use()
         (info.narSize, info.narSize != 0)
-        (info.narHash.to_string(Base16, true))
+        (info.narHash->to_string(Base16, true))
         (info.ultimate ? 1 : 0, info.ultimate)
         (concatStringsSep(" ", info.sigs), !info.sigs.empty())
         (renderContentAddress(info.ca), (bool) info.ca)
@@ -912,7 +905,7 @@ void LocalStore::registerValidPaths(const ValidPathInfos & infos)
         StorePathSet paths;
 
         for (auto & i : infos) {
-            assert(i.narHash.type == htSHA256);
+            assert(i.narHash && i.narHash->type == htSHA256);
             if (isValidPath_(*state, i.path))
                 updatePathInfo(*state, i);
             else
@@ -974,7 +967,7 @@ const PublicKeys & LocalStore::getPublicKeys()
 
 
 void LocalStore::addToStore(const ValidPathInfo & info, Source & source,
-    RepairFlag repair, CheckSigsFlag checkSigs, std::shared_ptr<FSAccessor> accessor)
+    RepairFlag repair, CheckSigsFlag checkSigs)
 {
     if (!info.narHash)
         throw Error("cannot add path '%s' because it lacks a hash", printStorePath(info.path));
@@ -1025,7 +1018,7 @@ void LocalStore::addToStore(const ValidPathInfo & info, Source & source,
 
             if (hashResult.first != info.narHash)
                 throw Error("hash mismatch importing path '%s';\n  wanted: %s\n  got:    %s",
-                    printStorePath(info.path), info.narHash.to_string(Base32, true), hashResult.first.to_string(Base32, true));
+                    printStorePath(info.path), info.narHash->to_string(Base32, true), hashResult.first.to_string(Base32, true));
 
             if (hashResult.second != info.narSize)
                 throw Error("size mismatch importing path '%s';\n  wanted: %s\n  got:   %s",
@@ -1045,79 +1038,26 @@ void LocalStore::addToStore(const ValidPathInfo & info, Source & source,
 }
 
 
-StorePath LocalStore::addToStoreFromDump(const string & dump, const string & name,
-    FileIngestionMethod method, HashType hashAlgo, RepairFlag repair)
-{
-    Hash h = hashString(hashAlgo, dump);
-
-    auto dstPath = makeFixedOutputPath(method, h, name);
-
-    addTempRoot(dstPath);
-
-    if (repair || !isValidPath(dstPath)) {
-
-        /* The first check above is an optimisation to prevent
-           unnecessary lock acquisition. */
-
-        auto realPath = Store::toRealPath(dstPath);
-
-        PathLocks outputLock({realPath});
-
-        if (repair || !isValidPath(dstPath)) {
-
-            deletePath(realPath);
-
-            autoGC();
-
-            if (method == FileIngestionMethod::Recursive) {
-                StringSource source(dump);
-                restorePath(realPath, source);
-            } else
-                writeFile(realPath, dump);
-
-            canonicalisePathMetaData(realPath, -1);
-
-            /* Register the SHA-256 hash of the NAR serialisation of
-               the path in the database.  We may just have computed it
-               above (if called with recursive == true and hashAlgo ==
-               sha256); otherwise, compute it here. */
-            HashResult hash;
-            if (method == FileIngestionMethod::Recursive) {
-                hash.first = hashAlgo == htSHA256 ? h : hashString(htSHA256, dump);
-                hash.second = dump.size();
-            } else
-                hash = hashPath(htSHA256, realPath);
-
-            optimisePath(realPath); // FIXME: combine with hashPath()
-
-            ValidPathInfo info(dstPath);
-            info.narHash = hash.first;
-            info.narSize = hash.second;
-            info.ca = FixedOutputHash { .method = method, .hash = h };
-            registerValidPath(info);
-        }
-
-        outputLock.setDeletion(true);
-    }
-
-    return dstPath;
-}
-
-
 StorePath LocalStore::addToStore(const string & name, const Path & _srcPath,
     FileIngestionMethod method, HashType hashAlgo, PathFilter & filter, RepairFlag repair)
 {
     Path srcPath(absPath(_srcPath));
+    auto source = sinkToSource([&](Sink & sink) {
+        if (method == FileIngestionMethod::Recursive)
+            dumpPath(srcPath, sink, filter);
+        else
+            readFile(srcPath, sink);
+    });
+    return addToStoreFromDump(*source, name, method, hashAlgo, repair);
+}
 
-    /* For computing the NAR hash. */
-    auto sha256Sink = std::make_unique<HashSink>(htSHA256);
 
-    /* For computing the store path. In recursive SHA-256 mode, this
-       is the same as the NAR hash, so no need to do it again. */
-    std::unique_ptr<HashSink> hashSink =
-        method == FileIngestionMethod::Recursive && hashAlgo == htSHA256
-        ? nullptr
-        : std::make_unique<HashSink>(hashAlgo);
+StorePath LocalStore::addToStoreFromDump(Source & source0, const string & name,
+    FileIngestionMethod method, HashType hashAlgo, RepairFlag repair)
+{
+    /* For computing the store path. */
+    auto hashSink = std::make_unique<HashSink>(hashAlgo);
+    TeeSource source { source0, *hashSink };
 
     /* Read the source path into memory, but only if it's up to
        narBufferSize bytes. If it's larger, write it to a temporary
@@ -1125,61 +1065,49 @@ StorePath LocalStore::addToStore(const string & name, const Path & _srcPath,
        destination store path is already valid, we just delete the
        temporary path. Otherwise, we move it to the destination store
        path. */
-    bool inMemory = true;
-    std::string nar;
+    bool inMemory = false;
 
-    auto source = sinkToSource([&](Sink & sink) {
+    std::string dump;
 
-        LambdaSink sink2([&](const unsigned char * buf, size_t len) {
-            (*sha256Sink)(buf, len);
-            if (hashSink) (*hashSink)(buf, len);
-
-            if (inMemory) {
-                if (nar.size() + len > settings.narBufferSize) {
-                    inMemory = false;
-                    sink << 1;
-                    sink((const unsigned char *) nar.data(), nar.size());
-                    nar.clear();
-                } else {
-                    nar.append((const char *) buf, len);
-                }
-            }
-
-            if (!inMemory) sink(buf, len);
-        });
-
-        if (method == FileIngestionMethod::Recursive)
-            dumpPath(srcPath, sink2, filter);
-        else
-            readFile(srcPath, sink2);
-    });
+    /* Fill out buffer, and decide whether we are working strictly in
+       memory based on whether we break out because the buffer is full
+       or the original source is empty */
+    while (dump.size() < settings.narBufferSize) {
+        auto oldSize = dump.size();
+        constexpr size_t chunkSize = 65536;
+        auto want = std::min(chunkSize, settings.narBufferSize - oldSize);
+        dump.resize(oldSize + want);
+        auto got = 0;
+        try {
+            got = source.read((uint8_t *) dump.data() + oldSize, want);
+        } catch (EndOfFile &) {
+            inMemory = true;
+            break;
+        }
+        dump.resize(oldSize + got);
+    }
 
     std::unique_ptr<AutoDelete> delTempDir;
     Path tempPath;
 
-    try {
-        /* Wait for the source coroutine to give us some dummy
-           data. This is so that we don't create the temporary
-           directory if the NAR fits in memory. */
-        readInt(*source);
+    if (!inMemory) {
+        /* Drain what we pulled so far, and then keep on pulling */
+        StringSource dumpSource { dump };
+        ChainSource bothSource { dumpSource, source };
 
         auto tempDir = createTempDir(realStoreDir, "add");
         delTempDir = std::make_unique<AutoDelete>(tempDir);
         tempPath = tempDir + "/x";
 
         if (method == FileIngestionMethod::Recursive)
-            restorePath(tempPath, *source);
+            restorePath(tempPath, bothSource);
         else
-            writeFile(tempPath, *source);
+            writeFile(tempPath, bothSource);
 
-    } catch (EndOfFile &) {
-        if (!inMemory) throw;
-        /* The NAR fits in memory, so we didn't do restorePath(). */
+        dump.clear();
     }
 
-    auto sha256 = sha256Sink->finish();
-
-    Hash hash = hashSink ? hashSink->finish().first : sha256.first;
+    auto [hash, size] = hashSink->finish();
 
     auto dstPath = makeFixedOutputPath(method, hash, name);
 
@@ -1201,25 +1129,34 @@ StorePath LocalStore::addToStore(const string & name, const Path & _srcPath,
             autoGC();
 
             if (inMemory) {
+                 StringSource dumpSource { dump };
                 /* Restore from the NAR in memory. */
-                StringSource source(nar);
                 if (method == FileIngestionMethod::Recursive)
-                    restorePath(realPath, source);
+                    restorePath(realPath, dumpSource);
                 else
-                    writeFile(realPath, source);
+                    writeFile(realPath, dumpSource);
             } else {
                 /* Move the temporary path we restored above. */
                 if (rename(tempPath.c_str(), realPath.c_str()))
                     throw Error("renaming '%s' to '%s'", tempPath, realPath);
             }
 
+            /* For computing the nar hash. In recursive SHA-256 mode, this
+               is the same as the store hash, so no need to do it again. */
+            auto narHash = std::pair { hash, size };
+            if (method != FileIngestionMethod::Recursive || hashAlgo != htSHA256) {
+                HashSink narSink { htSHA256 };
+                dumpPath(realPath, narSink);
+                narHash = narSink.finish();
+            }
+
             canonicalisePathMetaData(realPath, -1); // FIXME: merge into restorePath
 
             optimisePath(realPath);
 
             ValidPathInfo info(dstPath);
-            info.narHash = sha256.first;
-            info.narSize = sha256.second;
+            info.narHash = narHash.first;
+            info.narSize = narHash.second;
             info.ca = FixedOutputHash { .method = method, .hash = hash };
             registerValidPath(info);
         }
@@ -1377,9 +1314,9 @@ bool LocalStore::verifyStore(bool checkContents, RepairFlag repair)
 
                 std::unique_ptr<AbstractHashSink> hashSink;
                 if (!info->ca || !info->references.count(info->path))
-                    hashSink = std::make_unique<HashSink>(*info->narHash.type);
+                    hashSink = std::make_unique<HashSink>(info->narHash->type);
                 else
-                    hashSink = std::make_unique<HashModuloSink>(*info->narHash.type, std::string(info->path.hashPart()));
+                    hashSink = std::make_unique<HashModuloSink>(info->narHash->type, std::string(info->path.hashPart()));
 
                 dumpPath(Store::toRealPath(i), *hashSink);
                 auto current = hashSink->finish();
@@ -1388,7 +1325,7 @@ bool LocalStore::verifyStore(bool checkContents, RepairFlag repair)
                     logError({
                         .name = "Invalid hash - path modified",
                         .hint = hintfmt("path '%s' was modified! expected hash '%s', got '%s'",
-                        printStorePath(i), info->narHash.to_string(Base32, true), current.first.to_string(Base32, true))
+                        printStorePath(i), info->narHash->to_string(Base32, true), current.first.to_string(Base32, true))
                     });
                     if (repair) repairPath(i); else errors = true;
                 } else {

src/libstore/local-store.hh

@@ -29,8 +29,8 @@ struct Derivation;
 struct OptimiseStats
 {
     unsigned long filesLinked = 0;
-    unsigned long long bytesFreed = 0;
-    unsigned long long blocksFreed = 0;
+    uint64_t bytesFreed = 0;
+    uint64_t blocksFreed = 0;
 };
 
 
@@ -143,8 +143,7 @@ public:
         SubstitutablePathInfos & infos) override;
 
     void addToStore(const ValidPathInfo & info, Source & source,
-        RepairFlag repair, CheckSigsFlag checkSigs,
-        std::shared_ptr<FSAccessor> accessor) override;
+        RepairFlag repair, CheckSigsFlag checkSigs) override;
 
     StorePath addToStore(const string & name, const Path & srcPath,
         FileIngestionMethod method, HashType hashAlgo,
@@ -154,7 +153,7 @@ public:
        in `dump', which is either a NAR serialisation (if recursive ==
        true) or simply the contents of a regular file (if recursive ==
        false). */
-    StorePath addToStoreFromDump(const string & dump, const string & name,
+    StorePath addToStoreFromDump(Source & dump, const string & name,
         FileIngestionMethod method = FileIngestionMethod::Recursive, HashType hashAlgo = htSHA256, RepairFlag repair = NoRepair) override;
 
     StorePath addTextToStore(const string & name, const string & s,

src/libstore/local.mk

@@ -61,3 +61,6 @@ $(d)/build.cc:
 clean-files += $(d)/schema.sql.gen.hh
 
 $(eval $(call install-file-in, $(d)/nix-store.pc, $(prefix)/lib/pkgconfig, 0644))
+
+$(foreach i, $(wildcard src/libstore/builtins/*.hh), \
+  $(eval $(call install-file-in, $(i), $(includedir)/nix/builtins, 0644)))

src/libstore/misc.cc

@@ -111,14 +111,16 @@ void Store::computeFSClosure(const StorePath & startPath,
 std::optional<ContentAddress> getDerivationCA(const BasicDerivation & drv)
 {
     auto out = drv.outputs.find("out");
-    if (out != drv.outputs.end())
-        return out->second.hash;
+    if (out != drv.outputs.end()) {
+        if (auto v = std::get_if<DerivationOutputFixed>(&out->second.output))
+            return v->hash;
+    }
     return std::nullopt;
 }
 
 void Store::queryMissing(const std::vector<StorePathWithOutputs> & targets,
     StorePathSet & willBuild_, StorePathSet & willSubstitute_, StorePathSet & unknown_,
-    unsigned long long & downloadSize_, unsigned long long & narSize_)
+    uint64_t & downloadSize_, uint64_t & narSize_)
 {
     Activity act(*logger, lvlDebug, actUnknown, "querying info about missing paths");
 
@@ -130,8 +132,8 @@ void Store::queryMissing(const std::vector<StorePathWithOutputs> & targets,
     {
         std::unordered_set<std::string> done;
         StorePathSet & unknown, & willSubstitute, & willBuild;
-        unsigned long long & downloadSize;
-        unsigned long long & narSize;
+        uint64_t & downloadSize;
+        uint64_t & narSize;
     };
 
     struct DrvState
@@ -206,8 +208,8 @@ void Store::queryMissing(const std::vector<StorePathWithOutputs> & targets,
             PathSet invalid;
             for (auto & j : drv->outputs)
                 if (wantOutput(j.first, path.outputs)
-                    && !isValidPath(j.second.path))
-                    invalid.insert(printStorePath(j.second.path));
+                    && !isValidPath(j.second.path(*this, drv->name)))
+                    invalid.insert(printStorePath(j.second.path(*this, drv->name)));
             if (invalid.empty()) return;
 
             if (settings.useSubstitutes && parsedDrv.substitutesAllowed()) {

src/libstore/nar-accessor.cc

@@ -18,7 +18,7 @@ struct NarMember
 
     /* If this is a regular file, position of the contents of this
        file in the NAR. */
-    size_t start = 0, size = 0;
+    uint64_t start = 0, size = 0;
 
     std::string target;
 
@@ -34,17 +34,19 @@ struct NarAccessor : public FSAccessor
 
     NarMember root;
 
-    struct NarIndexer : ParseSink, StringSource
+    struct NarIndexer : ParseSink, Source
     {
         NarAccessor & acc;
+        Source & source;
 
         std::stack<NarMember *> parents;
 
-        std::string currentStart;
         bool isExec = false;
 
-        NarIndexer(NarAccessor & acc, const std::string & nar)
-            : StringSource(nar), acc(acc)
+        uint64_t pos = 0;
+
+        NarIndexer(NarAccessor & acc, Source & source)
+            : acc(acc), source(source)
         { }
 
         void createMember(const Path & path, NarMember member) {
@@ -77,33 +79,40 @@ struct NarAccessor : public FSAccessor
             parents.top()->isExecutable = true;
         }
 
-        void preallocateContents(unsigned long long size) override
+        void preallocateContents(uint64_t size) override
         {
-            currentStart = string(s, pos, 16);
-            assert(size <= std::numeric_limits<size_t>::max());
-            parents.top()->size = (size_t)size;
+            assert(size <= std::numeric_limits<uint64_t>::max());
+            parents.top()->size = (uint64_t) size;
             parents.top()->start = pos;
         }
 
-        void receiveContents(unsigned char * data, unsigned int len) override
-        {
-            // Sanity check
-            if (!currentStart.empty()) {
-                assert(len < 16 || currentStart == string((char *) data, 16));
-                currentStart.clear();
-            }
-        }
+        void receiveContents(unsigned char * data, size_t len) override
+        { }
 
         void createSymlink(const Path & path, const string & target) override
         {
             createMember(path,
                 NarMember{FSAccessor::Type::tSymlink, false, 0, 0, target});
         }
+
+        size_t read(unsigned char * data, size_t len) override
+        {
+            auto n = source.read(data, len);
+            pos += n;
+            return n;
+        }
     };
 
     NarAccessor(ref<const std::string> nar) : nar(nar)
     {
-        NarIndexer indexer(*this, *nar);
+        StringSource source(*nar);
+        NarIndexer indexer(*this, source);
+        parseDump(indexer, indexer);
+    }
+
+    NarAccessor(Source & source)
+    {
+        NarIndexer indexer(*this, source);
         parseDump(indexer, indexer);
     }
 
@@ -219,6 +228,11 @@ ref<FSAccessor> makeNarAccessor(ref<const std::string> nar)
     return make_ref<NarAccessor>(nar);
 }
 
+ref<FSAccessor> makeNarAccessor(Source & source)
+{
+    return make_ref<NarAccessor>(source);
+}
+
 ref<FSAccessor> makeLazyNarAccessor(const std::string & listing,
     GetNarBytes getNarBytes)
 {

src/libstore/nar-accessor.hh

@@ -6,10 +6,14 @@
 
 namespace nix {
 
+struct Source;
+
 /* Return an object that provides access to the contents of a NAR
    file. */
 ref<FSAccessor> makeNarAccessor(ref<const std::string> nar);
 
+ref<FSAccessor> makeNarAccessor(Source & source);
+
 /* Create a NAR accessor from a NAR listing (in the format produced by
    listNar()). The callback getNarBytes(offset, length) is used by the
    readFile() method of the accessor to get the contents of files

src/libstore/nar-info-disk-cache.cc

@@ -230,9 +230,9 @@ public:
                     (std::string(info->path.name()))
                     (narInfo ? narInfo->url : "", narInfo != 0)
                     (narInfo ? narInfo->compression : "", narInfo != 0)
-                    (narInfo && narInfo->fileHash ? narInfo->fileHash.to_string(Base32, true) : "", narInfo && narInfo->fileHash)
+                    (narInfo && narInfo->fileHash ? narInfo->fileHash->to_string(Base32, true) : "", narInfo && narInfo->fileHash)
                     (narInfo ? narInfo->fileSize : 0, narInfo != 0 && narInfo->fileSize)
-                    (info->narHash.to_string(Base32, true))
+                    (info->narHash->to_string(Base32, true))
                     (info->narSize)
                     (concatStringsSep(" ", info->shortRefs()))
                     (info->deriver ? std::string(info->deriver->to_string()) : "", (bool) info->deriver)

src/libstore/nar-info.cc

@@ -7,15 +7,14 @@ NarInfo::NarInfo(const Store & store, const std::string & s, const std::string &
     : ValidPathInfo(StorePath(StorePath::dummy)) // FIXME: hack
 {
     auto corrupt = [&]() {
-        throw Error("NAR info file '%1%' is corrupt", whence);
+        return Error("NAR info file '%1%' is corrupt", whence);
     };
 
     auto parseHashField = [&](const string & s) {
         try {
             return Hash(s);
         } catch (BadHash &) {
-            corrupt();
-            return Hash(); // never reached
+            throw corrupt();
         }
     };
 
@@ -25,12 +24,12 @@ NarInfo::NarInfo(const Store & store, const std::string & s, const std::string &
     while (pos < s.size()) {
 
         size_t colon = s.find(':', pos);
-        if (colon == std::string::npos) corrupt();
+        if (colon == std::string::npos) throw corrupt();
 
         std::string name(s, pos, colon - pos);
 
         size_t eol = s.find('\n', colon + 2);
-        if (eol == std::string::npos) corrupt();
+        if (eol == std::string::npos) throw corrupt();
 
         std::string value(s, colon + 2, eol - colon - 2);
 
@@ -45,16 +44,16 @@ NarInfo::NarInfo(const Store & store, const std::string & s, const std::string &
         else if (name == "FileHash")
             fileHash = parseHashField(value);
         else if (name == "FileSize") {
-            if (!string2Int(value, fileSize)) corrupt();
+            if (!string2Int(value, fileSize)) throw corrupt();
         }
         else if (name == "NarHash")
             narHash = parseHashField(value);
         else if (name == "NarSize") {
-            if (!string2Int(value, narSize)) corrupt();
+            if (!string2Int(value, narSize)) throw corrupt();
         }
         else if (name == "References") {
             auto refs = tokenizeString<Strings>(value, " ");
-            if (!references.empty()) corrupt();
+            if (!references.empty()) throw corrupt();
             for (auto & r : refs)
                 references.insert(StorePath(r));
         }
@@ -67,7 +66,7 @@ NarInfo::NarInfo(const Store & store, const std::string & s, const std::string &
         else if (name == "Sig")
             sigs.insert(value);
         else if (name == "CA") {
-            if (ca) corrupt();
+            if (ca) throw corrupt();
             // FIXME: allow blank ca or require skipping field?
             ca = parseContentAddressOpt(value);
         }
@@ -77,7 +76,7 @@ NarInfo::NarInfo(const Store & store, const std::string & s, const std::string &
 
     if (compression == "") compression = "bzip2";
 
-    if (!havePath || url.empty() || narSize == 0 || !narHash) corrupt();
+    if (!havePath || url.empty() || narSize == 0 || !narHash) throw corrupt();
 }
 
 std::string NarInfo::to_string(const Store & store) const
@@ -87,11 +86,11 @@ std::string NarInfo::to_string(const Store & store) const
     res += "URL: " + url + "\n";
     assert(compression != "");
     res += "Compression: " + compression + "\n";
-    assert(fileHash.type == htSHA256);
-    res += "FileHash: " + fileHash.to_string(Base32, true) + "\n";
+    assert(fileHash && fileHash->type == htSHA256);
+    res += "FileHash: " + fileHash->to_string(Base32, true) + "\n";
     res += "FileSize: " + std::to_string(fileSize) + "\n";
-    assert(narHash.type == htSHA256);
-    res += "NarHash: " + narHash.to_string(Base32, true) + "\n";
+    assert(narHash && narHash->type == htSHA256);
+    res += "NarHash: " + narHash->to_string(Base32, true) + "\n";
     res += "NarSize: " + std::to_string(narSize) + "\n";
 
     res += "References: " + concatStringsSep(" ", shortRefs()) + "\n";

src/libstore/nar-info.hh

@@ -10,7 +10,7 @@ struct NarInfo : ValidPathInfo
 {
     std::string url;
     std::string compression;
-    Hash fileHash;
+    std::optional<Hash> fileHash;
     uint64_t fileSize = 0;
     std::string system;
 

src/libstore/optimise-store.cc

@@ -282,7 +282,7 @@ void LocalStore::optimiseStore(OptimiseStats & stats)
     }
 }
 
-static string showBytes(unsigned long long bytes)
+static string showBytes(uint64_t bytes)
 {
     return (format("%.2f MiB") % (bytes / (1024.0 * 1024.0))).str();
 }

src/libstore/path.cc

@@ -2,8 +2,6 @@
 
 namespace nix {
 
-MakeError(BadStorePath, Error);
-
 static void checkName(std::string_view path, std::string_view name)
 {
     if (name.empty())

src/libstore/profiles.cc

@@ -12,30 +12,24 @@
 namespace nix {
 
 
-static bool cmpGensByNumber(const Generation & a, const Generation & b)
-{
-    return a.number < b.number;
-}
-
-
 /* Parse a generation name of the format
    `<profilename>-<number>-link'. */
-static int parseName(const string & profileName, const string & name)
+static std::optional<GenerationNumber> parseName(const string & profileName, const string & name)
 {
-    if (string(name, 0, profileName.size() + 1) != profileName + "-") return -1;
+    if (string(name, 0, profileName.size() + 1) != profileName + "-") return {};
     string s = string(name, profileName.size() + 1);
     string::size_type p = s.find("-link");
-    if (p == string::npos) return -1;
-    int n;
+    if (p == string::npos) return {};
+    unsigned int n;
     if (string2Int(string(s, 0, p), n) && n >= 0)
         return n;
     else
-        return -1;
+        return {};
 }
 
 
 
-Generations findGenerations(Path profile, int & curGen)
+std::pair<Generations, std::optional<GenerationNumber>> findGenerations(Path profile)
 {
     Generations gens;
 
@@ -43,30 +37,34 @@ Generations findGenerations(Path profile, int & curGen)
     auto profileName = std::string(baseNameOf(profile));
 
     for (auto & i : readDirectory(profileDir)) {
-        int n;
-        if ((n = parseName(profileName, i.name)) != -1) {
-            Generation gen;
-            gen.path = profileDir + "/" + i.name;
-            gen.number = n;
+        if (auto n = parseName(profileName, i.name)) {
+            auto path = profileDir + "/" + i.name;
             struct stat st;
-            if (lstat(gen.path.c_str(), &st) != 0)
-                throw SysError("statting '%1%'", gen.path);
-            gen.creationTime = st.st_mtime;
-            gens.push_back(gen);
+            if (lstat(path.c_str(), &st) != 0)
+                throw SysError("statting '%1%'", path);
+            gens.push_back({
+                .number = *n,
+                .path = path,
+                .creationTime = st.st_mtime
+            });
         }
     }
 
-    gens.sort(cmpGensByNumber);
+    gens.sort([](const Generation & a, const Generation & b)
+    {
+        return a.number < b.number;
+    });
 
-    curGen = pathExists(profile)
+    return {
+        gens,
+        pathExists(profile)
         ? parseName(profileName, readLink(profile))
-        : -1;
-
-    return gens;
+        : std::nullopt
+    };
 }
 
 
-static void makeName(const Path & profile, unsigned int num,
+static void makeName(const Path & profile, GenerationNumber num,
     Path & outLink)
 {
     Path prefix = (format("%1%-%2%") % profile % num).str();
@@ -78,10 +76,9 @@ Path createGeneration(ref<LocalFSStore> store, Path profile, Path outPath)
 {
     /* The new generation number should be higher than old the
        previous ones. */
-    int dummy;
-    Generations gens = findGenerations(profile, dummy);
+    auto [gens, dummy] = findGenerations(profile);
 
-    unsigned int num;
+    GenerationNumber num;
     if (gens.size() > 0) {
         Generation last = gens.back();
 
@@ -121,7 +118,7 @@ static void removeFile(const Path & path)
 }
 
 
-void deleteGeneration(const Path & profile, unsigned int gen)
+void deleteGeneration(const Path & profile, GenerationNumber gen)
 {
     Path generation;
     makeName(profile, gen, generation);
@@ -129,7 +126,7 @@ void deleteGeneration(const Path & profile, unsigned int gen)
 }
 
 
-static void deleteGeneration2(const Path & profile, unsigned int gen, bool dryRun)
+static void deleteGeneration2(const Path & profile, GenerationNumber gen, bool dryRun)
 {
     if (dryRun)
         printInfo(format("would remove generation %1%") % gen);
@@ -140,31 +137,29 @@ static void deleteGeneration2(const Path & profile, unsigned int gen, bool dryRu
 }
 
 
-void deleteGenerations(const Path & profile, const std::set<unsigned int> & gensToDelete, bool dryRun)
+void deleteGenerations(const Path & profile, const std::set<GenerationNumber> & gensToDelete, bool dryRun)
 {
     PathLocks lock;
     lockProfile(lock, profile);
 
-    int curGen;
-    Generations gens = findGenerations(profile, curGen);
+    auto [gens, curGen] = findGenerations(profile);
 
-    if (gensToDelete.find(curGen) != gensToDelete.end())
+    if (gensToDelete.count(*curGen))
         throw Error("cannot delete current generation of profile %1%'", profile);
 
     for (auto & i : gens) {
-        if (gensToDelete.find(i.number) == gensToDelete.end()) continue;
+        if (!gensToDelete.count(i.number)) continue;
         deleteGeneration2(profile, i.number, dryRun);
     }
 }
 
-void deleteGenerationsGreaterThan(const Path & profile, int max, bool dryRun)
+void deleteGenerationsGreaterThan(const Path & profile, GenerationNumber max, bool dryRun)
 {
     PathLocks lock;
     lockProfile(lock, profile);
 
-    int curGen;
     bool fromCurGen = false;
-    Generations gens = findGenerations(profile, curGen);
+    auto [gens, curGen] = findGenerations(profile);
     for (auto i = gens.rbegin(); i != gens.rend(); ++i) {
         if (i->number == curGen) {
             fromCurGen = true;
@@ -186,8 +181,7 @@ void deleteOldGenerations(const Path & profile, bool dryRun)
     PathLocks lock;
     lockProfile(lock, profile);
 
-    int curGen;
-    Generations gens = findGenerations(profile, curGen);
+    auto [gens, curGen] = findGenerations(profile);
 
     for (auto & i : gens)
         if (i.number != curGen)
@@ -200,8 +194,7 @@ void deleteGenerationsOlderThan(const Path & profile, time_t t, bool dryRun)
     PathLocks lock;
     lockProfile(lock, profile);
 
-    int curGen;
-    Generations gens = findGenerations(profile, curGen);
+    auto [gens, curGen] = findGenerations(profile);
 
     bool canDelete = false;
     for (auto i = gens.rbegin(); i != gens.rend(); ++i)

src/libstore/profiles.hh

@@ -9,37 +9,32 @@
 namespace nix {
 
 
+typedef unsigned int GenerationNumber;
+
 struct Generation
 {
-    int number;
+    GenerationNumber number;
     Path path;
     time_t creationTime;
-    Generation()
-    {
-        number = -1;
-    }
-    operator bool() const
-    {
-        return number != -1;
-    }
 };
 
-typedef list<Generation> Generations;
+typedef std::list<Generation> Generations;
 
 
 /* Returns the list of currently present generations for the specified
-   profile, sorted by generation number. */
-Generations findGenerations(Path profile, int & curGen);
+   profile, sorted by generation number. Also returns the number of
+   the current generation. */
+std::pair<Generations, std::optional<GenerationNumber>> findGenerations(Path profile);
 
 class LocalFSStore;
 
 Path createGeneration(ref<LocalFSStore> store, Path profile, Path outPath);
 
-void deleteGeneration(const Path & profile, unsigned int gen);
+void deleteGeneration(const Path & profile, GenerationNumber gen);
 
-void deleteGenerations(const Path & profile, const std::set<unsigned int> & gensToDelete, bool dryRun);
+void deleteGenerations(const Path & profile, const std::set<GenerationNumber> & gensToDelete, bool dryRun);
 
-void deleteGenerationsGreaterThan(const Path & profile, const int max, bool dryRun);
+void deleteGenerationsGreaterThan(const Path & profile, GenerationNumber max, bool dryRun);
 
 void deleteOldGenerations(const Path & profile, bool dryRun);
 

src/libstore/references.cc

@@ -48,13 +48,12 @@ static void search(const unsigned char * s, size_t len,
 
 struct RefScanSink : Sink
 {
-    HashSink hashSink;
     StringSet hashes;
     StringSet seen;
 
     string tail;
 
-    RefScanSink() : hashSink(htSHA256) { }
+    RefScanSink() { }
 
     void operator () (const unsigned char * data, size_t len);
 };
@@ -62,8 +61,6 @@ struct RefScanSink : Sink
 
 void RefScanSink::operator () (const unsigned char * data, size_t len)
 {
-    hashSink(data, len);
-
     /* It's possible that a reference spans the previous and current
        fragment, so search in the concatenation of the tail of the
        previous fragment and the start of the current fragment. */
@@ -79,10 +76,12 @@ void RefScanSink::operator () (const unsigned char * data, size_t len)
 }
 
 
-PathSet scanForReferences(const string & path,
-    const PathSet & refs, HashResult & hash)
+std::pair<PathSet, HashResult> scanForReferences(const string & path,
+    const PathSet & refs)
 {
-    RefScanSink sink;
+    RefScanSink refsSink;
+    HashSink hashSink { htSHA256 };
+    TeeSink sink { refsSink, hashSink };
     std::map<string, Path> backMap;
 
     /* For efficiency (and a higher hit rate), just search for the
@@ -97,7 +96,7 @@ PathSet scanForReferences(const string & path,
         assert(s.size() == refLength);
         assert(backMap.find(s) == backMap.end());
         // parseHash(htSHA256, s);
-        sink.hashes.insert(s);
+        refsSink.hashes.insert(s);
         backMap[s] = i;
     }
 
@@ -106,15 +105,15 @@ PathSet scanForReferences(const string & path,
 
     /* Map the hashes found back to their store paths. */
     PathSet found;
-    for (auto & i : sink.seen) {
+    for (auto & i : refsSink.seen) {
         std::map<string, Path>::iterator j;
         if ((j = backMap.find(i)) == backMap.end()) abort();
         found.insert(j->second);
     }
 
-    hash = sink.hashSink.finish();
+    auto hash = hashSink.finish();
 
-    return found;
+    return std::pair<PathSet, HashResult>(found, hash);
 }
 
 

src/libstore/references.hh

@@ -5,8 +5,7 @@
 
 namespace nix {
 
-PathSet scanForReferences(const Path & path, const PathSet & refs,
-    HashResult & hash);
+std::pair<PathSet, HashResult> scanForReferences(const Path & path, const PathSet & refs);
 
 struct RewritingSink : Sink
 {

src/libstore/remote-fs-accessor.cc

@@ -16,26 +16,26 @@ RemoteFSAccessor::RemoteFSAccessor(ref<Store> store, const Path & cacheDir)
         createDirs(cacheDir);
 }
 
-Path RemoteFSAccessor::makeCacheFile(const Path & storePath, const std::string & ext)
+Path RemoteFSAccessor::makeCacheFile(std::string_view hashPart, const std::string & ext)
 {
     assert(cacheDir != "");
-    return fmt("%s/%s.%s", cacheDir, store->parseStorePath(storePath).hashPart(), ext);
+    return fmt("%s/%s.%s", cacheDir, hashPart, ext);
 }
 
-void RemoteFSAccessor::addToCache(const Path & storePath, const std::string & nar,
+void RemoteFSAccessor::addToCache(std::string_view hashPart, const std::string & nar,
     ref<FSAccessor> narAccessor)
 {
-    nars.emplace(storePath, narAccessor);
+    nars.emplace(hashPart, narAccessor);
 
     if (cacheDir != "") {
         try {
             std::ostringstream str;
             JSONPlaceholder jsonRoot(str);
             listNar(jsonRoot, narAccessor, "", true);
-            writeFile(makeCacheFile(storePath, "ls"), str.str());
+            writeFile(makeCacheFile(hashPart, "ls"), str.str());
 
             /* FIXME: do this asynchronously. */
-            writeFile(makeCacheFile(storePath, "nar"), nar);
+            writeFile(makeCacheFile(hashPart, "nar"), nar);
 
         } catch (...) {
             ignoreException();
@@ -47,23 +47,22 @@ std::pair<ref<FSAccessor>, Path> RemoteFSAccessor::fetch(const Path & path_)
 {
     auto path = canonPath(path_);
 
-    auto storePath = store->toStorePath(path);
-    std::string restPath = std::string(path, storePath.size());
+    auto [storePath, restPath] = store->toStorePath(path);
 
-    if (!store->isValidPath(store->parseStorePath(storePath)))
-        throw InvalidPath("path '%1%' is not a valid store path", storePath);
+    if (!store->isValidPath(storePath))
+        throw InvalidPath("path '%1%' is not a valid store path", store->printStorePath(storePath));
 
-    auto i = nars.find(storePath);
+    auto i = nars.find(std::string(storePath.hashPart()));
     if (i != nars.end()) return {i->second, restPath};
 
     StringSink sink;
     std::string listing;
     Path cacheFile;
 
-    if (cacheDir != "" && pathExists(cacheFile = makeCacheFile(storePath, "nar"))) {
+    if (cacheDir != "" && pathExists(cacheFile = makeCacheFile(storePath.hashPart(), "nar"))) {
 
         try {
-            listing = nix::readFile(makeCacheFile(storePath, "ls"));
+            listing = nix::readFile(makeCacheFile(storePath.hashPart(), "ls"));
 
             auto narAccessor = makeLazyNarAccessor(listing,
                 [cacheFile](uint64_t offset, uint64_t length) {
@@ -81,7 +80,7 @@ std::pair<ref<FSAccessor>, Path> RemoteFSAccessor::fetch(const Path & path_)
                     return buf;
                 });
 
-            nars.emplace(storePath, narAccessor);
+            nars.emplace(storePath.hashPart(), narAccessor);
             return {narAccessor, restPath};
 
         } catch (SysError &) { }
@@ -90,15 +89,15 @@ std::pair<ref<FSAccessor>, Path> RemoteFSAccessor::fetch(const Path & path_)
             *sink.s = nix::readFile(cacheFile);
 
             auto narAccessor = makeNarAccessor(sink.s);
-            nars.emplace(storePath, narAccessor);
+            nars.emplace(storePath.hashPart(), narAccessor);
             return {narAccessor, restPath};
 
         } catch (SysError &) { }
     }
 
-    store->narFromPath(store->parseStorePath(storePath), sink);
+    store->narFromPath(storePath, sink);
     auto narAccessor = makeNarAccessor(sink.s);
-    addToCache(storePath, *sink.s, narAccessor);
+    addToCache(storePath.hashPart(), *sink.s, narAccessor);
     return {narAccessor, restPath};
 }
 

src/libstore/remote-fs-accessor.hh

@@ -10,7 +10,7 @@ class RemoteFSAccessor : public FSAccessor
 {
     ref<Store> store;
 
-    std::map<Path, ref<FSAccessor>> nars;
+    std::map<std::string, ref<FSAccessor>> nars;
 
     Path cacheDir;
 
@@ -18,9 +18,9 @@ class RemoteFSAccessor : public FSAccessor
 
     friend class BinaryCacheStore;
 
-    Path makeCacheFile(const Path & storePath, const std::string & ext);
+    Path makeCacheFile(std::string_view hashPart, const std::string & ext);
 
-    void addToCache(const Path & storePath, const std::string & nar,
+    void addToCache(std::string_view hashPart, const std::string & nar,
         ref<FSAccessor> narAccessor);
 
 public:

src/libstore/remote-store.cc

@@ -489,7 +489,7 @@ std::optional<StorePath> RemoteStore::queryPathFromHashPart(const std::string &
 
 
 void RemoteStore::addToStore(const ValidPathInfo & info, Source & source,
-    RepairFlag repair, CheckSigsFlag checkSigs, std::shared_ptr<FSAccessor> accessor)
+    RepairFlag repair, CheckSigsFlag checkSigs)
 {
     auto conn(getConnection());
 
@@ -521,14 +521,89 @@ void RemoteStore::addToStore(const ValidPathInfo & info, Source & source,
         conn->to << wopAddToStoreNar
                  << printStorePath(info.path)
                  << (info.deriver ? printStorePath(*info.deriver) : "")
-                 << info.narHash.to_string(Base16, false);
+                 << info.narHash->to_string(Base16, false);
         writeStorePaths(*this, conn->to, info.references);
         conn->to << info.registrationTime << info.narSize
                  << info.ultimate << info.sigs << renderContentAddress(info.ca)
                  << repair << !checkSigs;
-        bool tunnel = GET_PROTOCOL_MINOR(conn->daemonVersion) >= 21;
-        if (!tunnel) copyNAR(source, conn->to);
-        conn.processStderr(0, tunnel ? &source : nullptr);
+
+        if (GET_PROTOCOL_MINOR(conn->daemonVersion) >= 23) {
+
+            std::exception_ptr ex;
+
+            struct FramedSink : BufferedSink
+            {
+                ConnectionHandle & conn;
+                std::exception_ptr & ex;
+
+                FramedSink(ConnectionHandle & conn, std::exception_ptr & ex) : conn(conn), ex(ex)
+                { }
+
+                ~FramedSink()
+                {
+                    try {
+                        conn->to << 0;
+                        conn->to.flush();
+                    } catch (...) {
+                        ignoreException();
+                    }
+                }
+
+                void write(const unsigned char * data, size_t len) override
+                {
+                    /* Don't send more data if the remote has
+                       encountered an error. */
+                    if (ex) {
+                        auto ex2 = ex;
+                        ex = nullptr;
+                        std::rethrow_exception(ex2);
+                    }
+                    conn->to << len;
+                    conn->to(data, len);
+                };
+            };
+
+            /* Handle log messages / exceptions from the remote on a
+               separate thread. */
+            std::thread stderrThread([&]()
+            {
+                try {
+                    conn.processStderr();
+                } catch (...) {
+                    ex = std::current_exception();
+                }
+            });
+
+            Finally joinStderrThread([&]()
+            {
+                if (stderrThread.joinable()) {
+                    stderrThread.join();
+                    if (ex) {
+                        try {
+                            std::rethrow_exception(ex);
+                        } catch (...) {
+                            ignoreException();
+                        }
+                    }
+                }
+            });
+
+            {
+                FramedSink sink(conn, ex);
+                copyNAR(source, sink);
+                sink.flush();
+            }
+
+            stderrThread.join();
+            if (ex)
+                std::rethrow_exception(ex);
+
+        } else if (GET_PROTOCOL_MINOR(conn->daemonVersion) >= 21) {
+            conn.processStderr(0, &source);
+        } else {
+            copyNAR(source, conn->to);
+            conn.processStderr(0, nullptr);
+        }
     }
 }
 
@@ -730,7 +805,7 @@ void RemoteStore::addSignatures(const StorePath & storePath, const StringSet & s
 
 void RemoteStore::queryMissing(const std::vector<StorePathWithOutputs> & targets,
     StorePathSet & willBuild, StorePathSet & willSubstitute, StorePathSet & unknown,
-    unsigned long long & downloadSize, unsigned long long & narSize)
+    uint64_t & downloadSize, uint64_t & narSize)
 {
     {
         auto conn(getConnection());

src/libstore/remote-store.hh

@@ -60,8 +60,7 @@ public:
         SubstitutablePathInfos & infos) override;
 
     void addToStore(const ValidPathInfo & info, Source & nar,
-        RepairFlag repair, CheckSigsFlag checkSigs,
-        std::shared_ptr<FSAccessor> accessor) override;
+        RepairFlag repair, CheckSigsFlag checkSigs) override;
 
     StorePath addToStore(const string & name, const Path & srcPath,
         FileIngestionMethod method = FileIngestionMethod::Recursive, HashType hashAlgo = htSHA256,
@@ -95,7 +94,7 @@ public:
 
     void queryMissing(const std::vector<StorePathWithOutputs> & targets,
         StorePathSet & willBuild, StorePathSet & willSubstitute, StorePathSet & unknown,
-        unsigned long long & downloadSize, unsigned long long & narSize) override;
+        uint64_t & downloadSize, uint64_t & narSize) override;
 
     void connect() override;
 

src/libstore/s3-binary-cache-store.cc

@@ -7,7 +7,6 @@
 #include "globals.hh"
 #include "compression.hh"
 #include "filetransfer.hh"
-#include "istringstream_nocopy.hh"
 
 #include <aws/core/Aws.h>
 #include <aws/core/VersionConfig.h>
@@ -262,12 +261,11 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
     std::shared_ptr<TransferManager> transferManager;
     std::once_flag transferManagerCreated;
 
-    void uploadFile(const std::string & path, const std::string & data,
+    void uploadFile(const std::string & path,
+        std::shared_ptr<std::basic_iostream<char>> istream,
         const std::string & mimeType,
         const std::string & contentEncoding)
     {
-        auto stream = std::make_shared<istringstream_nocopy>(data);
-
         auto maxThreads = std::thread::hardware_concurrency();
 
         static std::shared_ptr<Aws::Utils::Threading::PooledThreadExecutor>
@@ -307,7 +305,7 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
 
             std::shared_ptr<TransferHandle> transferHandle =
                 transferManager->UploadFile(
-                    stream, bucketName, path, mimeType,
+                    istream, bucketName, path, mimeType,
                     Aws::Map<Aws::String, Aws::String>(),
                     nullptr /*, contentEncoding */);
 
@@ -333,9 +331,7 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
             if (contentEncoding != "")
                 request.SetContentEncoding(contentEncoding);
 
-            auto stream = std::make_shared<istringstream_nocopy>(data);
-
-            request.SetBody(stream);
+            request.SetBody(istream);
 
             auto result = checkAws(fmt("AWS error uploading '%s'", path),
                 s3Helper.client->PutObject(request));
@@ -347,25 +343,31 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
             std::chrono::duration_cast<std::chrono::milliseconds>(now2 - now1)
                 .count();
 
-        printInfo(format("uploaded 's3://%1%/%2%' (%3% bytes) in %4% ms") %
-                  bucketName % path % data.size() % duration);
+        printInfo("uploaded 's3://%s/%s' in %d ms",
+            bucketName, path, duration);
 
         stats.putTimeMs += duration;
-        stats.putBytes += data.size();
         stats.put++;
     }
 
-    void upsertFile(const std::string & path, const std::string & data,
+    void upsertFile(const std::string & path,
+        std::shared_ptr<std::basic_iostream<char>> istream,
         const std::string & mimeType) override
     {
+        auto compress = [&](std::string compression)
+        {
+            auto compressed = nix::compress(compression, StreamToSourceAdapter(istream).drain());
+            return std::make_shared<std::stringstream>(std::move(*compressed));
+        };
+
         if (narinfoCompression != "" && hasSuffix(path, ".narinfo"))
-            uploadFile(path, *compress(narinfoCompression, data), mimeType, narinfoCompression);
+            uploadFile(path, compress(narinfoCompression), mimeType, narinfoCompression);
         else if (lsCompression != "" && hasSuffix(path, ".ls"))
-            uploadFile(path, *compress(lsCompression, data), mimeType, lsCompression);
+            uploadFile(path, compress(lsCompression), mimeType, lsCompression);
         else if (logCompression != "" && hasPrefix(path, "log/"))
-            uploadFile(path, *compress(logCompression, data), mimeType, logCompression);
+            uploadFile(path, compress(logCompression), mimeType, logCompression);
         else
-            uploadFile(path, data, mimeType, "");
+            uploadFile(path, istream, mimeType, "");
     }
 
     void getFile(const std::string & path, Sink & sink) override
@@ -410,7 +412,7 @@ struct S3BinaryCacheStoreImpl : public S3BinaryCacheStore
             for (auto object : contents) {
                 auto & key = object.GetKey();
                 if (key.size() != 40 || !hasSuffix(key, ".narinfo")) continue;
-                paths.insert(parseStorePath(storeDir + "/" + key.substr(0, key.size() - 8) + "-unknown"));
+                paths.insert(parseStorePath(storeDir + "/" + key.substr(0, key.size() - 8) + "-" + MissingName));
             }
 
             marker = res.GetNextMarker();

src/libstore/s3-binary-cache-store.hh

@@ -19,7 +19,6 @@ public:
     struct Stats
     {
         std::atomic<uint64_t> put{0};
-        std::atomic<uint64_t> putBytes{0};
         std::atomic<uint64_t> putTimeMs{0};
         std::atomic<uint64_t> get{0};
         std::atomic<uint64_t> getBytes{0};

src/libstore/sqlite.cc

@@ -61,6 +61,11 @@ void SQLite::exec(const std::string & stmt)
     });
 }
 
+uint64_t SQLite::getLastInsertedRowId()
+{
+    return sqlite3_last_insert_rowid(db);
+}
+
 void SQLiteStmt::create(sqlite3 * db, const string & sql)
 {
     checkInterrupt();
@@ -95,10 +100,10 @@ SQLiteStmt::Use::~Use()
     sqlite3_reset(stmt);
 }
 
-SQLiteStmt::Use & SQLiteStmt::Use::operator () (const std::string & value, bool notNull)
+SQLiteStmt::Use & SQLiteStmt::Use::operator () (std::string_view value, bool notNull)
 {
     if (notNull) {
-        if (sqlite3_bind_text(stmt, curArg++, value.c_str(), -1, SQLITE_TRANSIENT) != SQLITE_OK)
+        if (sqlite3_bind_text(stmt, curArg++, value.data(), -1, SQLITE_TRANSIENT) != SQLITE_OK)
             throwSQLiteError(stmt.db, "binding argument");
     } else
         bind();

src/libstore/sqlite.hh

@@ -26,6 +26,8 @@ struct SQLite
     void isCache();
 
     void exec(const std::string & stmt);
+
+    uint64_t getLastInsertedRowId();
 };
 
 /* RAII wrapper to create and destroy SQLite prepared statements. */
@@ -54,7 +56,7 @@ struct SQLiteStmt
         ~Use();
 
         /* Bind the next parameter. */
-        Use & operator () (const std::string & value, bool notNull = true);
+        Use & operator () (std::string_view value, bool notNull = true);
         Use & operator () (const unsigned char * data, size_t len, bool notNull = true);
         Use & operator () (int64_t value, bool notNull = true);
         Use & bind(); // null

src/libstore/store-api.cc

@@ -21,15 +21,15 @@ bool Store::isInStore(const Path & path) const
 }
 
 
-Path Store::toStorePath(const Path & path) const
+std::pair<StorePath, Path> Store::toStorePath(const Path & path) const
 {
     if (!isInStore(path))
         throw Error("path '%1%' is not in the Nix store", path);
     Path::size_type slash = path.find('/', storeDir.size() + 1);
     if (slash == Path::npos)
-        return path;
+        return {parseStorePath(path), ""};
     else
-        return Path(path, 0, slash);
+        return {parseStorePath(std::string_view(path).substr(0, slash)), path.substr(slash)};
 }
 
 
@@ -42,14 +42,14 @@ Path Store::followLinksToStore(std::string_view _path) const
         path = absPath(target, dirOf(path));
     }
     if (!isInStore(path))
-        throw NotInStore("path '%1%' is not in the Nix store", path);
+        throw BadStorePath("path '%1%' is not in the Nix store", path);
     return path;
 }
 
 
 StorePath Store::followLinksToStorePath(std::string_view path) const
 {
-    return parseStorePath(toStorePath(followLinksToStore(path)));
+    return toStorePath(followLinksToStore(path)).first;
 }
 
 
@@ -239,20 +239,73 @@ StorePath Store::computeStorePathForText(const string & name, const string & s,
 }
 
 
+/*
+The aim of this function is to compute in one pass the correct ValidPathInfo for
+the files that we are trying to add to the store. To accomplish that in one
+pass, given the different kind of inputs that we can take (normal nar archives,
+nar archives with non SHA-256 hashes, and flat files), we set up a net of sinks
+and aliases. Also, since the dataflow is obfuscated by this, we include here a
+graphviz diagram:
+
+digraph graphname {
+    node [shape=box]
+    fileSource -> narSink
+    narSink [style=dashed]
+    narSink -> unsualHashTee [style = dashed, label = "Recursive && !SHA-256"]
+    narSink -> narHashSink [style = dashed, label = "else"]
+    unsualHashTee -> narHashSink
+    unsualHashTee -> caHashSink
+    fileSource -> parseSink
+    parseSink [style=dashed]
+    parseSink-> fileSink [style = dashed, label = "Flat"]
+    parseSink -> blank [style = dashed, label = "Recursive"]
+    fileSink -> caHashSink
+}
+*/
 ValidPathInfo Store::addToStoreSlow(std::string_view name, const Path & srcPath,
     FileIngestionMethod method, HashType hashAlgo,
     std::optional<Hash> expectedCAHash)
 {
-    /* FIXME: inefficient: we're reading/hashing 'tmpFile' three
-       times. */
+    HashSink narHashSink { htSHA256 };
+    HashSink caHashSink { hashAlgo };
 
-    auto [narHash, narSize] = hashPath(htSHA256, srcPath);
+    /* Note that fileSink and unusualHashTee must be mutually exclusive, since
+       they both write to caHashSink. Note that that requisite is currently true
+       because the former is only used in the flat case. */
+    RetrieveRegularNARSink fileSink { caHashSink };
+    TeeSink unusualHashTee { narHashSink, caHashSink };
 
-    auto hash = method == FileIngestionMethod::Recursive
-        ? hashAlgo == htSHA256
-          ? narHash
-          : hashPath(hashAlgo, srcPath).first
-        : hashFile(hashAlgo, srcPath);
+    auto & narSink = method == FileIngestionMethod::Recursive && hashAlgo != htSHA256
+        ? static_cast<Sink &>(unusualHashTee)
+        : narHashSink;
+
+    /* Functionally, this means that fileSource will yield the content of
+       srcPath. The fact that we use scratchpadSink as a temporary buffer here
+       is an implementation detail. */
+    auto fileSource = sinkToSource([&](Sink & scratchpadSink) {
+        dumpPath(srcPath, scratchpadSink);
+    });
+
+    /* tapped provides the same data as fileSource, but we also write all the
+       information to narSink. */
+    TeeSource tapped { *fileSource, narSink };
+
+    ParseSink blank;
+    auto & parseSink = method == FileIngestionMethod::Flat
+        ? fileSink
+        : blank;
+
+    /* The information that flows from tapped (besides being replicated in
+       narSink), is now put in parseSink. */
+    parseDump(parseSink, tapped);
+
+    /* We extract the result of the computation from the sink by calling
+       finish. */
+    auto [narHash, narSize] = narHashSink.finish();
+
+    auto hash = method == FileIngestionMethod::Recursive && hashAlgo == htSHA256
+        ? narHash
+        : caHashSink.finish().first;
 
     if (expectedCAHash && expectedCAHash != hash)
         throw Error("hash mismatch for '%s'", srcPath);
@@ -263,8 +316,8 @@ ValidPathInfo Store::addToStoreSlow(std::string_view name, const Path & srcPath,
     info.ca = FixedOutputHash { .method = method, .hash = hash };
 
     if (!isValidPath(info.path)) {
-        auto source = sinkToSource([&](Sink & sink) {
-            dumpPath(srcPath, sink);
+        auto source = sinkToSource([&](Sink & scratchpadSink) {
+            dumpPath(srcPath, scratchpadSink);
         });
         addToStore(info, *source);
     }
@@ -368,6 +421,14 @@ ref<const ValidPathInfo> Store::queryPathInfo(const StorePath & storePath)
 }
 
 
+static bool goodStorePath(const StorePath & expected, const StorePath & actual)
+{
+    return
+        expected.hashPart() == actual.hashPart()
+        && (expected.name() == Store::MissingName || expected.name() == actual.name());
+}
+
+
 void Store::queryPathInfo(const StorePath & storePath,
     Callback<ref<const ValidPathInfo>> callback) noexcept
 {
@@ -395,7 +456,7 @@ void Store::queryPathInfo(const StorePath & storePath,
                     state_->pathInfoCache.upsert(hashPart,
                         res.first == NarInfoDiskCache::oInvalid ? PathInfoCacheValue{} : PathInfoCacheValue{ .value = res.second });
                     if (res.first == NarInfoDiskCache::oInvalid ||
-                        res.second->path != storePath)
+                        !goodStorePath(storePath, res.second->path))
                         throw InvalidPath("path '%s' is not valid", printStorePath(storePath));
                 }
                 return callback(ref<const ValidPathInfo>(res.second));
@@ -407,7 +468,7 @@ void Store::queryPathInfo(const StorePath & storePath,
     auto callbackPtr = std::make_shared<decltype(callback)>(std::move(callback));
 
     queryPathInfoUncached(storePath,
-        {[this, storePath{printStorePath(storePath)}, hashPart, callbackPtr](std::future<std::shared_ptr<const ValidPathInfo>> fut) {
+        {[this, storePathS{printStorePath(storePath)}, hashPart, callbackPtr](std::future<std::shared_ptr<const ValidPathInfo>> fut) {
 
             try {
                 auto info = fut.get();
@@ -420,9 +481,11 @@ void Store::queryPathInfo(const StorePath & storePath,
                     state_->pathInfoCache.upsert(hashPart, PathInfoCacheValue { .value = info });
                 }
 
-                if (!info || info->path != parseStorePath(storePath)) {
+                auto storePath = parseStorePath(storePathS);
+
+                if (!info || !goodStorePath(storePath, info->path)) {
                     stats.narInfoMissing++;
-                    throw InvalidPath("path '%s' is not valid", storePath);
+                    throw InvalidPath("path '%s' is not valid", storePathS);
                 }
 
                 (*callbackPtr)(ref<const ValidPathInfo>(info));
@@ -492,7 +555,7 @@ string Store::makeValidityRegistration(const StorePathSet & paths,
         auto info = queryPathInfo(i);
 
         if (showHash) {
-            s += info->narHash.to_string(Base16, false) + "\n";
+            s += info->narHash->to_string(Base16, false) + "\n";
             s += (format("%1%\n") % info->narSize).str();
         }
 
@@ -524,7 +587,7 @@ void Store::pathInfoToJSON(JSONPlaceholder & jsonOut, const StorePathSet & store
             auto info = queryPathInfo(storePath);
 
             jsonPath
-                .attr("narHash", info->narHash.to_string(hashBase, true))
+                .attr("narHash", info->narHash->to_string(hashBase, true))
                 .attr("narSize", info->narSize);
 
             {
@@ -567,7 +630,7 @@ void Store::pathInfoToJSON(JSONPlaceholder & jsonOut, const StorePathSet & store
                     if (!narInfo->url.empty())
                         jsonPath.attr("url", narInfo->url);
                     if (narInfo->fileHash)
-                        jsonPath.attr("downloadHash", narInfo->fileHash.to_string(Base32, true));
+                        jsonPath.attr("downloadHash", narInfo->fileHash->to_string(hashBase, true));
                     if (narInfo->fileSize)
                         jsonPath.attr("downloadSize", narInfo->fileSize);
                     if (showClosureSize)
@@ -692,9 +755,9 @@ std::map<StorePath, StorePath> copyPaths(ref<Store> srcStore, ref<Store> dstStor
 {
     auto valid = dstStore->queryValidPaths(storePaths, substitute);
 
-    PathSet missing;
+    StorePathSet missing;
     for (auto & path : storePaths)
-        if (!valid.count(path)) missing.insert(srcStore->printStorePath(path));
+        if (!valid.count(path)) missing.insert(path);
 
     std::map<StorePath, StorePath> pathsMap;
     for (auto & path : storePaths)
@@ -715,12 +778,10 @@ std::map<StorePath, StorePath> copyPaths(ref<Store> srcStore, ref<Store> dstStor
 
     ThreadPool pool;
 
-    processGraph<Path>(pool,
-        PathSet(missing.begin(), missing.end()),
-
-        [&](const Path & storePathS) {
-            auto storePath = srcStore->parseStorePath(storePathS);
+    processGraph<StorePath>(pool,
+        StorePathSet(missing.begin(), missing.end()),
 
+        [&](const StorePath & storePath) {
             auto info = srcStore->queryPathInfo(storePath);
             auto storePathForDst = storePath;
             if (info->ca && info->references.empty()) {
@@ -732,22 +793,21 @@ std::map<StorePath, StorePath> copyPaths(ref<Store> srcStore, ref<Store> dstStor
             }
             pathsMap.insert_or_assign(storePath, storePathForDst);
 
-            if (dstStore->isValidPath(storePathForDst)) {
+            if (dstStore->isValidPath(storePath)) {
                 nrDone++;
                 showProgress();
-                return PathSet();
+                return StorePathSet();
             }
 
             bytesExpected += info->narSize;
             act.setExpected(actCopyPath, bytesExpected);
 
-            return srcStore->printStorePathSet(info->references);
+            return info->references;
         },
 
-        [&](const Path & storePathS) {
+        [&](const StorePath & storePath) {
             checkInterrupt();
 
-            auto storePath = srcStore->parseStorePath(storePathS);
             auto info = srcStore->queryPathInfo(storePath);
 
             auto storePathForDst = storePath;
@@ -769,7 +829,7 @@ std::map<StorePath, StorePath> copyPaths(ref<Store> srcStore, ref<Store> dstStor
                     nrFailed++;
                     if (!settings.keepGoing)
                         throw e;
-                    logger->log(lvlError, fmt("could not copy %s: %s", storePathS, e.what()));
+                    logger->log(lvlError, fmt("could not copy %s: %s", dstStore->printStorePath(storePath), e.what()));
                     showProgress();
                     return;
                 }
@@ -845,7 +905,7 @@ std::string ValidPathInfo::fingerprint(const Store & store) const
             store.printStorePath(path));
     return
         "1;" + store.printStorePath(path) + ";"
-        + narHash.to_string(Base32, true) + ";"
+        + narHash->to_string(Base32, true) + ";"
         + std::to_string(narSize) + ";"
         + concatStringsSep(",", store.printStorePathSet(references));
 }
@@ -954,12 +1014,20 @@ ref<Store> openStore(const std::string & uri_,
     throw Error("don't know how to open Nix store '%s'", uri);
 }
 
+static bool isNonUriPath(const std::string & spec) {
+    return
+        // is not a URL
+        spec.find("://") == std::string::npos
+        // Has at least one path separator, and so isn't a single word that
+        // might be special like "auto"
+        && spec.find("/") != std::string::npos;
+}
 
 StoreType getStoreType(const std::string & uri, const std::string & stateDir)
 {
     if (uri == "daemon") {
         return tDaemon;
-    } else if (uri == "local" || hasPrefix(uri, "/")) {
+    } else if (uri == "local" || isNonUriPath(uri)) {
         return tLocal;
     } else if (uri == "" || uri == "auto") {
         if (access(stateDir.c_str(), R_OK | W_OK) == 0)
@@ -983,8 +1051,9 @@ static RegisterStoreImplementation regStore([](
             return std::shared_ptr<Store>(std::make_shared<UDSRemoteStore>(params));
         case tLocal: {
             Store::Params params2 = params;
-            if (hasPrefix(uri, "/"))
-                params2["root"] = uri;
+            if (isNonUriPath(uri)) {
+                params2["root"] = absPath(uri);
+            }
             return std::shared_ptr<Store>(std::make_shared<LocalStore>(params2));
         }
         default:

src/libstore/store-api.hh

@@ -31,7 +31,7 @@ MakeError(InvalidPath, Error);
 MakeError(Unsupported, Error);
 MakeError(SubstituteGone, Error);
 MakeError(SubstituterDisabled, Error);
-MakeError(NotInStore, Error);
+MakeError(BadStorePath, Error);
 
 
 class FSAccessor;
@@ -85,7 +85,7 @@ struct GCOptions
     StorePathSet pathsToDelete;
 
     /* Stop after at least `maxFreed' bytes have been freed. */
-    unsigned long long maxFreed{std::numeric_limits<unsigned long long>::max()};
+    uint64_t maxFreed{std::numeric_limits<uint64_t>::max()};
 };
 
 
@@ -97,7 +97,7 @@ struct GCResults
 
     /* For `gcReturnDead', `gcDeleteDead' and `gcDeleteSpecific', the
        number of bytes that would be or was freed. */
-    unsigned long long bytesFreed = 0;
+    uint64_t bytesFreed = 0;
 };
 
 
@@ -105,8 +105,8 @@ struct SubstitutablePathInfo
 {
     std::optional<StorePath> deriver;
     StorePathSet references;
-    unsigned long long downloadSize; /* 0 = unknown or inapplicable */
-    unsigned long long narSize; /* 0 = unknown */
+    uint64_t downloadSize; /* 0 = unknown or inapplicable */
+    uint64_t narSize; /* 0 = unknown */
 };
 
 typedef std::map<StorePath, SubstitutablePathInfo> SubstitutablePathInfos;
@@ -115,7 +115,8 @@ struct ValidPathInfo
 {
     StorePath path;
     std::optional<StorePath> deriver;
-    Hash narHash;
+    // TODO document this
+    std::optional<Hash> narHash;
     StorePathSet references;
     time_t registrationTime = 0;
     uint64_t narSize = 0; // 0 = unknown
@@ -317,9 +318,9 @@ public:
        the Nix store. */
     bool isStorePath(std::string_view path) const;
 
-    /* Chop off the parts after the top-level store name, e.g.,
-       /nix/store/abcd-foo/bar => /nix/store/abcd-foo. */
-    Path toStorePath(const Path & path) const;
+    /* Split a path like /nix/store/<hash>-<name>/<bla> into
+       /nix/store/<hash>-<name> and /<bla>. */
+    std::pair<StorePath, Path> toStorePath(const Path & path) const;
 
     /* Follow symlinks until we end up with a path in the Nix store. */
     Path followLinksToStore(std::string_view path) const;
@@ -388,13 +389,16 @@ public:
         SubstituteFlag maybeSubstitute = NoSubstitute);
 
     /* Query the set of all valid paths. Note that for some store
-       backends, the name part of store paths may be omitted
-       (i.e. you'll get /nix/store/<hash> rather than
+       backends, the name part of store paths may be replaced by 'x'
+       (i.e. you'll get /nix/store/<hash>-x rather than
        /nix/store/<hash>-<name>). Use queryPathInfo() to obtain the
-       full store path. */
+       full store path. FIXME: should return a set of
+       std::variant<StorePath, HashPart> to get rid of this hack. */
     virtual StorePathSet queryAllValidPaths()
     { unsupported("queryAllValidPaths"); }
 
+    constexpr static const char * MissingName = "x";
+
     /* Query information about a valid path. It is permitted to omit
        the name part of the store path. */
     ref<const ValidPathInfo> queryPathInfo(const StorePath & path);
@@ -444,8 +448,7 @@ public:
 
     /* Import a path into the store. */
     virtual void addToStore(const ValidPathInfo & info, Source & narSource,
-        RepairFlag repair = NoRepair, CheckSigsFlag checkSigs = CheckSigs,
-        std::shared_ptr<FSAccessor> accessor = 0) = 0;
+        RepairFlag repair = NoRepair, CheckSigsFlag checkSigs = CheckSigs) = 0;
 
     /* Copy the contents of a path to the store and register the
        validity the resulting path.  The resulting path is returned.
@@ -463,7 +466,7 @@ public:
         std::optional<Hash> expectedCAHash = {});
 
     // FIXME: remove?
-    virtual StorePath addToStoreFromDump(const string & dump, const string & name,
+    virtual StorePath addToStoreFromDump(Source & dump, const string & name,
         FileIngestionMethod method = FileIngestionMethod::Recursive, HashType hashAlgo = htSHA256, RepairFlag repair = NoRepair)
     {
         throw Error("addToStoreFromDump() is not supported by this store");
@@ -612,7 +615,7 @@ public:
        that will be substituted. */
     virtual void queryMissing(const std::vector<StorePathWithOutputs> & targets,
         StorePathSet & willBuild, StorePathSet & willSubstitute, StorePathSet & unknown,
-        unsigned long long & downloadSize, unsigned long long & narSize);
+        uint64_t & downloadSize, uint64_t & narSize);
 
     /* Sort a set of paths topologically under the references
        relation.  If p refers to q, then p precedes q in this list. */
@@ -628,8 +631,7 @@ public:
        the Nix store. Optionally, the contents of the NARs are
        preloaded into the specified FS accessor to speed up subsequent
        access. */
-    StorePaths importPaths(Source & source, std::shared_ptr<FSAccessor> accessor,
-        CheckSigsFlag checkSigs = CheckSigs);
+    StorePaths importPaths(Source & source, CheckSigsFlag checkSigs = CheckSigs);
 
     struct Stats
     {

src/libstore/worker-protocol.hh

@@ -6,7 +6,7 @@ namespace nix {
 #define WORKER_MAGIC_1 0x6e697863
 #define WORKER_MAGIC_2 0x6478696f
 
-#define PROTOCOL_VERSION 0x116
+#define PROTOCOL_VERSION 0x117
 #define GET_PROTOCOL_MAJOR(x) ((x) & 0xff00)
 #define GET_PROTOCOL_MINOR(x) ((x) & 0x00ff)
 

src/libutil/ansicolor.hh

@@ -11,7 +11,7 @@ namespace nix {
 #define ANSI_GREEN "\e[32;1m"
 #define ANSI_YELLOW "\e[33;1m"
 #define ANSI_BLUE "\e[34;1m"
-#define ANSI_MAGENTA "\e[35m;1m"
-#define ANSI_CYAN "\e[36m;1m"
+#define ANSI_MAGENTA "\e[35;1m"
+#define ANSI_CYAN "\e[36;1m"
 
 }

src/libutil/archive.cc

@@ -150,17 +150,17 @@ static void skipGeneric(Source & source)
 
 static void parseContents(ParseSink & sink, Source & source, const Path & path)
 {
-    unsigned long long size = readLongLong(source);
+    uint64_t size = readLongLong(source);
 
     sink.preallocateContents(size);
 
-    unsigned long long left = size;
+    uint64_t left = size;
     std::vector<unsigned char> buf(65536);
 
     while (left) {
         checkInterrupt();
         auto n = buf.size();
-        if ((unsigned long long)n > left) n = left;
+        if ((uint64_t)n > left) n = left;
         source(buf.data(), n);
         sink.receiveContents(buf.data(), n);
         left -= n;
@@ -323,7 +323,7 @@ struct RestoreSink : ParseSink
             throw SysError("fchmod");
     }
 
-    void preallocateContents(unsigned long long len)
+    void preallocateContents(uint64_t len)
     {
 #if HAVE_POSIX_FALLOCATE
         if (len) {
@@ -338,7 +338,7 @@ struct RestoreSink : ParseSink
 #endif
     }
 
-    void receiveContents(unsigned char * data, unsigned int len)
+    void receiveContents(unsigned char * data, size_t len)
     {
         writeFull(fd.get(), data, len);
     }

src/libutil/archive.hh

@@ -57,17 +57,35 @@ struct ParseSink
 
     virtual void createRegularFile(const Path & path) { };
     virtual void isExecutable() { };
-    virtual void preallocateContents(unsigned long long size) { };
-    virtual void receiveContents(unsigned char * data, unsigned int len) { };
+    virtual void preallocateContents(uint64_t size) { };
+    virtual void receiveContents(unsigned char * data, size_t len) { };
 
     virtual void createSymlink(const Path & path, const string & target) { };
 };
 
-struct TeeSink : ParseSink
+/* If the NAR archive contains a single file at top-level, then save
+   the contents of the file to `s'.  Otherwise barf. */
+struct RetrieveRegularNARSink : ParseSink
 {
-    TeeSource source;
+    bool regular = true;
+    Sink & sink;
 
-    TeeSink(Source & source) : source(source) { }
+    RetrieveRegularNARSink(Sink & sink) : sink(sink) { }
+
+    void createDirectory(const Path & path)
+    {
+        regular = false;
+    }
+
+    void receiveContents(unsigned char * data, size_t len)
+    {
+        sink(data, len);
+    }
+
+    void createSymlink(const Path & path, const string & target)
+    {
+        regular = false;
+    }
 };
 
 void parseDump(ParseSink & sink, Source & source);

src/libutil/args.cc

@@ -1,6 +1,8 @@
 #include "args.hh"
 #include "hash.hh"
 
+#include <glob.h>
+
 namespace nix {
 
 void Args::addFlag(Flag && flag_)
@@ -13,6 +15,20 @@ void Args::addFlag(Flag && flag_)
     if (flag->shortName) shortFlags[flag->shortName] = flag;
 }
 
+bool pathCompletions = false;
+std::shared_ptr<std::set<std::string>> completions;
+
+std::string completionMarker = "___COMPLETE___";
+
+std::optional<std::string> needsCompletion(std::string_view s)
+{
+    if (!completions) return {};
+    auto i = s.find(completionMarker);
+    if (i != std::string::npos)
+        return std::string(s.begin(), i);
+    return {};
+}
+
 void Args::parseCmdline(const Strings & _cmdline)
 {
     Strings pendingArgs;
@@ -20,6 +36,14 @@ void Args::parseCmdline(const Strings & _cmdline)
 
     Strings cmdline(_cmdline);
 
+    if (auto s = getEnv("NIX_GET_COMPLETIONS")) {
+        size_t n = std::stoi(*s);
+        assert(n > 0 && n <= cmdline.size());
+        *std::next(cmdline.begin(), n - 1) += completionMarker;
+        completions = std::make_shared<decltype(completions)::element_type>();
+        verbosity = lvlError;
+    }
+
     for (auto pos = cmdline.begin(); pos != cmdline.end(); ) {
 
         auto arg = *pos;
@@ -63,7 +87,7 @@ void Args::printHelp(const string & programName, std::ostream & out)
     for (auto & exp : expectedArgs) {
         std::cout << renderLabels({exp.label});
         // FIXME: handle arity > 1
-        if (exp.arity == 0) std::cout << "...";
+        if (exp.handler.arity == ArityAny) std::cout << "...";
         if (exp.optional) std::cout << "?";
     }
     std::cout << "\n";
@@ -99,18 +123,32 @@ bool Args::processFlag(Strings::iterator & pos, Strings::iterator end)
     auto process = [&](const std::string & name, const Flag & flag) -> bool {
         ++pos;
         std::vector<std::string> args;
+        bool anyCompleted = false;
         for (size_t n = 0 ; n < flag.handler.arity; ++n) {
             if (pos == end) {
                 if (flag.handler.arity == ArityAny) break;
                 throw UsageError("flag '%s' requires %d argument(s)", name, flag.handler.arity);
             }
+            if (flag.completer)
+                if (auto prefix = needsCompletion(*pos)) {
+                    anyCompleted = true;
+                    flag.completer(n, *prefix);
+                }
             args.push_back(*pos++);
         }
-        flag.handler.fun(std::move(args));
+        if (!anyCompleted)
+            flag.handler.fun(std::move(args));
         return true;
     };
 
     if (string(*pos, 0, 2) == "--") {
+        if (auto prefix = needsCompletion(*pos)) {
+            for (auto & [name, flag] : longFlags) {
+                if (!hiddenCategories.count(flag->category)
+                    && hasPrefix(name, std::string(*prefix, 2)))
+                    completions->insert("--" + name);
+            }
+        }
         auto i = longFlags.find(string(*pos, 2));
         if (i == longFlags.end()) return false;
         return process("--" + i->first, *i->second);
@@ -123,6 +161,14 @@ bool Args::processFlag(Strings::iterator & pos, Strings::iterator end)
         return process(std::string("-") + c, *i->second);
     }
 
+    if (auto prefix = needsCompletion(*pos)) {
+        if (prefix == "-") {
+            completions->insert("--");
+            for (auto & [flag, _] : shortFlags)
+                completions->insert(std::string("-") + flag);
+        }
+    }
+
     return false;
 }
 
@@ -138,12 +184,17 @@ bool Args::processArgs(const Strings & args, bool finish)
 
     bool res = false;
 
-    if ((exp.arity == 0 && finish) ||
-        (exp.arity > 0 && args.size() == exp.arity))
+    if ((exp.handler.arity == ArityAny && finish) ||
+        (exp.handler.arity != ArityAny && args.size() == exp.handler.arity))
     {
         std::vector<std::string> ss;
-        for (auto & s : args) ss.push_back(s);
-        exp.handler(std::move(ss));
+        for (const auto &[n, s] : enumerate(args)) {
+            ss.push_back(s);
+            if (exp.completer)
+                if (auto prefix = needsCompletion(s))
+                    exp.completer(n, *prefix);
+        }
+        exp.handler.fun(ss);
         expectedArgs.pop_front();
         res = true;
     }
@@ -154,6 +205,13 @@ bool Args::processArgs(const Strings & args, bool finish)
     return res;
 }
 
+static void hashTypeCompleter(size_t index, std::string_view prefix) 
+{
+    for (auto & type : hashTypes)
+        if (hasPrefix(type, prefix))
+            completions->insert(type);
+}
+
 Args::Flag Args::Flag::mkHashTypeFlag(std::string && longName, HashType * ht)
 {
     return Flag {
@@ -162,7 +220,8 @@ Args::Flag Args::Flag::mkHashTypeFlag(std::string && longName, HashType * ht)
         .labels = {"hash-algo"},
         .handler = {[ht](std::string s) {
             *ht = parseHashType(s);
-        }}
+        }},
+        .completer = hashTypeCompleter
     };
 }
 
@@ -174,10 +233,42 @@ Args::Flag Args::Flag::mkHashTypeOptFlag(std::string && longName, std::optional<
         .labels = {"hash-algo"},
         .handler = {[oht](std::string s) {
             *oht = std::optional<HashType> { parseHashType(s) };
-        }}
+        }},
+        .completer = hashTypeCompleter
     };
 }
 
+static void completePath(std::string_view prefix, bool onlyDirs)
+{
+    pathCompletions = true;
+    glob_t globbuf;
+    int flags = GLOB_NOESCAPE | GLOB_TILDE;
+    #ifdef GLOB_ONLYDIR
+    if (onlyDirs)
+        flags |= GLOB_ONLYDIR;
+    #endif
+    if (glob((std::string(prefix) + "*").c_str(), flags, nullptr, &globbuf) == 0) {
+        for (size_t i = 0; i < globbuf.gl_pathc; ++i) {
+            if (onlyDirs) {
+                auto st = lstat(globbuf.gl_pathv[i]);
+                if (!S_ISDIR(st.st_mode)) continue;
+            }
+            completions->insert(globbuf.gl_pathv[i]);
+        }
+        globfree(&globbuf);
+    }
+}
+
+void completePath(size_t, std::string_view prefix)
+{
+    completePath(prefix, false);
+}
+
+void completeDir(size_t, std::string_view prefix)
+{
+    completePath(prefix, true);
+}
+
 Strings argvToStrings(int argc, char * * argv)
 {
     Strings args;
@@ -225,18 +316,26 @@ void Command::printHelp(const string & programName, std::ostream & out)
 MultiCommand::MultiCommand(const Commands & commands)
     : commands(commands)
 {
-    expectedArgs.push_back(ExpectedArg{"command", 1, true, [=](std::vector<std::string> ss) {
-        assert(!command);
-        auto cmd = ss[0];
-        if (auto alias = get(deprecatedAliases, cmd)) {
-            warn("'%s' is a deprecated alias for '%s'", cmd, *alias);
-            cmd = *alias;
-        }
-        auto i = commands.find(cmd);
-        if (i == commands.end())
-            throw UsageError("'%s' is not a recognised command", cmd);
-        command = {cmd, i->second()};
-    }});
+    expectArgs({
+        .label = "command",
+        .optional = true,
+        .handler = {[=](std::string s) {
+            assert(!command);
+            if (auto alias = get(deprecatedAliases, s)) {
+                warn("'%s' is a deprecated alias for '%s'", s, *alias);
+                s = *alias;
+            }
+            if (auto prefix = needsCompletion(s)) {
+                for (auto & [name, command] : commands)
+                    if (hasPrefix(name, *prefix))
+                        completions->insert(name);
+            }
+            auto i = commands.find(s);
+            if (i == commands.end())
+                throw UsageError("'%s' is not a recognised command", s);
+            command = {s, i->second()};
+        }}
+    });
 
     categories[Command::catDefault] = "Available commands";
 }

src/libutil/args.hh

@@ -8,8 +8,6 @@
 
 namespace nix {
 
-MakeError(UsageError, Error);
-
 enum HashType : char;
 
 class Args
@@ -28,61 +26,67 @@ protected:
 
     static const size_t ArityAny = std::numeric_limits<size_t>::max();
 
+    struct Handler
+    {
+        std::function<void(std::vector<std::string>)> fun;
+        size_t arity;
+
+        Handler() {}
+
+        Handler(std::function<void(std::vector<std::string>)> && fun)
+            : fun(std::move(fun))
+            , arity(ArityAny)
+        { }
+
+        Handler(std::function<void()> && handler)
+            : fun([handler{std::move(handler)}](std::vector<std::string>) { handler(); })
+            , arity(0)
+        { }
+
+        Handler(std::function<void(std::string)> && handler)
+            : fun([handler{std::move(handler)}](std::vector<std::string> ss) {
+                handler(std::move(ss[0]));
+              })
+            , arity(1)
+        { }
+
+        Handler(std::function<void(std::string, std::string)> && handler)
+            : fun([handler{std::move(handler)}](std::vector<std::string> ss) {
+                handler(std::move(ss[0]), std::move(ss[1]));
+              })
+            , arity(2)
+        { }
+
+        Handler(std::vector<std::string> * dest)
+            : fun([=](std::vector<std::string> ss) { *dest = ss; })
+            , arity(ArityAny)
+        { }
+
+        template<class T>
+        Handler(T * dest)
+            : fun([=](std::vector<std::string> ss) { *dest = ss[0]; })
+            , arity(1)
+        { }
+
+        template<class T>
+        Handler(T * dest, const T & val)
+            : fun([=](std::vector<std::string> ss) { *dest = val; })
+            , arity(0)
+        { }
+    };
+
     /* Flags. */
     struct Flag
     {
         typedef std::shared_ptr<Flag> ptr;
 
-        struct Handler
-        {
-            std::function<void(std::vector<std::string>)> fun;
-            size_t arity;
-
-            Handler() {}
-
-            Handler(std::function<void(std::vector<std::string>)> && fun)
-                : fun(std::move(fun))
-                , arity(ArityAny)
-            { }
-
-            Handler(std::function<void()> && handler)
-                : fun([handler{std::move(handler)}](std::vector<std::string>) { handler(); })
-                , arity(0)
-            { }
-
-            Handler(std::function<void(std::string)> && handler)
-                : fun([handler{std::move(handler)}](std::vector<std::string> ss) {
-                    handler(std::move(ss[0]));
-                  })
-                , arity(1)
-            { }
-
-            Handler(std::function<void(std::string, std::string)> && handler)
-                : fun([handler{std::move(handler)}](std::vector<std::string> ss) {
-                    handler(std::move(ss[0]), std::move(ss[1]));
-                  })
-                , arity(2)
-            { }
-
-            template<class T>
-            Handler(T * dest)
-                : fun([=](std::vector<std::string> ss) { *dest = ss[0]; })
-                , arity(1)
-            { }
-
-            template<class T>
-            Handler(T * dest, const T & val)
-                : fun([=](std::vector<std::string> ss) { *dest = val; })
-                , arity(0)
-            { }
-        };
-
         std::string longName;
         char shortName = 0;
         std::string description;
         std::string category;
         Strings labels;
         Handler handler;
+        std::function<void(size_t, std::string_view)> completer;
 
         static Flag mkHashTypeFlag(std::string && longName, HashType * ht);
         static Flag mkHashTypeOptFlag(std::string && longName, std::optional<HashType> * oht);
@@ -99,9 +103,9 @@ protected:
     struct ExpectedArg
     {
         std::string label;
-        size_t arity; // 0 = any
-        bool optional;
-        std::function<void(std::vector<std::string>)> handler;
+        bool optional = false;
+        Handler handler;
+        std::function<void(size_t, std::string_view)> completer;
     };
 
     std::list<ExpectedArg> expectedArgs;
@@ -175,20 +179,28 @@ public:
         });
     }
 
+    void expectArgs(ExpectedArg && arg)
+    {
+        expectedArgs.emplace_back(std::move(arg));
+    }
+
     /* Expect a string argument. */
     void expectArg(const std::string & label, string * dest, bool optional = false)
     {
-        expectedArgs.push_back(ExpectedArg{label, 1, optional, [=](std::vector<std::string> ss) {
-            *dest = ss[0];
-        }});
+        expectArgs({
+            .label = label,
+            .optional = true,
+            .handler = {dest}
+        });
     }
 
     /* Expect 0 or more arguments. */
     void expectArgs(const std::string & label, std::vector<std::string> * dest)
     {
-        expectedArgs.push_back(ExpectedArg{label, 0, false, [=](std::vector<std::string> ss) {
-            *dest = std::move(ss);
-        }});
+        expectArgs({
+            .label = label,
+            .handler = {dest}
+        });
     }
 
     friend class MultiCommand;
@@ -259,4 +271,13 @@ typedef std::vector<std::pair<std::string, std::string>> Table2;
 
 void printTable(std::ostream & out, const Table2 & table);
 
+extern std::shared_ptr<std::set<std::string>> completions;
+extern bool pathCompletions;
+
+std::optional<std::string> needsCompletion(std::string_view s);
+
+void completePath(size_t, std::string_view prefix);
+
+void completeDir(size_t, std::string_view prefix);
+
 }

src/libutil/error.hh

@@ -191,6 +191,7 @@ public:
     }
 
 MakeError(Error, BaseError);
+MakeError(UsageError, Error);
 
 class SysError : public Error
 {

src/libutil/hash.cc

@@ -8,7 +8,6 @@
 #include "hash.hh"
 #include "archive.hh"
 #include "util.hh"
-#include "istringstream_nocopy.hh"
 
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -16,16 +15,22 @@
 
 namespace nix {
 
+static size_t regularHashSize(HashType type) {
+    switch (type) {
+    case htMD5: return md5HashSize;
+    case htSHA1: return sha1HashSize;
+    case htSHA256: return sha256HashSize;
+    case htSHA512: return sha512HashSize;
+    }
+    abort();
+}
+
+std::set<std::string> hashTypes = { "md5", "sha1", "sha256", "sha512" };
+
 
 void Hash::init()
 {
-    assert(type);
-    switch (*type) {
-    case htMD5: hashSize = md5HashSize; break;
-    case htSHA1: hashSize = sha1HashSize; break;
-    case htSHA256: hashSize = sha256HashSize; break;
-    case htSHA512: hashSize = sha512HashSize; break;
-    }
+    hashSize = regularHashSize(type);
     assert(hashSize <= maxHashSize);
     memset(hash, 0, maxHashSize);
 }
@@ -106,17 +111,11 @@ string printHash16or32(const Hash & hash)
 }
 
 
-HashType assertInitHashType(const Hash & h)
-{
-    assert(h.type);
-    return *h.type;
-}
-
 std::string Hash::to_string(Base base, bool includeType) const
 {
     std::string s;
     if (base == SRI || includeType) {
-        s += printHashType(assertInitHashType(*this));
+        s += printHashType(type);
         s += base == SRI ? '-' : ':';
     }
     switch (base) {
@@ -137,60 +136,66 @@ std::string Hash::to_string(Base base, bool includeType) const
 Hash::Hash(std::string_view s, HashType type) : Hash(s, std::optional { type }) { }
 Hash::Hash(std::string_view s) : Hash(s, std::optional<HashType>{}) { }
 
-Hash::Hash(std::string_view s, std::optional<HashType> type)
-    : type(type)
+Hash::Hash(std::string_view original, std::optional<HashType> optType)
 {
+    auto rest = original;
+
     size_t pos = 0;
     bool isSRI = false;
 
-    auto sep = s.find(':');
-    if (sep == string::npos) {
-        sep = s.find('-');
-        if (sep != string::npos) {
-            isSRI = true;
-        } else if (! type)
-            throw BadHash("hash '%s' does not include a type", s);
+    // Parse the has type before the separater, if there was one.
+    std::optional<HashType> optParsedType;
+    {
+        auto sep = rest.find(':');
+        if (sep == std::string_view::npos) {
+            sep = rest.find('-');
+            if (sep != std::string_view::npos)
+                isSRI = true;
+        }
+        if (sep != std::string_view::npos) {
+            auto hashRaw = rest.substr(0, sep);
+            optParsedType = parseHashType(hashRaw);
+            rest = rest.substr(sep + 1);
+        }
     }
 
-    if (sep != string::npos) {
-        string hts = string(s, 0, sep);
-        this->type = parseHashType(hts);
-        if (!this->type)
-            throw BadHash("unknown hash type '%s'", hts);
-        if (type && type != this->type)
-            throw BadHash("hash '%s' should have type '%s'", s, printHashType(*type));
-        pos = sep + 1;
+    // Either the string or user must provide the type, if they both do they
+    // must agree.
+    if (!optParsedType && !optType) {
+        throw BadHash("hash '%s' does not include a type, nor is the type otherwise known from context.", rest);
+    } else {
+        this->type = optParsedType ? *optParsedType : *optType;
+        if (optParsedType && optType && *optParsedType != *optType)
+            throw BadHash("hash '%s' should have type '%s'", original, printHashType(*optType));
     }
 
     init();
 
-    size_t size = s.size() - pos;
-
-    if (!isSRI && size == base16Len()) {
+    if (!isSRI && rest.size() == base16Len()) {
 
         auto parseHexDigit = [&](char c) {
             if (c >= '0' && c <= '9') return c - '0';
             if (c >= 'A' && c <= 'F') return c - 'A' + 10;
             if (c >= 'a' && c <= 'f') return c - 'a' + 10;
-            throw BadHash("invalid base-16 hash '%s'", s);
+            throw BadHash("invalid base-16 hash '%s'", original);
         };
 
         for (unsigned int i = 0; i < hashSize; i++) {
             hash[i] =
-                parseHexDigit(s[pos + i * 2]) << 4
-                | parseHexDigit(s[pos + i * 2 + 1]);
+                parseHexDigit(rest[pos + i * 2]) << 4
+                | parseHexDigit(rest[pos + i * 2 + 1]);
         }
     }
 
-    else if (!isSRI && size == base32Len()) {
+    else if (!isSRI && rest.size() == base32Len()) {
 
-        for (unsigned int n = 0; n < size; ++n) {
-            char c = s[pos + size - n - 1];
+        for (unsigned int n = 0; n < rest.size(); ++n) {
+            char c = rest[rest.size() - n - 1];
             unsigned char digit;
             for (digit = 0; digit < base32Chars.size(); ++digit) /* !!! slow */
                 if (base32Chars[digit] == c) break;
             if (digit >= 32)
-                throw BadHash("invalid base-32 hash '%s'", s);
+                throw BadHash("invalid base-32 hash '%s'", original);
             unsigned int b = n * 5;
             unsigned int i = b / 8;
             unsigned int j = b % 8;
@@ -200,21 +205,21 @@ Hash::Hash(std::string_view s, std::optional<HashType> type)
                 hash[i + 1] |= digit >> (8 - j);
             } else {
                 if (digit >> (8 - j))
-                    throw BadHash("invalid base-32 hash '%s'", s);
+                    throw BadHash("invalid base-32 hash '%s'", original);
             }
         }
     }
 
-    else if (isSRI || size == base64Len()) {
-        auto d = base64Decode(s.substr(pos));
+    else if (isSRI || rest.size() == base64Len()) {
+        auto d = base64Decode(rest);
         if (d.size() != hashSize)
-            throw BadHash("invalid %s hash '%s'", isSRI ? "SRI" : "base-64", s);
+            throw BadHash("invalid %s hash '%s'", isSRI ? "SRI" : "base-64", original);
         assert(hashSize);
         memcpy(hash, d.data(), hashSize);
     }
 
     else
-        throw BadHash("hash '%s' has wrong length for hash type '%s'", s, printHashType(*type));
+        throw BadHash("hash '%s' has wrong length for hash type '%s'", rest, printHashType(this->type));
 }
 
 Hash newHashAllowEmpty(std::string hashStr, std::optional<HashType> ht)
@@ -267,7 +272,7 @@ static void finish(HashType ht, Ctx & ctx, unsigned char * hash)
 }
 
 
-Hash hashString(HashType ht, const string & s)
+Hash hashString(HashType ht, std::string_view s)
 {
     Ctx ctx;
     Hash hash(ht);
@@ -334,7 +339,7 @@ HashResult hashPath(
 
 Hash compressHash(const Hash & hash, unsigned int newSize)
 {
-    Hash h;
+    Hash h(hash.type);
     h.hashSize = newSize;
     for (unsigned int i = 0; i < hash.hashSize; ++i)
         h.hash[i % newSize] ^= hash.hash[i];
@@ -342,7 +347,7 @@ Hash compressHash(const Hash & hash, unsigned int newSize)
 }
 
 
-std::optional<HashType> parseHashTypeOpt(const string & s)
+std::optional<HashType> parseHashTypeOpt(std::string_view s)
 {
     if (s == "md5") return htMD5;
     else if (s == "sha1") return htSHA1;
@@ -351,7 +356,7 @@ std::optional<HashType> parseHashTypeOpt(const string & s)
     else return std::optional<HashType> {};
 }
 
-HashType parseHashType(const string & s)
+HashType parseHashType(std::string_view s)
 {
     auto opt_h = parseHashTypeOpt(s);
     if (opt_h)

src/libutil/hash.hh

@@ -18,6 +18,8 @@ const int sha1HashSize = 20;
 const int sha256HashSize = 32;
 const int sha512HashSize = 64;
 
+extern std::set<std::string> hashTypes;
+
 extern const string base32Chars;
 
 enum Base : int { Base64, Base32, Base16, SRI };
@@ -25,14 +27,11 @@ enum Base : int { Base64, Base32, Base16, SRI };
 
 struct Hash
 {
-    static const unsigned int maxHashSize = 64;
-    unsigned int hashSize = 0;
-    unsigned char hash[maxHashSize] = {};
+    constexpr static size_t maxHashSize = 64;
+    size_t hashSize = 0;
+    uint8_t hash[maxHashSize] = {};
 
-    std::optional<HashType> type = {};
-
-    /* Create an unset hash object. */
-    Hash() { };
+    HashType type;
 
     /* Create a zero-filled hash object. */
     Hash(HashType type) : type(type) { init(); };
@@ -105,14 +104,14 @@ Hash newHashAllowEmpty(std::string hashStr, std::optional<HashType> ht);
 string printHash16or32(const Hash & hash);
 
 /* Compute the hash of the given string. */
-Hash hashString(HashType ht, const string & s);
+Hash hashString(HashType ht, std::string_view s);
 
 /* Compute the hash of the given file. */
 Hash hashFile(HashType ht, const Path & path);
 
 /* Compute the hash of the given path.  The hash is defined as
    (essentially) hashString(ht, dumpPath(path)). */
-typedef std::pair<Hash, unsigned long long> HashResult;
+typedef std::pair<Hash, uint64_t> HashResult;
 HashResult hashPath(HashType ht, const Path & path,
     PathFilter & filter = defaultPathFilter);
 
@@ -121,9 +120,10 @@ HashResult hashPath(HashType ht, const Path & path,
 Hash compressHash(const Hash & hash, unsigned int newSize);
 
 /* Parse a string representing a hash type. */
-HashType parseHashType(const string & s);
+HashType parseHashType(std::string_view s);
+
 /* Will return nothing on parse error */
-std::optional<HashType> parseHashTypeOpt(const string & s);
+std::optional<HashType> parseHashTypeOpt(std::string_view s);
 
 /* And the reverse. */
 string printHashType(HashType ht);
@@ -141,7 +141,7 @@ class HashSink : public BufferedSink, public AbstractHashSink
 private:
     HashType ht;
     Ctx * ctx;
-    unsigned long long bytes;
+    uint64_t bytes;
 
 public:
     HashSink(HashType ht);

src/libutil/istringstream_nocopy.hh

@@ -1,92 +0,0 @@
-/* This file provides a variant of std::istringstream that doesn't
-   copy its string argument. This is useful for large strings. The
-   caller must ensure that the string object is not destroyed while
-   it's referenced by this object. */
-
-#pragma once
-
-#include <string>
-#include <iostream>
-
-template <class CharT, class Traits = std::char_traits<CharT>, class Allocator = std::allocator<CharT>>
-class basic_istringbuf_nocopy : public std::basic_streambuf<CharT, Traits>
-{
-public:
-    typedef std::basic_string<CharT, Traits, Allocator> string_type;
-
-    typedef typename std::basic_streambuf<CharT, Traits>::off_type off_type;
-
-    typedef typename std::basic_streambuf<CharT, Traits>::pos_type pos_type;
-
-    typedef typename std::basic_streambuf<CharT, Traits>::int_type int_type;
-
-    typedef typename std::basic_streambuf<CharT, Traits>::traits_type traits_type;
-
-private:
-    const string_type & s;
-
-    off_type off;
-
-public:
-    basic_istringbuf_nocopy(const string_type & s) : s{s}, off{0}
-    {
-    }
-
-private:
-    pos_type seekoff(off_type off, std::ios_base::seekdir dir, std::ios_base::openmode which)
-    {
-        if (which & std::ios_base::in) {
-            this->off = dir == std::ios_base::beg
-                ? off
-                : (dir == std::ios_base::end
-                    ? s.size() + off
-                    : this->off + off);
-        }
-        return pos_type(this->off);
-    }
-
-    pos_type seekpos(pos_type pos, std::ios_base::openmode which)
-    {
-        return seekoff(pos, std::ios_base::beg, which);
-    }
-
-    std::streamsize showmanyc()
-    {
-        return s.size() - off;
-    }
-
-    int_type underflow()
-    {
-        if (typename string_type::size_type(off) == s.size())
-            return traits_type::eof();
-        return traits_type::to_int_type(s[off]);
-    }
-
-    int_type uflow()
-    {
-        if (typename string_type::size_type(off) == s.size())
-            return traits_type::eof();
-        return traits_type::to_int_type(s[off++]);
-    }
-
-    int_type pbackfail(int_type ch)
-    {
-        if (off == 0 || (ch != traits_type::eof() && ch != s[off - 1]))
-            return traits_type::eof();
-
-        return traits_type::to_int_type(s[--off]);
-    }
-
-};
-
-template <class CharT, class Traits = std::char_traits<CharT>, class Allocator = std::allocator<CharT>>
-class basic_istringstream_nocopy : public std::basic_iostream<CharT, Traits>
-{
-    typedef basic_istringbuf_nocopy<CharT, Traits, Allocator> buf_type;
-    buf_type buf;
-public:
-    basic_istringstream_nocopy(const typename buf_type::string_type & s) :
-        std::basic_iostream<CharT, Traits>(&buf), buf(s) {};
-};
-
-typedef basic_istringstream_nocopy<char> istringstream_nocopy;

src/libutil/serialise.cc

@@ -322,5 +322,18 @@ void StringSink::operator () (const unsigned char * data, size_t len)
     s->append((const char *) data, len);
 }
 
+size_t ChainSource::read(unsigned char * data, size_t len)
+{
+    if (useSecond) {
+        return source2.read(data, len);
+    } else {
+        try {
+            return source1.read(data, len);
+        } catch (EndOfFile &) {
+            useSecond = true;
+            return this->read(data, len);
+        }
+    }
+}
 
 }

src/libutil/serialise.hh

@@ -166,17 +166,30 @@ struct StringSource : Source
 };
 
 
-/* Adapter class of a Source that saves all data read to `s'. */
+/* A sink that writes all incoming data to two other sinks. */
+struct TeeSink : Sink
+{
+    Sink & sink1, & sink2;
+    TeeSink(Sink & sink1, Sink & sink2) : sink1(sink1), sink2(sink2) { }
+    virtual void operator () (const unsigned char * data, size_t len)
+    {
+        sink1(data, len);
+        sink2(data, len);
+    }
+};
+
+
+/* Adapter class of a Source that saves all data read to a sink. */
 struct TeeSource : Source
 {
     Source & orig;
-    ref<std::string> data;
-    TeeSource(Source & orig)
-        : orig(orig), data(make_ref<std::string>()) { }
+    Sink & sink;
+    TeeSource(Source & orig, Sink & sink)
+        : orig(orig), sink(sink) { }
     size_t read(unsigned char * data, size_t len)
     {
         size_t n = orig.read(data, len);
-        this->data->append((const char *) data, n);
+        sink(data, n);
         return n;
     }
 };
@@ -243,6 +256,19 @@ struct LambdaSource : Source
     }
 };
 
+/* Chain two sources together so after the first is exhausted, the second is
+   used */
+struct ChainSource : Source
+{
+    Source & source1, & source2;
+    bool useSecond = false;
+    ChainSource(Source & s1, Source & s2)
+        : source1(s1), source2(s2)
+    { }
+
+    size_t read(unsigned char * data, size_t len) override;
+};
+
 
 /* Convert a function that feeds data into a Sink into a Source. The
    Source executes the function as a coroutine. */
@@ -286,14 +312,14 @@ T readNum(Source & source)
     source(buf, sizeof(buf));
 
     uint64_t n =
-        ((unsigned long long) buf[0]) |
-        ((unsigned long long) buf[1] << 8) |
-        ((unsigned long long) buf[2] << 16) |
-        ((unsigned long long) buf[3] << 24) |
-        ((unsigned long long) buf[4] << 32) |
-        ((unsigned long long) buf[5] << 40) |
-        ((unsigned long long) buf[6] << 48) |
-        ((unsigned long long) buf[7] << 56);
+        ((uint64_t) buf[0]) |
+        ((uint64_t) buf[1] << 8) |
+        ((uint64_t) buf[2] << 16) |
+        ((uint64_t) buf[3] << 24) |
+        ((uint64_t) buf[4] << 32) |
+        ((uint64_t) buf[5] << 40) |
+        ((uint64_t) buf[6] << 48) |
+        ((uint64_t) buf[7] << 56);
 
     if (n > std::numeric_limits<T>::max())
         throw SerialisationError("serialised integer %d is too large for type '%s'", n, typeid(T).name());
@@ -336,4 +362,27 @@ Source & operator >> (Source & in, bool & b)
 }
 
 
+/* An adapter that converts a std::basic_istream into a source. */
+struct StreamToSourceAdapter : Source
+{
+    std::shared_ptr<std::basic_istream<char>> istream;
+
+    StreamToSourceAdapter(std::shared_ptr<std::basic_istream<char>> istream)
+        : istream(istream)
+    { }
+
+    size_t read(unsigned char * data, size_t len) override
+    {
+        if (!istream->read((char *) data, len)) {
+            if (istream->eof()) {
+                if (istream->gcount() == 0)
+                    throw EndOfFile("end of file");
+            } else
+                throw Error("I/O error in StreamToSourceAdapter");
+        }
+        return istream->gcount();
+    }
+};
+
+
 }

src/libutil/util.cc

@@ -23,6 +23,7 @@
 #include <sys/types.h>
 #include <sys/socket.h>
 #include <sys/wait.h>
+#include <sys/time.h>
 #include <sys/un.h>
 #include <unistd.h>
 
@@ -79,7 +80,7 @@ void replaceEnv(std::map<std::string, std::string> newEnv)
 }
 
 
-Path absPath(Path path, std::optional<Path> dir)
+Path absPath(Path path, std::optional<Path> dir, bool resolveSymlinks)
 {
     if (path[0] != '/') {
         if (!dir) {
@@ -100,7 +101,7 @@ Path absPath(Path path, std::optional<Path> dir)
         }
         path = *dir + "/" + path;
     }
-    return canonPath(path);
+    return canonPath(path, resolveSymlinks);
 }
 
 
@@ -345,7 +346,6 @@ void writeFile(const Path & path, Source & source, mode_t mode)
     }
 }
 
-
 string readLine(int fd)
 {
     string s;
@@ -374,7 +374,7 @@ void writeLine(int fd, string s)
 }
 
 
-static void _deletePath(int parentfd, const Path & path, unsigned long long & bytesFreed)
+static void _deletePath(int parentfd, const Path & path, uint64_t & bytesFreed)
 {
     checkInterrupt();
 
@@ -414,7 +414,7 @@ static void _deletePath(int parentfd, const Path & path, unsigned long long & by
     }
 }
 
-static void _deletePath(const Path & path, unsigned long long & bytesFreed)
+static void _deletePath(const Path & path, uint64_t & bytesFreed)
 {
     Path dir = dirOf(path);
     if (dir == "")
@@ -435,12 +435,12 @@ static void _deletePath(const Path & path, unsigned long long & bytesFreed)
 
 void deletePath(const Path & path)
 {
-    unsigned long long dummy;
+    uint64_t dummy;
     deletePath(path, dummy);
 }
 
 
-void deletePath(const Path & path, unsigned long long & bytesFreed)
+void deletePath(const Path & path, uint64_t & bytesFreed)
 {
     //Activity act(*logger, lvlDebug, format("recursively deleting path '%1%'") % path);
     bytesFreed = 0;
@@ -581,20 +581,31 @@ Paths createDirs(const Path & path)
 }
 
 
-void createSymlink(const Path & target, const Path & link)
+void createSymlink(const Path & target, const Path & link,
+    std::optional<time_t> mtime)
 {
     if (symlink(target.c_str(), link.c_str()))
         throw SysError("creating symlink from '%1%' to '%2%'", link, target);
+    if (mtime) {
+        struct timeval times[2];
+        times[0].tv_sec = *mtime;
+        times[0].tv_usec = 0;
+        times[1].tv_sec = *mtime;
+        times[1].tv_usec = 0;
+        if (lutimes(link.c_str(), times))
+            throw SysError("setting time of symlink '%s'", link);
+    }
 }
 
 
-void replaceSymlink(const Path & target, const Path & link)
+void replaceSymlink(const Path & target, const Path & link,
+    std::optional<time_t> mtime)
 {
     for (unsigned int n = 0; true; n++) {
         Path tmp = canonPath(fmt("%s/.%d_%s", dirOf(link), n, baseNameOf(link)));
 
         try {
-            createSymlink(target, tmp);
+            createSymlink(target, tmp, mtime);
         } catch (SysError & e) {
             if (e.errNo == EEXIST) continue;
             throw;
@@ -1006,12 +1017,14 @@ std::vector<char *> stringsToCharPtrs(const Strings & ss)
     return res;
 }
 
-
+// Output = "standard out" output stream
 string runProgram(Path program, bool searchPath, const Strings & args,
     const std::optional<std::string> & input)
 {
     RunOptions opts(program, args);
     opts.searchPath = searchPath;
+    // This allows you to refer to a program with a pathname relative to the
+    // PATH variable.
     opts.input = input;
 
     auto res = runProgram(opts);
@@ -1022,6 +1035,7 @@ string runProgram(Path program, bool searchPath, const Strings & args,
     return res.second;
 }
 
+// Output = error code + "standard out" output stream
 std::pair<int, std::string> runProgram(const RunOptions & options_)
 {
     RunOptions options(options_);
@@ -1094,6 +1108,8 @@ void runProgram2(const RunOptions & options)
 
         if (options.searchPath)
             execvp(options.program.c_str(), stringsToCharPtrs(args_).data());
+            // This allows you to refer to a program with a pathname relative
+            // to the PATH variable.
         else
             execv(options.program.c_str(), stringsToCharPtrs(args_).data());
 
@@ -1565,7 +1581,7 @@ AutoCloseFD createUnixDomainSocket(const Path & path, mode_t mode)
 
     struct sockaddr_un addr;
     addr.sun_family = AF_UNIX;
-    if (path.size() >= sizeof(addr.sun_path))
+    if (path.size() + 1 >= sizeof(addr.sun_path))
         throw Error("socket path '%1%' is too long", path);
     strcpy(addr.sun_path, path.c_str());
 

src/libutil/util.hh

@@ -49,7 +49,9 @@ void clearEnv();
 /* Return an absolutized path, resolving paths relative to the
    specified directory, or the current directory otherwise.  The path
    is also canonicalised. */
-Path absPath(Path path, std::optional<Path> dir = {});
+Path absPath(Path path,
+    std::optional<Path> dir = {},
+    bool resolveSymlinks = false);
 
 /* Canonicalise a path by removing all `.' or `..' components and
    double or trailing slashes.  Optionally resolves all symlink
@@ -123,7 +125,7 @@ void writeLine(int fd, string s);
    second variant returns the number of bytes and blocks freed. */
 void deletePath(const Path & path);
 
-void deletePath(const Path & path, unsigned long long & bytesFreed);
+void deletePath(const Path & path, uint64_t & bytesFreed);
 
 std::string getUserName();
 
@@ -147,10 +149,12 @@ Path getDataDir();
 Paths createDirs(const Path & path);
 
 /* Create a symlink. */
-void createSymlink(const Path & target, const Path & link);
+void createSymlink(const Path & target, const Path & link,
+    std::optional<time_t> mtime = {});
 
 /* Atomically create or replace a symlink. */
-void replaceSymlink(const Path & target, const Path & link);
+void replaceSymlink(const Path & target, const Path & link,
+    std::optional<time_t> mtime = {});
 
 
 /* Wrappers arount read()/write() that read/write exactly the

src/nix-build/nix-build.cc

@@ -174,7 +174,7 @@ static void _main(int argc, char * * argv)
         else if (*arg == "--run-env") // obsolete
             runEnv = true;
 
-        else if (*arg == "--command" || *arg == "--run") {
+        else if (runEnv && (*arg == "--command" || *arg == "--run")) {
             if (*arg == "--run")
                 interactive = false;
             envCommand = getArg(*arg, arg, end) + "\nexit";
@@ -192,7 +192,7 @@ static void _main(int argc, char * * argv)
         else if (*arg == "--pure") pure = true;
         else if (*arg == "--impure") pure = false;
 
-        else if (*arg == "--packages" || *arg == "-p")
+        else if (runEnv && (*arg == "--packages" || *arg == "-p"))
             packages = true;
 
         else if (inShebang && *arg == "-i") {
@@ -325,7 +325,7 @@ static void _main(int argc, char * * argv)
     auto buildPaths = [&](const std::vector<StorePathWithOutputs> & paths) {
         /* Note: we do this even when !printMissing to efficiently
            fetch binary cache data. */
-        unsigned long long downloadSize, narSize;
+        uint64_t downloadSize, narSize;
         StorePathSet willBuild, willSubstitute, unknown;
         store->queryMissing(paths,
             willBuild, willSubstitute, unknown, downloadSize, narSize);

src/nix-collect-garbage/nix-collect-garbage.cc

@@ -67,10 +67,8 @@ static int _main(int argc, char * * argv)
                 deleteOlderThan = getArg(*arg, arg, end);
             }
             else if (*arg == "--dry-run") dryRun = true;
-            else if (*arg == "--max-freed") {
-                long long maxFreed = getIntArg<long long>(*arg, arg, end, true);
-                options.maxFreed = maxFreed >= 0 ? maxFreed : 0;
-            }
+            else if (*arg == "--max-freed")
+                options.maxFreed = std::max(getIntArg<int64_t>(*arg, arg, end, true), (int64_t) 0);
             else
                 return false;
             return true;

src/nix-env/nix-env.cc

@@ -381,7 +381,8 @@ static void queryInstSources(EvalState & state,
 
                 if (path.isDerivation()) {
                     elem.setDrvPath(state.store->printStorePath(path));
-                    elem.setOutPath(state.store->printStorePath(state.store->derivationFromPath(path).findOutput("out")));
+                    auto outputs = state.store->queryDerivationOutputMap(path);
+                    elem.setOutPath(state.store->printStorePath(outputs.at("out")));
                     if (name.size() >= drvExtension.size() &&
                         string(name, name.size() - drvExtension.size()) == drvExtension)
                         name = string(name, 0, name.size() - drvExtension.size());
@@ -1208,18 +1209,17 @@ static void opSwitchProfile(Globals & globals, Strings opFlags, Strings opArgs)
 }
 
 
-static const int prevGen = -2;
+static constexpr GenerationNumber prevGen = std::numeric_limits<GenerationNumber>::max();
 
 
-static void switchGeneration(Globals & globals, int dstGen)
+static void switchGeneration(Globals & globals, GenerationNumber dstGen)
 {
     PathLocks lock;
     lockProfile(lock, globals.profile);
 
-    int curGen;
-    Generations gens = findGenerations(globals.profile, curGen);
+    auto [gens, curGen] = findGenerations(globals.profile);
 
-    Generation dst;
+    std::optional<Generation> dst;
     for (auto & i : gens)
         if ((dstGen == prevGen && i.number < curGen) ||
             (dstGen >= 0 && i.number == dstGen))
@@ -1227,18 +1227,16 @@ static void switchGeneration(Globals & globals, int dstGen)
 
     if (!dst) {
         if (dstGen == prevGen)
-            throw Error("no generation older than the current (%1%) exists",
-                curGen);
+            throw Error("no generation older than the current (%1%) exists", curGen.value_or(0));
         else
             throw Error("generation %1% does not exist", dstGen);
     }
 
-    printInfo(format("switching from generation %1% to %2%")
-        % curGen % dst.number);
+    printInfo("switching from generation %1% to %2%", curGen.value_or(0), dst->number);
 
     if (globals.dryRun) return;
 
-    switchLink(globals.profile, dst.path);
+    switchLink(globals.profile, dst->path);
 }
 
 
@@ -1249,7 +1247,7 @@ static void opSwitchGeneration(Globals & globals, Strings opFlags, Strings opArg
     if (opArgs.size() != 1)
         throw UsageError("exactly one argument expected");
 
-    int dstGen;
+    GenerationNumber dstGen;
     if (!string2Int(opArgs.front(), dstGen))
         throw UsageError("expected a generation number");
 
@@ -1278,8 +1276,7 @@ static void opListGenerations(Globals & globals, Strings opFlags, Strings opArgs
     PathLocks lock;
     lockProfile(lock, globals.profile);
 
-    int curGen;
-    Generations gens = findGenerations(globals.profile, curGen);
+    auto [gens, curGen] = findGenerations(globals.profile);
 
     RunPager pager;
 
@@ -1308,14 +1305,14 @@ static void opDeleteGenerations(Globals & globals, Strings opFlags, Strings opAr
         if(opArgs.front().size() < 2)
             throw Error("invalid number of generations ‘%1%’", opArgs.front());
         string str_max = string(opArgs.front(), 1, opArgs.front().size());
-        int max;
+        GenerationNumber max;
         if (!string2Int(str_max, max) || max == 0)
             throw Error("invalid number of generations to keep ‘%1%’", opArgs.front());
         deleteGenerationsGreaterThan(globals.profile, max, globals.dryRun);
     } else {
-        std::set<unsigned int> gens;
+        std::set<GenerationNumber> gens;
         for (auto & i : opArgs) {
-            unsigned int n;
+            GenerationNumber n;
             if (!string2Int(i, n))
                 throw UsageError("invalid generation number '%1%'", i);
             gens.insert(n);

src/nix-prefetch-url/nix-prefetch-url.cc

@@ -154,7 +154,7 @@ static int _main(int argc, char * * argv)
         /* If an expected hash is given, the file may already exist in
            the store. */
         std::optional<Hash> expectedHash;
-        Hash hash;
+        Hash hash(ht);
         std::optional<StorePath> storePath;
         if (args.size() == 2) {
             expectedHash = Hash(args[1], ht);