Merge remote-tracking branch 'john-ericson/enum-FileIngestionMethod' into no-stringly-typed-derivation-output

src/build-remote/build-remote.cc

@@ -17,7 +17,7 @@
 #include "store-api.hh"
 #include "derivations.hh"
 #include "local-store.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 using namespace nix;
 using std::cin;

src/error-demo/error-demo.cc

@@ -0,0 +1,66 @@
+#include "error.hh"
+#include "nixexpr.hh"
+
+#include <iostream>
+#include <optional>
+
+int main()
+{
+    using namespace nix;
+
+    // In each program where errors occur, this has to be set.
+    ErrorInfo::programName = std::optional("error-demo");
+
+    // Error in a program; no hint and no nix code.
+    printErrorInfo(
+        ErrorInfo { .level = elError,
+                    .name = "name",
+                    .description = "error description",
+                  });
+
+    // Warning with name, description, and hint.
+    // The hintfmt function makes all the substituted text yellow.
+    printErrorInfo(
+        ErrorInfo { .level = elWarning,
+                    .name = "name",
+                    .description = "error description",
+                    .hint =  std::optional(
+                                 hintfmt("there was a %1%", "warning")),
+                  });
+
+
+    // Warning with nix file, line number, column, and the lines of
+    // code where a warning occurred.
+    SymbolTable testTable;
+    auto problem_file = testTable.create("myfile.nix");
+
+    printErrorInfo(
+        ErrorInfo{
+            .level = elWarning,
+            .name = "warning name",
+            .description = "warning description",
+            .hint = hintfmt("this hint has %1% templated %2%!!", "yellow", "values"),
+            .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13),
+                .prevLineOfCode = std::nullopt,
+                .errLineOfCode = "this is the problem line of code",
+                .nextLineOfCode = std::nullopt
+            }});
+
+    // Error with previous and next lines of code.
+    printErrorInfo(
+        ErrorInfo{
+            .level = elError,
+            .name = "error name",
+            .description = "error description",
+            .hint = hintfmt("this hint has %1% templated %2%!!", "yellow", "values"),
+            .nixCode = NixCode {
+                .errPos = Pos(problem_file, 40, 13),
+                .prevLineOfCode = std::optional("previous line of code"),
+                .errLineOfCode = "this is the problem line of code",
+                .nextLineOfCode = std::optional("next line of code"),
+            }});
+
+
+    return 0;
+}

src/error-demo/local.mk

@@ -0,0 +1,12 @@
+programs += error-demo
+
+error-demo_DIR := $(d)
+
+error-demo_SOURCES := \
+  $(wildcard $(d)/*.cc) \
+
+error-demo_CXXFLAGS += -I src/libutil -I src/libexpr
+
+error-demo_LIBS = libutil libexpr
+
+error-demo_LDFLAGS = -pthread $(SODIUM_LIBS) $(EDITLINE_LIBS) $(BOOST_LDFLAGS) -lboost_context -lboost_thread -lboost_system

src/libexpr/attr-path.cc

@@ -37,9 +37,6 @@ std::pair<Value *, Pos> findAlongAttrPath(EvalState & state, const string & attr
 {
     Strings tokens = parseAttrPath(attrPath);
 
-    Error attrError =
-        Error(format("attribute selection path '%1%' does not match expression") % attrPath);
-
     Value * v = &vIn;
     Pos pos = noPos;
 

src/libexpr/common-eval-args.cc

@@ -1,31 +1,36 @@
 #include "common-eval-args.hh"
 #include "shared.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "util.hh"
 #include "eval.hh"
+#include "fetchers.hh"
+#include "store-api.hh"
 
 namespace nix {
 
 MixEvalArgs::MixEvalArgs()
 {
-    mkFlag()
-        .longName("arg")
-        .description("argument to be passed to Nix functions")
-        .labels({"name", "expr"})
-        .handler([&](std::vector<std::string> ss) { autoArgs[ss[0]] = 'E' + ss[1]; });
+    addFlag({
+        .longName = "arg",
+        .description = "argument to be passed to Nix functions",
+        .labels = {"name", "expr"},
+        .handler = {[&](std::string name, std::string expr) { autoArgs[name] = 'E' + expr; }}
+    });
 
-    mkFlag()
-        .longName("argstr")
-        .description("string-valued argument to be passed to Nix functions")
-        .labels({"name", "string"})
-        .handler([&](std::vector<std::string> ss) { autoArgs[ss[0]] = 'S' + ss[1]; });
+    addFlag({
+        .longName = "argstr",
+        .description = "string-valued argument to be passed to Nix functions",
+        .labels = {"name", "string"},
+        .handler = {[&](std::string name, std::string s) { autoArgs[name] = 'S' + s; }},
+    });
 
-    mkFlag()
-        .shortName('I')
-        .longName("include")
-        .description("add a path to the list of locations used to look up <...> file names")
-        .label("path")
-        .handler([&](std::string s) { searchPath.push_back(s); });
+    addFlag({
+        .longName = "include",
+        .shortName = 'I',
+        .description = "add a path to the list of locations used to look up <...> file names",
+        .labels = {"path"},
+        .handler = {[&](std::string s) { searchPath.push_back(s); }}
+    });
 }
 
 Bindings * MixEvalArgs::getAutoArgs(EvalState & state)
@@ -46,9 +51,9 @@ Bindings * MixEvalArgs::getAutoArgs(EvalState & state)
 Path lookupFileArg(EvalState & state, string s)
 {
     if (isUri(s)) {
-        CachedDownloadRequest request(s);
-        request.unpack = true;
-        return getDownloader()->downloadCached(state.store, request).path;
+        return state.store->toRealPath(
+            fetchers::downloadTarball(
+                state.store, resolveUri(s), "source", false).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-inline.hh

@@ -57,7 +57,7 @@ inline void EvalState::forceAttrs(Value & v)
 
 inline void EvalState::forceAttrs(Value & v, const Pos & pos)
 {
-    forceValue(v);
+    forceValue(v, pos);
     if (v.type != tAttrs)
         throwTypeError("value is %1% while a set was expected, at %2%", v, pos);
 }
@@ -73,7 +73,7 @@ inline void EvalState::forceList(Value & v)
 
 inline void EvalState::forceList(Value & v, const Pos & pos)
 {
-    forceValue(v);
+    forceValue(v, pos);
     if (!v.isList())
         throwTypeError("value is %1% while a list was expected, at %2%", v, pos);
 }

src/libexpr/eval.cc

@@ -5,7 +5,7 @@
 #include "derivations.hh"
 #include "globals.hh"
 #include "eval-inline.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "json.hh"
 #include "function-trace.hh"
 
@@ -22,6 +22,8 @@
 
 #if HAVE_BOEHMGC
 
+#define GC_INCLUDE_NEW
+
 #include <gc/gc.h>
 #include <gc/gc_cpp.h>
 
@@ -56,6 +58,12 @@ static char * dupStringWithLen(const char * s, size_t size)
 }
 
 
+RootValue allocRootValue(Value * v)
+{
+    return std::allocate_shared<Value *>(traceable_allocator<Value *>(), v);
+}
+
+
 static void printValue(std::ostream & str, std::set<const Value *> & active, const Value & v)
 {
     checkInterrupt();
@@ -1256,7 +1264,7 @@ void ExprWith::eval(EvalState & state, Env & env, Value & v)
 
 void ExprIf::eval(EvalState & state, Env & env, Value & v)
 {
-    (state.evalBool(env, cond) ? then : else_)->eval(state, env, v);
+    (state.evalBool(env, cond, pos) ? then : else_)->eval(state, env, v);
 }
 
 
@@ -1502,7 +1510,7 @@ NixFloat EvalState::forceFloat(Value & v, const Pos & pos)
 
 bool EvalState::forceBool(Value & v, const Pos & pos)
 {
-    forceValue(v);
+    forceValue(v, pos);
     if (v.type != tBool)
         throwTypeError("value is %1% while a Boolean was expected, at %2%", v, pos);
     return v.boolean;
@@ -1517,7 +1525,7 @@ bool EvalState::isFunctor(Value & fun)
 
 void EvalState::forceFunction(Value & v, const Pos & pos)
 {
-    forceValue(v);
+    forceValue(v, pos);
     if (v.type != tLambda && v.type != tPrimOp && v.type != tPrimOpApp && !isFunctor(v))
         throwTypeError("value is %1% while a function was expected, at %2%", v, pos);
 }
@@ -1594,7 +1602,7 @@ std::optional<string> EvalState::tryAttrsToString(const Pos & pos, Value & v,
 string EvalState::coerceToString(const Pos & pos, Value & v, PathSet & context,
     bool coerceMore, bool copyToStore)
 {
-    forceValue(v);
+    forceValue(v, pos);
 
     string s;
 

src/libexpr/json-to-value.cc

@@ -4,7 +4,6 @@
 #include <nlohmann/json.hpp>
 
 using json = nlohmann::json;
-using std::unique_ptr;
 
 namespace nix {
 
@@ -13,69 +12,69 @@ namespace nix {
 class JSONSax : nlohmann::json_sax<json> {
     class JSONState {
     protected:
-        unique_ptr<JSONState> parent;
-        Value * v;
+        std::unique_ptr<JSONState> parent;
+        RootValue v;
     public:
-        virtual unique_ptr<JSONState> resolve(EvalState &)
+        virtual std::unique_ptr<JSONState> resolve(EvalState &)
         {
             throw std::logic_error("tried to close toplevel json parser state");
-        };
-        explicit JSONState(unique_ptr<JSONState>&& p) : parent(std::move(p)), v(nullptr) {};
-        explicit JSONState(Value* v) : v(v) {};
-        JSONState(JSONState& p) = delete;
-        Value& value(EvalState & state)
+        }
+        explicit JSONState(std::unique_ptr<JSONState> && p) : parent(std::move(p)) {}
+        explicit JSONState(Value * v) : v(allocRootValue(v)) {}
+        JSONState(JSONState & p) = delete;
+        Value & value(EvalState & state)
         {
-            if (v == nullptr)
-                v = state.allocValue();
-            return *v;
-        };
-        virtual ~JSONState() {};
-        virtual void add() {};
+            if (!v)
+                v = allocRootValue(state.allocValue());
+            return **v;
+        }
+        virtual ~JSONState() {}
+        virtual void add() {}
     };
 
     class JSONObjectState : public JSONState {
         using JSONState::JSONState;
-        ValueMap attrs = ValueMap();
-        virtual unique_ptr<JSONState> resolve(EvalState & state) override
+        ValueMap attrs;
+        std::unique_ptr<JSONState> resolve(EvalState & state) override
         {
-            Value& v = parent->value(state);
+            Value & v = parent->value(state);
             state.mkAttrs(v, attrs.size());
             for (auto & i : attrs)
                 v.attrs->push_back(Attr(i.first, i.second));
             return std::move(parent);
         }
-        virtual void add() override { v = nullptr; };
+        void add() override { v = nullptr; }
     public:
-        void key(string_t& name, EvalState & state)
+        void key(string_t & name, EvalState & state)
         {
-            attrs[state.symbols.create(name)] = &value(state);
+            attrs.insert_or_assign(state.symbols.create(name), &value(state));
         }
     };
 
     class JSONListState : public JSONState {
-        ValueVector values = ValueVector();
-        virtual unique_ptr<JSONState> resolve(EvalState & state) override
+        ValueVector values;
+        std::unique_ptr<JSONState> resolve(EvalState & state) override
         {
-            Value& v = parent->value(state);
+            Value & v = parent->value(state);
             state.mkList(v, values.size());
             for (size_t n = 0; n < values.size(); ++n) {
                 v.listElems()[n] = values[n];
             }
             return std::move(parent);
         }
-        virtual void add() override {
-            values.push_back(v);
+        void add() override {
+            values.push_back(*v);
             v = nullptr;
-        };
+        }
     public:
-        JSONListState(unique_ptr<JSONState>&& p, std::size_t reserve) : JSONState(std::move(p))
+        JSONListState(std::unique_ptr<JSONState> && p, std::size_t reserve) : JSONState(std::move(p))
         {
             values.reserve(reserve);
         }
     };
 
     EvalState & state;
-    unique_ptr<JSONState> rs;
+    std::unique_ptr<JSONState> rs;
 
     template<typename T, typename... Args> inline bool handle_value(T f, Args... args)
     {
@@ -107,12 +106,12 @@ public:
         return handle_value(mkInt, val);
     }
 
-    bool number_float(number_float_t val, const string_t& s)
+    bool number_float(number_float_t val, const string_t & s)
     {
         return handle_value(mkFloat, val);
     }
 
-    bool string(string_t& val)
+    bool string(string_t & val)
     {
         return handle_value<void(Value&, const char*)>(mkString, val.c_str());
     }
@@ -123,7 +122,7 @@ public:
         return true;
     }
 
-    bool key(string_t& name)
+    bool key(string_t & name)
     {
         dynamic_cast<JSONObjectState*>(rs.get())->key(name, state);
         return true;

src/libexpr/local.mk

@@ -6,7 +6,9 @@ libexpr_DIR := $(d)
 
 libexpr_SOURCES := $(wildcard $(d)/*.cc) $(wildcard $(d)/primops/*.cc) $(d)/lexer-tab.cc $(d)/parser-tab.cc
 
-libexpr_LIBS = libutil libstore libnixrust
+libexpr_CXXFLAGS += -I src/libutil -I src/libstore -I src/libfetchers -I src/libmain -I src/libexpr
+
+libexpr_LIBS = libutil libstore libfetchers libnixrust
 
 libexpr_LDFLAGS =
 ifneq ($(OS), FreeBSD)

src/libexpr/nixexpr.hh

@@ -209,9 +209,10 @@ struct ExprList : Expr
 
 struct Formal
 {
+    Pos pos;
     Symbol name;
     Expr * def;
-    Formal(const Symbol & name, Expr * def) : name(name), def(def) { };
+    Formal(const Pos & pos, const Symbol & name, Expr * def) : pos(pos), name(name), def(def) { };
 };
 
 struct Formals
@@ -261,8 +262,9 @@ struct ExprWith : Expr
 
 struct ExprIf : Expr
 {
+    Pos pos;
     Expr * cond, * then, * else_;
-    ExprIf(Expr * cond, Expr * then, Expr * else_) : cond(cond), then(then), else_(else_) { };
+    ExprIf(const Pos & pos, Expr * cond, Expr * then, Expr * else_) : pos(pos), cond(cond), then(then), else_(else_) { };
     COMMON_METHODS
 };
 

src/libexpr/parser.y

@@ -335,7 +335,7 @@ expr_function
   ;
 
 expr_if
-  : IF expr THEN expr ELSE expr { $$ = new ExprIf($2, $4, $6); }
+  : IF expr THEN expr ELSE expr { $$ = new ExprIf(CUR_POS, $2, $4, $6); }
   | expr_op
   ;
 
@@ -531,8 +531,8 @@ formals
   ;
 
 formal
-  : ID { $$ = new Formal(data->symbols.create($1), 0); }
-  | ID '?' expr { $$ = new Formal(data->symbols.create($1), $3); }
+  : ID { $$ = new Formal(CUR_POS, data->symbols.create($1), 0); }
+  | ID '?' expr { $$ = new Formal(CUR_POS, data->symbols.create($1), $3); }
   ;
 
 %%
@@ -544,7 +544,8 @@ formal
 #include <unistd.h>
 
 #include "eval.hh"
-#include "download.hh"
+#include "filetransfer.hh"
+#include "fetchers.hh"
 #include "store-api.hh"
 
 
@@ -687,10 +688,9 @@ std::pair<bool, std::string> EvalState::resolveSearchPathElem(const SearchPathEl
 
     if (isUri(elem.second)) {
         try {
-            CachedDownloadRequest request(elem.second);
-            request.unpack = true;
-            res = { true, getDownloader()->downloadCached(store, request).path };
-        } catch (DownloadError & e) {
+            res = { true, store->toRealPath(fetchers::downloadTarball(
+                        store, resolveUri(elem.second), "source", false).storePath) };
+        } catch (FileTransferError & e) {
             printError(format("warning: Nix search path entry '%1%' cannot be downloaded, ignoring") % elem.second);
             res = { false, "" };
         }

src/libexpr/primops.cc

@@ -1,6 +1,5 @@
 #include "archive.hh"
 #include "derivations.hh"
-#include "download.hh"
 #include "eval-inline.hh"
 #include "eval.hh"
 #include "globals.hh"
@@ -122,16 +121,16 @@ static void prim_scopedImport(EvalState & state, const Pos & pos, Value * * args
         }
         w.attrs->sort();
 
-        static Value * fun = nullptr;
+        static RootValue fun;
         if (!fun) {
-            fun = state.allocValue();
+            fun = allocRootValue(state.allocValue());
             state.eval(state.parseExprFromString(
                 #include "imported-drv-to-derivation.nix.gen.hh"
-                , "/"), *fun);
+                , "/"), **fun);
         }
 
-        state.forceFunction(*fun, pos);
-        mkApp(v, *fun, w);
+        state.forceFunction(**fun, pos);
+        mkApp(v, **fun, w);
         state.forceAttrs(v, pos);
     } else {
         state.forceAttrs(*args[0]);
@@ -242,7 +241,7 @@ void prim_exec(EvalState & state, const Pos & pos, Value * * args, Value & v)
 /* Return a string representing the type of the expression. */
 static void prim_typeOf(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     string t;
     switch (args[0]->type) {
         case tInt: t = "int"; break;
@@ -270,7 +269,7 @@ static void prim_typeOf(EvalState & state, const Pos & pos, Value * * args, Valu
 /* Determine whether the argument is the null value. */
 static void prim_isNull(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tNull);
 }
 
@@ -278,7 +277,7 @@ static void prim_isNull(EvalState & state, const Pos & pos, Value * * args, Valu
 /* Determine whether the argument is a function. */
 static void prim_isFunction(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     bool res;
     switch (args[0]->type) {
         case tLambda:
@@ -297,21 +296,21 @@ static void prim_isFunction(EvalState & state, const Pos & pos, Value * * args,
 /* Determine whether the argument is an integer. */
 static void prim_isInt(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tInt);
 }
 
 /* Determine whether the argument is a float. */
 static void prim_isFloat(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tFloat);
 }
 
 /* Determine whether the argument is a string. */
 static void prim_isString(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tString);
 }
 
@@ -319,14 +318,14 @@ static void prim_isString(EvalState & state, const Pos & pos, Value * * args, Va
 /* Determine whether the argument is a Boolean. */
 static void prim_isBool(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tBool);
 }
 
 /* Determine whether the argument is a path. */
 static void prim_isPath(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tPath);
 }
 
@@ -383,7 +382,7 @@ static void prim_genericClosure(EvalState & state, const Pos & pos, Value * * ar
         args[0]->attrs->find(state.symbols.create("operator"));
     if (op == args[0]->attrs->end())
         throw EvalError(format("attribute 'operator' required, at %1%") % pos);
-    state.forceValue(*op->value);
+    state.forceValue(*op->value, pos);
 
     /* Construct the closure by applying the operator to element of
        `workSet', adding the result to `workSet', continuing until
@@ -402,7 +401,7 @@ static void prim_genericClosure(EvalState & state, const Pos & pos, Value * * ar
             e->attrs->find(state.symbols.create("key"));
         if (key == e->attrs->end())
             throw EvalError(format("attribute 'key' required, at %1%") % pos);
-        state.forceValue(*key->value);
+        state.forceValue(*key->value, pos);
 
         if (!doneKeys.insert(key->value).second) continue;
         res.push_back(e);
@@ -414,7 +413,7 @@ static void prim_genericClosure(EvalState & state, const Pos & pos, Value * * ar
 
         /* Add the values returned by the operator to the work set. */
         for (unsigned int n = 0; n < call.listSize(); ++n) {
-            state.forceValue(*call.listElems()[n]);
+            state.forceValue(*call.listElems()[n], pos);
             workSet.push_back(call.listElems()[n]);
         }
     }
@@ -446,7 +445,7 @@ static void prim_throw(EvalState & state, const Pos & pos, Value * * args, Value
 static void prim_addErrorContext(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     try {
-        state.forceValue(*args[1]);
+        state.forceValue(*args[1], pos);
         v = *args[1];
     } catch (Error & e) {
         PathSet context;
@@ -462,7 +461,7 @@ static void prim_tryEval(EvalState & state, const Pos & pos, Value * * args, Val
 {
     state.mkAttrs(v, 2);
     try {
-        state.forceValue(*args[0]);
+        state.forceValue(*args[0], pos);
         v.attrs->push_back(Attr(state.sValue, args[0]));
         mkBool(*state.allocAttr(v, state.symbols.create("success")), true);
     } catch (AssertionError & e) {
@@ -484,8 +483,8 @@ static void prim_getEnv(EvalState & state, const Pos & pos, Value * * args, Valu
 /* Evaluate the first argument, then return the second argument. */
 static void prim_seq(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
-    state.forceValue(*args[1]);
+    state.forceValue(*args[0], pos);
+    state.forceValue(*args[1], pos);
     v = *args[1];
 }
 
@@ -495,7 +494,7 @@ static void prim_seq(EvalState & state, const Pos & pos, Value * * args, Value &
 static void prim_deepSeq(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     state.forceValueDeep(*args[0]);
-    state.forceValue(*args[1]);
+    state.forceValue(*args[1], pos);
     v = *args[1];
 }
 
@@ -504,12 +503,12 @@ static void prim_deepSeq(EvalState & state, const Pos & pos, Value * * args, Val
    return the second expression.  Useful for debugging. */
 static void prim_trace(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     if (args[0]->type == tString)
         printError(format("trace: %1%") % args[0]->string.s);
     else
         printError(format("trace: %1%") % *args[0]);
-    state.forceValue(*args[1]);
+    state.forceValue(*args[1], pos);
     v = *args[1];
 }
 
@@ -563,7 +562,7 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
 
     std::optional<std::string> outputHash;
     std::string outputHashAlgo;
-    FileIngestionMethod outputHashRecursive = FileIngestionMethod::Flat;
+    auto ingestionMethod = FileIngestionMethod::Flat;
 
     StringSet outputs;
     outputs.insert("out");
@@ -574,8 +573,8 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
         vomit("processing attribute '%1%'", key);
 
         auto handleHashMode = [&](const std::string & s) {
-            if (s == "recursive") outputHashRecursive = FileIngestionMethod::Recursive;
-            else if (s == "flat") outputHashRecursive = FileIngestionMethod::Flat;
+            if (s == "recursive") ingestionMethod = FileIngestionMethod::Recursive;
+            else if (s == "flat") ingestionMethod = FileIngestionMethod::Flat;
             else throw EvalError("invalid value '%s' for 'outputHashMode' attribute, at %s", s, posDrvName);
         };
 
@@ -600,7 +599,7 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
         try {
 
             if (ignoreNulls) {
-                state.forceValue(*i->value);
+                state.forceValue(*i->value, pos);
                 if (i->value->type == tNull) continue;
             }
 
@@ -722,11 +721,11 @@ static void prim_derivationStrict(EvalState & state, const Pos & pos, Value * *
         HashType ht = outputHashAlgo.empty() ? htUnknown : parseHashType(outputHashAlgo);
         Hash h(*outputHash, ht);
 
-        auto outPath = state.store->makeFixedOutputPath(outputHashRecursive, h, drvName);
+        auto outPath = state.store->makeFixedOutputPath(ingestionMethod, h, drvName);
         if (!jsonObject) drv.env["out"] = state.store->printStorePath(outPath);
         drv.outputs.insert_or_assign("out", DerivationOutput(
                 std::move(outPath),
-                FileSystemHash(outputHashRecursive, std::move(h))));
+                FileSystemHash(ingestionMethod, std::move(h))));
     }
 
     else {
@@ -1021,7 +1020,9 @@ static void prim_toFile(EvalState & state, const Pos & pos, Value * * args, Valu
 
     for (auto path : context) {
         if (path.at(0) != '/')
-            throw EvalError(format("in 'toFile': the file '%1%' cannot refer to derivation outputs, at %2%") % name % pos);
+            throw EvalError(format(
+                "in 'toFile': the file named '%1%' must not contain a reference "
+                "to a derivation but contains (%2%), at %3%") % name % path % pos);
         refs.insert(state.store->parseStorePath(path));
     }
 
@@ -1091,7 +1092,7 @@ static void prim_filterSource(EvalState & state, const Pos & pos, Value * * args
     if (!context.empty())
         throw EvalError(format("string '%1%' cannot refer to other paths, at %2%") % path % pos);
 
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     if (args[0]->type != tLambda)
         throw TypeError(format("first argument in call to 'filterSource' is not a function but %1%, at %2%") % showType(*args[0]) % pos);
 
@@ -1117,7 +1118,7 @@ static void prim_path(EvalState & state, const Pos & pos, Value * * args, Value
         } else if (attr.name == state.sName)
             name = state.forceStringNoCtx(*attr.value, *attr.pos);
         else if (n == "filter") {
-            state.forceValue(*attr.value);
+            state.forceValue(*attr.value, pos);
             filterFun = attr.value;
         } else if (n == "recursive")
             recursive = FileIngestionMethod { state.forceBool(*attr.value, *attr.pos) };
@@ -1188,7 +1189,7 @@ void prim_getAttr(EvalState & state, const Pos & pos, Value * * args, Value & v)
         throw EvalError(format("attribute '%1%' missing, at %2%") % attr % pos);
     // !!! add to stack trace?
     if (state.countCalls && i->pos) state.attrSelects[*i->pos]++;
-    state.forceValue(*i->value);
+    state.forceValue(*i->value, pos);
     v = *i->value;
 }
 
@@ -1218,7 +1219,7 @@ static void prim_hasAttr(EvalState & state, const Pos & pos, Value * * args, Val
 /* Determine whether the argument is a set. */
 static void prim_isAttrs(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->type == tAttrs);
 }
 
@@ -1344,7 +1345,7 @@ static void prim_catAttrs(EvalState & state, const Pos & pos, Value * * args, Va
 */
 static void prim_functionArgs(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     if (args[0]->type != tLambda)
         throw TypeError(format("'functionArgs' requires a function, at %1%") % pos);
 
@@ -1354,9 +1355,12 @@ static void prim_functionArgs(EvalState & state, const Pos & pos, Value * * args
     }
 
     state.mkAttrs(v, args[0]->lambda.fun->formals->formals.size());
-    for (auto & i : args[0]->lambda.fun->formals->formals)
+    for (auto & i : args[0]->lambda.fun->formals->formals) {
         // !!! should optimise booleans (allocate only once)
-        mkBool(*state.allocAttr(v, i.name), i.def);
+        Value * value = state.allocValue();
+        v.attrs->push_back(Attr(i.name, value, &i.pos));
+        mkBool(*value, i.def);
+    }
     v.attrs->sort();
 }
 
@@ -1387,7 +1391,7 @@ static void prim_mapAttrs(EvalState & state, const Pos & pos, Value * * args, Va
 /* Determine whether the argument is a list. */
 static void prim_isList(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
+    state.forceValue(*args[0], pos);
     mkBool(v, args[0]->isList());
 }
 
@@ -1397,7 +1401,7 @@ static void elemAt(EvalState & state, const Pos & pos, Value & list, int n, Valu
     state.forceList(list, pos);
     if (n < 0 || (unsigned int) n >= list.listSize())
         throw Error(format("list index %1% is out of bounds, at %2%") % n % pos);
-    state.forceValue(*list.listElems()[n]);
+    state.forceValue(*list.listElems()[n], pos);
     v = *list.listElems()[n];
 }
 
@@ -1520,9 +1524,9 @@ static void prim_foldlStrict(EvalState & state, const Pos & pos, Value * * args,
             vCur = n == args[2]->listSize() - 1 ? &v : state.allocValue();
             state.callFunction(vTmp, *args[2]->listElems()[n], *vCur, pos);
         }
-        state.forceValue(v);
+        state.forceValue(v, pos);
     } else {
-        state.forceValue(*args[1]);
+        state.forceValue(*args[1], pos);
         v = *args[1];
     }
 }
@@ -1587,7 +1591,7 @@ static void prim_sort(EvalState & state, const Pos & pos, Value * * args, Value
     auto len = args[1]->listSize();
     state.mkList(v, len);
     for (unsigned int n = 0; n < len; ++n) {
-        state.forceValue(*args[1]->listElems()[n]);
+        state.forceValue(*args[1]->listElems()[n], pos);
         v.listElems()[n] = args[1]->listElems()[n];
     }
 
@@ -1622,7 +1626,7 @@ static void prim_partition(EvalState & state, const Pos & pos, Value * * args, V
 
     for (unsigned int n = 0; n < len; ++n) {
         auto vElem = args[1]->listElems()[n];
-        state.forceValue(*vElem);
+        state.forceValue(*vElem, pos);
         Value res;
         state.callFunction(*args[0], *vElem, res, pos);
         if (state.forceBool(res, pos))
@@ -1753,8 +1757,8 @@ static void prim_bitXor(EvalState & state, const Pos & pos, Value * * args, Valu
 
 static void prim_lessThan(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
-    state.forceValue(*args[0]);
-    state.forceValue(*args[1]);
+    state.forceValue(*args[0], pos);
+    state.forceValue(*args[1], pos);
     CompareValues comp;
     mkBool(v, comp(args[0], args[1]));
 }
@@ -2045,68 +2049,6 @@ static void prim_splitVersion(EvalState & state, const Pos & pos, Value * * args
 }
 
 
-/*************************************************************
- * Networking
- *************************************************************/
-
-
-void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
-    const string & who, bool unpack, const std::string & defaultName)
-{
-    CachedDownloadRequest request("");
-    request.unpack = unpack;
-    request.name = defaultName;
-
-    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")
-                request.uri = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else if (n == "sha256")
-                request.expectedHash = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA256);
-            else if (n == "name")
-                request.name = state.forceStringNoCtx(*attr.value, *attr.pos);
-            else
-                throw EvalError(format("unsupported argument '%1%' to '%2%', at %3%") % attr.name % who % attr.pos);
-        }
-
-        if (request.uri.empty())
-            throw EvalError(format("'url' argument required, at %1%") % pos);
-
-    } else
-        request.uri = state.forceStringNoCtx(*args[0], pos);
-
-    state.checkURI(request.uri);
-
-    if (evalSettings.pureEval && !request.expectedHash)
-        throw Error("in pure evaluation mode, '%s' requires a 'sha256' argument", who);
-
-    auto res = getDownloader()->downloadCached(state.store, request);
-
-    if (state.allowedPaths)
-        state.allowedPaths->insert(res.path);
-
-    mkString(v, res.storePath, PathSet({res.storePath}));
-}
-
-
-static void prim_fetchurl(EvalState & state, const Pos & pos, Value * * args, Value & v)
-{
-    fetch(state, pos, args, v, "fetchurl", false, "");
-}
-
-
-static void prim_fetchTarball(EvalState & state, const Pos & pos, Value * * args, Value & v)
-{
-    fetch(state, pos, args, v, "fetchTarball", true, "source");
-}
-
-
 /*************************************************************
  * Primop registration
  *************************************************************/
@@ -2289,10 +2231,6 @@ void EvalState::createBaseEnv()
     addPrimOp("derivationStrict", 1, prim_derivationStrict);
     addPrimOp("placeholder", 1, prim_placeholder);
 
-    // Networking
-    addPrimOp("__fetchurl", 1, prim_fetchurl);
-    addPrimOp("fetchTarball", 1, prim_fetchTarball);
-
     /* Add a wrapper around the derivation primop that computes the
        `drvPath' and `outPath' attributes lazily. */
     string path = canonPath(settings.nixDataDir + "/nix/corepkgs/derivation.nix", true);

src/libexpr/primops.hh

@@ -20,6 +20,7 @@ struct RegisterPrimOp
    them. */
 /* Load a ValueInitializer from a DSO and return whatever it initializes */
 void prim_importNative(EvalState & state, const Pos & pos, Value * * args, Value & v);
+
 /* Execute a program and parse its output */
 void prim_exec(EvalState & state, const Pos & pos, Value * * args, Value & v);
 

src/libexpr/primops/fetchGit.cc

@@ -1,203 +1,19 @@
 #include "primops.hh"
 #include "eval-inline.hh"
-#include "download.hh"
 #include "store-api.hh"
-#include "pathlocks.hh"
 #include "hash.hh"
-#include "tarfile.hh"
-
-#include <sys/time.h>
-
-#include <regex>
-
-#include <nlohmann/json.hpp>
-
-using namespace std::string_literals;
+#include "fetchers.hh"
+#include "url.hh"
 
 namespace nix {
 
-struct GitInfo
-{
-    Path storePath;
-    std::string rev;
-    std::string shortRev;
-    uint64_t revCount = 0;
-};
-
-std::regex revRegex("^[0-9a-fA-F]{40}$");
-
-GitInfo exportGit(ref<Store> store, const std::string & uri,
-    std::optional<std::string> ref, std::string rev,
-    const std::string & name)
-{
-    if (evalSettings.pureEval && rev == "")
-        throw Error("in pure evaluation mode, 'fetchGit' requires a Git revision");
-
-    if (!ref && rev == "" && hasPrefix(uri, "/") && pathExists(uri + "/.git")) {
-
-        bool clean = true;
-
-        try {
-            runProgram("git", true, { "-C", uri, "diff-index", "--quiet", "HEAD", "--" });
-        } catch (ExecError & e) {
-            if (!WIFEXITED(e.status) || WEXITSTATUS(e.status) != 1) throw;
-            clean = false;
-        }
-
-        if (!clean) {
-
-            /* This is an unclean working tree. So copy all tracked files. */
-            GitInfo gitInfo;
-            gitInfo.rev = "0000000000000000000000000000000000000000";
-            gitInfo.shortRev = std::string(gitInfo.rev, 0, 7);
-
-            auto files = tokenizeString<std::set<std::string>>(
-                runProgram("git", true, { "-C", uri, "ls-files", "-z" }), "\0"s);
-
-            PathFilter filter = [&](const Path & p) -> bool {
-                assert(hasPrefix(p, uri));
-                std::string file(p, uri.size() + 1);
-
-                auto st = lstat(p);
-
-                if (S_ISDIR(st.st_mode)) {
-                    auto prefix = file + "/";
-                    auto i = files.lower_bound(prefix);
-                    return i != files.end() && hasPrefix(*i, prefix);
-                }
-
-                return files.count(file);
-            };
-
-            gitInfo.storePath = store->printStorePath(store->addToStore("source", uri, FileIngestionMethod::Recursive, htSHA256, filter));
-
-            return gitInfo;
-        }
-
-        // clean working tree, but no ref or rev specified.  Use 'HEAD'.
-        rev = chomp(runProgram("git", true, { "-C", uri, "rev-parse", "HEAD" }));
-        ref = "HEAD"s;
-    }
-
-    if (!ref) ref = "HEAD"s;
-
-    if (rev != "" && !std::regex_match(rev, revRegex))
-        throw Error("invalid Git revision '%s'", rev);
-
-    deletePath(getCacheDir() + "/nix/git");
-
-    Path cacheDir = getCacheDir() + "/nix/gitv2/" + hashString(htSHA256, uri).to_string(Base32, false);
-
-    if (!pathExists(cacheDir)) {
-        createDirs(dirOf(cacheDir));
-        runProgram("git", true, { "init", "--bare", cacheDir });
-    }
-
-    Path localRefFile;
-    if (ref->compare(0, 5, "refs/") == 0)
-        localRefFile = cacheDir + "/" + *ref;
-    else
-        localRefFile = cacheDir + "/refs/heads/" + *ref;
-
-    bool doFetch;
-    time_t now = time(0);
-    /* If a rev was specified, we need to fetch if it's not in the
-       repo. */
-    if (rev != "") {
-        try {
-            runProgram("git", true, { "-C", cacheDir, "cat-file", "-e", rev });
-            doFetch = false;
-        } catch (ExecError & e) {
-            if (WIFEXITED(e.status)) {
-                doFetch = true;
-            } else {
-                throw;
-            }
-        }
-    } else {
-        /* If the local ref is older than ‘tarball-ttl’ seconds, do a
-           git fetch to update the local ref to the remote ref. */
-        struct stat st;
-        doFetch = stat(localRefFile.c_str(), &st) != 0 ||
-            (uint64_t) st.st_mtime + settings.tarballTtl <= (uint64_t) now;
-    }
-    if (doFetch)
-    {
-        Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Git repository '%s'", uri));
-
-        // FIXME: git stderr messes up our progress indicator, so
-        // we're using --quiet for now. Should process its stderr.
-        runProgram("git", true, { "-C", cacheDir, "fetch", "--quiet", "--force", "--", uri, fmt("%s:%s", *ref, *ref) });
-
-        struct timeval times[2];
-        times[0].tv_sec = now;
-        times[0].tv_usec = 0;
-        times[1].tv_sec = now;
-        times[1].tv_usec = 0;
-
-        utimes(localRefFile.c_str(), times);
-    }
-
-    // FIXME: check whether rev is an ancestor of ref.
-    GitInfo gitInfo;
-    gitInfo.rev = rev != "" ? rev : chomp(readFile(localRefFile));
-    gitInfo.shortRev = std::string(gitInfo.rev, 0, 7);
-
-    printTalkative("using revision %s of repo '%s'", gitInfo.rev, uri);
-
-    std::string storeLinkName = hashString(htSHA512, name + std::string("\0"s) + gitInfo.rev).to_string(Base32, false);
-    Path storeLink = cacheDir + "/" + storeLinkName + ".link";
-    PathLocks storeLinkLock({storeLink}, fmt("waiting for lock on '%1%'...", storeLink)); // FIXME: broken
-
-    try {
-        auto json = nlohmann::json::parse(readFile(storeLink));
-
-        assert(json["name"] == name && json["rev"] == gitInfo.rev);
-
-        gitInfo.storePath = json["storePath"];
-
-        if (store->isValidPath(store->parseStorePath(gitInfo.storePath))) {
-            gitInfo.revCount = json["revCount"];
-            return gitInfo;
-        }
-
-    } catch (SysError & e) {
-        if (e.errNo != ENOENT) throw;
-    }
-
-    auto source = sinkToSource([&](Sink & sink) {
-        RunOptions gitOptions("git", { "-C", cacheDir, "archive", gitInfo.rev });
-        gitOptions.standardOut = &sink;
-        runProgram2(gitOptions);
-    });
-
-    Path tmpDir = createTempDir();
-    AutoDelete delTmpDir(tmpDir, true);
-
-    unpackTarfile(*source, tmpDir);
-
-    gitInfo.storePath = store->printStorePath(store->addToStore(name, tmpDir));
-
-    gitInfo.revCount = std::stoull(runProgram("git", true, { "-C", cacheDir, "rev-list", "--count", gitInfo.rev }));
-
-    nlohmann::json json;
-    json["storePath"] = gitInfo.storePath;
-    json["uri"] = uri;
-    json["name"] = name;
-    json["rev"] = gitInfo.rev;
-    json["revCount"] = gitInfo.revCount;
-
-    writeFile(storeLink, json.dump());
-
-    return gitInfo;
-}
-
 static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     std::string url;
     std::optional<std::string> ref;
-    std::string rev;
+    std::optional<Hash> rev;
     std::string name = "source";
+    bool fetchSubmodules = false;
     PathSet context;
 
     state.forceValue(*args[0]);
@@ -213,9 +29,11 @@ static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Va
             else if (n == "ref")
                 ref = state.forceStringNoCtx(*attr.value, *attr.pos);
             else if (n == "rev")
-                rev = state.forceStringNoCtx(*attr.value, *attr.pos);
+                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("unsupported argument '%s' to 'fetchGit', at %s", attr.name, *attr.pos);
         }
@@ -230,17 +48,36 @@ static void prim_fetchGit(EvalState & state, const Pos & pos, Value * * args, Va
     // whitelist. Ah well.
     state.checkURI(url);
 
-    auto gitInfo = exportGit(state.store, url, ref, rev, name);
+    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);
-    mkString(*state.allocAttr(v, state.sOutPath), gitInfo.storePath, PathSet({gitInfo.storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), gitInfo.rev);
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), gitInfo.shortRev);
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), gitInfo.revCount);
+    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(state.store->toRealPath(gitInfo.storePath));
+        state.allowedPaths->insert(tree.actualPath);
 }
 
 static RegisterPrimOp r("fetchGit", 1, prim_fetchGit);

src/libexpr/primops/fetchMercurial.cc

@@ -1,174 +1,18 @@
 #include "primops.hh"
 #include "eval-inline.hh"
-#include "download.hh"
 #include "store-api.hh"
-#include "pathlocks.hh"
-
-#include <sys/time.h>
+#include "fetchers.hh"
+#include "url.hh"
 
 #include <regex>
 
-#include <nlohmann/json.hpp>
-
-using namespace std::string_literals;
-
 namespace nix {
 
-struct HgInfo
-{
-    Path storePath;
-    std::string branch;
-    std::string rev;
-    uint64_t revCount = 0;
-};
-
-std::regex commitHashRegex("^[0-9a-fA-F]{40}$");
-
-HgInfo exportMercurial(ref<Store> store, const std::string & uri,
-    std::string rev, const std::string & name)
-{
-    if (evalSettings.pureEval && rev == "")
-        throw Error("in pure evaluation mode, 'fetchMercurial' requires a Mercurial revision");
-
-    if (rev == "" && hasPrefix(uri, "/") && pathExists(uri + "/.hg")) {
-
-        bool clean = runProgram("hg", true, { "status", "-R", uri, "--modified", "--added", "--removed" }) == "";
-
-        if (!clean) {
-
-            /* This is an unclean working tree. So copy all tracked
-               files. */
-
-            printTalkative("copying unclean Mercurial working tree '%s'", uri);
-
-            HgInfo hgInfo;
-            hgInfo.rev = "0000000000000000000000000000000000000000";
-            hgInfo.branch = chomp(runProgram("hg", true, { "branch", "-R", uri }));
-
-            auto files = tokenizeString<std::set<std::string>>(
-                runProgram("hg", true, { "status", "-R", uri, "--clean", "--modified", "--added", "--no-status", "--print0" }), "\0"s);
-
-            PathFilter filter = [&](const Path & p) -> bool {
-                assert(hasPrefix(p, uri));
-                std::string file(p, uri.size() + 1);
-
-                auto st = lstat(p);
-
-                if (S_ISDIR(st.st_mode)) {
-                    auto prefix = file + "/";
-                    auto i = files.lower_bound(prefix);
-                    return i != files.end() && hasPrefix(*i, prefix);
-                }
-
-                return files.count(file);
-            };
-
-            hgInfo.storePath = store->printStorePath(store->addToStore("source", uri, FileIngestionMethod::Recursive, htSHA256, filter));
-
-            return hgInfo;
-        }
-    }
-
-    if (rev == "") rev = "default";
-
-    Path cacheDir = fmt("%s/nix/hg/%s", getCacheDir(), hashString(htSHA256, uri).to_string(Base32, false));
-
-    Path stampFile = fmt("%s/.hg/%s.stamp", cacheDir, hashString(htSHA512, rev).to_string(Base32, false));
-
-    /* If we haven't pulled this repo less than ‘tarball-ttl’ seconds,
-       do so now. */
-    time_t now = time(0);
-    struct stat st;
-    if (stat(stampFile.c_str(), &st) != 0 ||
-        (uint64_t) st.st_mtime + settings.tarballTtl <= (uint64_t) now)
-    {
-        /* Except that if this is a commit hash that we already have,
-           we don't have to pull again. */
-        if (!(std::regex_match(rev, commitHashRegex)
-                && pathExists(cacheDir)
-                && runProgram(
-                    RunOptions("hg", { "log", "-R", cacheDir, "-r", rev, "--template", "1" })
-                    .killStderr(true)).second == "1"))
-        {
-            Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Mercurial repository '%s'", uri));
-
-            if (pathExists(cacheDir)) {
-                try {
-                    runProgram("hg", true, { "pull", "-R", cacheDir, "--", uri });
-                }
-                catch (ExecError & e) {
-                    string transJournal = cacheDir + "/.hg/store/journal";
-                    /* hg throws "abandoned transaction" error only if this file exists */
-                    if (pathExists(transJournal)) {
-                        runProgram("hg", true, { "recover", "-R", cacheDir });
-                        runProgram("hg", true, { "pull", "-R", cacheDir, "--", uri });
-                    } else {
-                        throw ExecError(e.status, fmt("'hg pull' %s", statusToString(e.status)));
-                    }
-                }
-            } else {
-                createDirs(dirOf(cacheDir));
-                runProgram("hg", true, { "clone", "--noupdate", "--", uri, cacheDir });
-            }
-        }
-
-        writeFile(stampFile, "");
-    }
-
-    auto tokens = tokenizeString<std::vector<std::string>>(
-        runProgram("hg", true, { "log", "-R", cacheDir, "-r", rev, "--template", "{node} {rev} {branch}" }));
-    assert(tokens.size() == 3);
-
-    HgInfo hgInfo;
-    hgInfo.rev = tokens[0];
-    hgInfo.revCount = std::stoull(tokens[1]);
-    hgInfo.branch = tokens[2];
-
-    std::string storeLinkName = hashString(htSHA512, name + std::string("\0"s) + hgInfo.rev).to_string(Base32, false);
-    Path storeLink = fmt("%s/.hg/%s.link", cacheDir, storeLinkName);
-
-    try {
-        auto json = nlohmann::json::parse(readFile(storeLink));
-
-        assert(json["name"] == name && json["rev"] == hgInfo.rev);
-
-        hgInfo.storePath = json["storePath"];
-
-        if (store->isValidPath(store->parseStorePath(hgInfo.storePath))) {
-            printTalkative("using cached Mercurial store path '%s'", hgInfo.storePath);
-            return hgInfo;
-        }
-
-    } catch (SysError & e) {
-        if (e.errNo != ENOENT) throw;
-    }
-
-    Path tmpDir = createTempDir();
-    AutoDelete delTmpDir(tmpDir, true);
-
-    runProgram("hg", true, { "archive", "-R", cacheDir, "-r", rev, tmpDir });
-
-    deletePath(tmpDir + "/.hg_archival.txt");
-
-    hgInfo.storePath = store->printStorePath(store->addToStore(name, tmpDir));
-
-    nlohmann::json json;
-    json["storePath"] = hgInfo.storePath;
-    json["uri"] = uri;
-    json["name"] = name;
-    json["branch"] = hgInfo.branch;
-    json["rev"] = hgInfo.rev;
-    json["revCount"] = hgInfo.revCount;
-
-    writeFile(storeLink, json.dump());
-
-    return hgInfo;
-}
-
 static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * args, Value & v)
 {
     std::string url;
-    std::string rev;
+    std::optional<Hash> rev;
+    std::optional<std::string> ref;
     std::string name = "source";
     PathSet context;
 
@@ -182,8 +26,15 @@ static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * ar
             string n(attr.name);
             if (n == "url")
                 url = state.coerceToString(*attr.pos, *attr.value, context, false, false);
-            else if (n == "rev")
-                rev = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else if (n == "rev") {
+                // Ugly: unlike fetchGit, here the "rev" attribute can
+                // be both a revision or a branch/tag name.
+                auto value = state.forceStringNoCtx(*attr.value, *attr.pos);
+                if (std::regex_match(value, revRegex))
+                    rev = Hash(value, htSHA1);
+                else
+                    ref = value;
+            }
             else if (n == "name")
                 name = state.forceStringNoCtx(*attr.value, *attr.pos);
             else
@@ -200,18 +51,35 @@ static void prim_fetchMercurial(EvalState & state, const Pos & pos, Value * * ar
     // whitelist. Ah well.
     state.checkURI(url);
 
-    auto hgInfo = exportMercurial(state.store, url, rev, name);
+    if (evalSettings.pureEval && !rev)
+        throw Error("in pure evaluation mode, 'fetchMercurial' requires a Mercurial revision");
+
+    fetchers::Attrs attrs;
+    attrs.insert_or_assign("type", "hg");
+    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);
+
+    // FIXME: use name
+    auto [tree, input2] = input->fetchTree(state.store);
 
     state.mkAttrs(v, 8);
-    mkString(*state.allocAttr(v, state.sOutPath), hgInfo.storePath, PathSet({hgInfo.storePath}));
-    mkString(*state.allocAttr(v, state.symbols.create("branch")), hgInfo.branch);
-    mkString(*state.allocAttr(v, state.symbols.create("rev")), hgInfo.rev);
-    mkString(*state.allocAttr(v, state.symbols.create("shortRev")), std::string(hgInfo.rev, 0, 12));
-    mkInt(*state.allocAttr(v, state.symbols.create("revCount")), hgInfo.revCount);
+    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());
+    // 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")), std::string(rev2.gitRev(), 0, 12));
+    if (tree.info.revCount)
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
     v.attrs->sort();
 
     if (state.allowedPaths)
-        state.allowedPaths->insert(state.store->toRealPath(hgInfo.storePath));
+        state.allowedPaths->insert(tree.actualPath);
 }
 
 static RegisterPrimOp r("fetchMercurial", 1, prim_fetchMercurial);

src/libexpr/primops/fetchTree.cc

@@ -0,0 +1,165 @@
+#include "primops.hh"
+#include "eval-inline.hh"
+#include "store-api.hh"
+#include "fetchers.hh"
+#include "filetransfer.hh"
+
+#include <ctime>
+#include <iomanip>
+
+namespace nix {
+
+void emitTreeAttrs(
+    EvalState & state,
+    const fetchers::Tree & tree,
+    std::shared_ptr<const fetchers::Input> input,
+    Value & v)
+{
+    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));
+
+    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());
+    }
+
+    if (tree.info.revCount)
+        mkInt(*state.allocAttr(v, state.symbols.create("revCount")), *tree.info.revCount);
+
+    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")));
+
+    v.attrs->sort();
+}
+
+static void prim_fetchTree(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    settings.requireExperimentalFeature("flakes");
+
+    std::shared_ptr<const fetchers::Input> input;
+    PathSet context;
+
+    state.forceValue(*args[0]);
+
+    if (args[0]->type == tAttrs) {
+        state.forceAttrs(*args[0], pos);
+
+        fetchers::Attrs attrs;
+
+        for (auto & attr : *args[0]->attrs) {
+            state.forceValue(*attr.value);
+            if (attr.value->type == tString)
+                attrs.emplace(attr.name, attr.value->string.s);
+            else if (attr.value->type == tBool)
+                attrs.emplace(attr.name, attr.value->boolean);
+            else
+                throw TypeError("fetchTree argument '%s' is %s while a string or Boolean is expected",
+                    attr.name, showType(*attr.value));
+        }
+
+        if (!attrs.count("type"))
+            throw Error("attribute 'type' is missing in call to 'fetchTree', at %s", pos);
+
+        input = fetchers::inputFromAttrs(attrs);
+    } else
+        input = fetchers::inputFromURL(state.coerceToString(pos, *args[0], context, false, false));
+
+    if (evalSettings.pureEval && !input->isImmutable())
+        throw Error("in pure evaluation mode, 'fetchTree' requires an immutable input");
+
+    // FIXME: use fetchOrSubstituteTree
+    auto [tree, input2] = input->fetchTree(state.store);
+
+    if (state.allowedPaths)
+        state.allowedPaths->insert(tree.actualPath);
+
+    emitTreeAttrs(state, tree, input2, v);
+}
+
+static RegisterPrimOp r("fetchTree", 1, prim_fetchTree);
+
+static void fetch(EvalState & state, const Pos & pos, Value * * args, Value & v,
+    const string & who, bool unpack, std::string name)
+{
+    std::optional<std::string> url;
+    std::optional<Hash> expectedHash;
+
+    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.forceStringNoCtx(*attr.value, *attr.pos);
+            else if (n == "sha256")
+                expectedHash = Hash(state.forceStringNoCtx(*attr.value, *attr.pos), htSHA256);
+            else if (n == "name")
+                name = state.forceStringNoCtx(*attr.value, *attr.pos);
+            else
+                throw EvalError("unsupported argument '%s' to '%s', at %s",
+                    attr.name, who, *attr.pos);
+        }
+
+        if (!url)
+            throw EvalError("'url' argument required, at %s", pos);
+
+    } else
+        url = state.forceStringNoCtx(*args[0], pos);
+
+    url = resolveUri(*url);
+
+    state.checkURI(*url);
+
+    if (name == "")
+        name = baseNameOf(*url);
+
+    if (evalSettings.pureEval && !expectedHash)
+        throw Error("in pure evaluation mode, '%s' requires a 'sha256' argument", who);
+
+    auto storePath =
+        unpack
+        ? fetchers::downloadTarball(state.store, *url, name, (bool) expectedHash).storePath
+        : fetchers::downloadFile(state.store, *url, name, (bool) expectedHash).storePath;
+
+    auto path = state.store->toRealPath(storePath);
+
+    if (expectedHash) {
+        auto hash = unpack
+            ? state.store->queryPathInfo(storePath)->narHash
+            : 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(), hash.to_string());
+    }
+
+    if (state.allowedPaths)
+        state.allowedPaths->insert(path);
+
+    mkString(v, path, PathSet({path}));
+}
+
+static void prim_fetchurl(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    fetch(state, pos, args, v, "fetchurl", false, "");
+}
+
+static void prim_fetchTarball(EvalState & state, const Pos & pos, Value * * args, Value & v)
+{
+    fetch(state, pos, args, v, "fetchTarball", true, "source");
+}
+
+static RegisterPrimOp r2("__fetchurl", 1, prim_fetchurl);
+static RegisterPrimOp r3("fetchTarball", 1, prim_fetchTarball);
+
+}

src/libexpr/primops/fromTOML.cc

@@ -1,7 +1,7 @@
 #include "primops.hh"
 #include "eval-inline.hh"
 
-#include "cpptoml/cpptoml.h"
+#include "../../cpptoml/cpptoml.h"
 
 namespace nix {
 

src/libexpr/value.hh

@@ -253,12 +253,17 @@ void mkPath(Value & v, const char * s);
 
 
 #if HAVE_BOEHMGC
-typedef std::vector<Value *, gc_allocator<Value *> > ValueVector;
-typedef std::map<Symbol, Value *, std::less<Symbol>, gc_allocator<std::pair<const Symbol, Value *> > > ValueMap;
+typedef std::vector<Value *, traceable_allocator<Value *> > ValueVector;
+typedef std::map<Symbol, Value *, std::less<Symbol>, traceable_allocator<std::pair<const Symbol, Value *> > > ValueMap;
 #else
 typedef std::vector<Value *> ValueVector;
 typedef std::map<Symbol, Value *> ValueMap;
 #endif
 
 
+/* A value allocated in traceable memory. */
+typedef std::shared_ptr<Value *> RootValue;
+
+RootValue allocRootValue(Value * v);
+
 }

src/libfetchers/attrs.cc

@@ -0,0 +1,107 @@
+#include "attrs.hh"
+#include "fetchers.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+Attrs jsonToAttrs(const nlohmann::json & json)
+{
+    Attrs attrs;
+
+    for (auto & i : json.items()) {
+        if (i.value().is_number())
+            attrs.emplace(i.key(), i.value().get<int64_t>());
+        else if (i.value().is_string())
+            attrs.emplace(i.key(), i.value().get<std::string>());
+        else if (i.value().is_boolean())
+            attrs.emplace(i.key(), i.value().get<bool>());
+        else
+            throw Error("unsupported input attribute type in lock file");
+    }
+
+    return attrs;
+}
+
+nlohmann::json attrsToJson(const Attrs & attrs)
+{
+    nlohmann::json json;
+    for (auto & attr : attrs) {
+        if (auto v = std::get_if<int64_t>(&attr.second)) {
+            json[attr.first] = *v;
+        } else if (auto v = std::get_if<std::string>(&attr.second)) {
+            json[attr.first] = *v;
+        } else if (auto v = std::get_if<Explicit<bool>>(&attr.second)) {
+            json[attr.first] = v->t;
+        } else abort();
+    }
+    return json;
+}
+
+std::optional<std::string> maybeGetStrAttr(const Attrs & attrs, const std::string & name)
+{
+    auto i = attrs.find(name);
+    if (i == attrs.end()) return {};
+    if (auto v = std::get_if<std::string>(&i->second))
+        return *v;
+    throw Error("input attribute '%s' is not a string %s", name, attrsToJson(attrs).dump());
+}
+
+std::string getStrAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetStrAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+std::optional<int64_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))
+        return *v;
+    throw Error("input attribute '%s' is not an integer", name);
+}
+
+int64_t getIntAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetIntAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+std::optional<bool> maybeGetBoolAttr(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))
+        return *v;
+    throw Error("input attribute '%s' is not a Boolean", name);
+}
+
+bool getBoolAttr(const Attrs & attrs, const std::string & name)
+{
+    auto s = maybeGetBoolAttr(attrs, name);
+    if (!s)
+        throw Error("input attribute '%s' is missing", name);
+    return *s;
+}
+
+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)) {
+            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);
+        } else if (auto v = std::get_if<Explicit<bool>>(&attr.second)) {
+            query.insert_or_assign(attr.first, v->t ? "1" : "0");
+        } else abort();
+    }
+    return query;
+}
+
+}

src/libfetchers/attrs.hh

@@ -0,0 +1,39 @@
+#pragma once
+
+#include "types.hh"
+
+#include <variant>
+
+#include <nlohmann/json_fwd.hpp>
+
+namespace nix::fetchers {
+
+/* Wrap bools to prevent string literals (i.e. 'char *') from being
+   cast to a bool in Attr. */
+template<typename T>
+struct Explicit {
+    T t;
+};
+
+typedef std::variant<std::string, int64_t, Explicit<bool>> Attr;
+typedef std::map<std::string, Attr> Attrs;
+
+Attrs jsonToAttrs(const nlohmann::json & json);
+
+nlohmann::json attrsToJson(const Attrs & attrs);
+
+std::optional<std::string> maybeGetStrAttr(const Attrs & attrs, const std::string & name);
+
+std::string getStrAttr(const Attrs & attrs, const std::string & name);
+
+std::optional<int64_t> maybeGetIntAttr(const Attrs & attrs, const std::string & name);
+
+int64_t getIntAttr(const Attrs & attrs, const std::string & name);
+
+std::optional<bool> maybeGetBoolAttr(const Attrs & attrs, const std::string & name);
+
+bool getBoolAttr(const Attrs & attrs, const std::string & name);
+
+std::map<std::string, std::string> attrsToQuery(const Attrs & attrs);
+
+}

src/libfetchers/cache.cc

@@ -0,0 +1,121 @@
+#include "cache.hh"
+#include "sqlite.hh"
+#include "sync.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+static const char * schema = R"sql(
+
+create table if not exists Cache (
+    input     text not null,
+    info      text not null,
+    path      text not null,
+    immutable integer not null,
+    timestamp integer not null,
+    primary key (input)
+);
+)sql";
+
+struct CacheImpl : Cache
+{
+    struct State
+    {
+        SQLite db;
+        SQLiteStmt add, lookup;
+    };
+
+    Sync<State> _state;
+
+    CacheImpl()
+    {
+        auto state(_state.lock());
+
+        auto dbPath = getCacheDir() + "/nix/fetcher-cache-v1.sqlite";
+        createDirs(dirOf(dbPath));
+
+        state->db = SQLite(dbPath);
+        state->db.isCache();
+        state->db.exec(schema);
+
+        state->add.create(state->db,
+            "insert or replace into Cache(input, info, path, immutable, timestamp) values (?, ?, ?, ?, ?)");
+
+        state->lookup.create(state->db,
+            "select info, path, immutable, timestamp from Cache where input = ?");
+    }
+
+    void add(
+        ref<Store> store,
+        const Attrs & inAttrs,
+        const Attrs & infoAttrs,
+        const StorePath & storePath,
+        bool immutable) override
+    {
+        _state.lock()->add.use()
+            (attrsToJson(inAttrs).dump())
+            (attrsToJson(infoAttrs).dump())
+            (store->printStorePath(storePath))
+            (immutable)
+            (time(0)).exec();
+    }
+
+    std::optional<std::pair<Attrs, StorePath>> lookup(
+        ref<Store> store,
+        const Attrs & inAttrs) override
+    {
+        if (auto res = lookupExpired(store, inAttrs)) {
+            if (!res->expired)
+                return std::make_pair(std::move(res->infoAttrs), std::move(res->storePath));
+            debug("ignoring expired cache entry '%s'",
+                attrsToJson(inAttrs).dump());
+        }
+        return {};
+    }
+
+    std::optional<Result> lookupExpired(
+        ref<Store> store,
+        const Attrs & inAttrs) override
+    {
+        auto state(_state.lock());
+
+        auto inAttrsJson = attrsToJson(inAttrs).dump();
+
+        auto stmt(state->lookup.use()(inAttrsJson));
+        if (!stmt.next()) {
+            debug("did not find cache entry for '%s'", inAttrsJson);
+            return {};
+        }
+
+        auto infoJson = stmt.getStr(0);
+        auto storePath = store->parseStorePath(stmt.getStr(1));
+        auto immutable = stmt.getInt(2) != 0;
+        auto timestamp = stmt.getInt(3);
+
+        store->addTempRoot(storePath);
+        if (!store->isValidPath(storePath)) {
+            // FIXME: we could try to substitute 'storePath'.
+            debug("ignoring disappeared cache entry '%s'", inAttrsJson);
+            return {};
+        }
+
+        debug("using cache entry '%s' -> '%s', '%s'",
+            inAttrsJson, infoJson, store->printStorePath(storePath));
+
+        return Result {
+            .expired = !immutable && (settings.tarballTtl.get() == 0 || timestamp + settings.tarballTtl < time(0)),
+            .infoAttrs = jsonToAttrs(nlohmann::json::parse(infoJson)),
+            .storePath = std::move(storePath)
+        };
+    }
+};
+
+ref<Cache> getCache()
+{
+    static auto cache = std::make_shared<CacheImpl>();
+    return ref<Cache>(cache);
+}
+
+}

src/libfetchers/cache.hh

@@ -0,0 +1,34 @@
+#pragma once
+
+#include "fetchers.hh"
+
+namespace nix::fetchers {
+
+struct Cache
+{
+    virtual void add(
+        ref<Store> store,
+        const Attrs & inAttrs,
+        const Attrs & infoAttrs,
+        const StorePath & storePath,
+        bool immutable) = 0;
+
+    virtual std::optional<std::pair<Attrs, StorePath>> lookup(
+        ref<Store> store,
+        const Attrs & inAttrs) = 0;
+
+    struct Result
+    {
+        bool expired = false;
+        Attrs infoAttrs;
+        StorePath storePath;
+    };
+
+    virtual std::optional<Result> lookupExpired(
+        ref<Store> store,
+        const Attrs & inAttrs) = 0;
+};
+
+ref<Cache> getCache();
+
+}

src/libfetchers/fetchers.cc

@@ -0,0 +1,75 @@
+#include "fetchers.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+namespace nix::fetchers {
+
+std::unique_ptr<std::vector<std::unique_ptr<InputScheme>>> inputSchemes = nullptr;
+
+void registerInputScheme(std::unique_ptr<InputScheme> && inputScheme)
+{
+    if (!inputSchemes) inputSchemes = std::make_unique<std::vector<std::unique_ptr<InputScheme>>>();
+    inputSchemes->push_back(std::move(inputScheme));
+}
+
+std::unique_ptr<Input> inputFromURL(const ParsedURL & url)
+{
+    for (auto & inputScheme : *inputSchemes) {
+        auto res = inputScheme->inputFromURL(url);
+        if (res) return res;
+    }
+    throw Error("input '%s' is unsupported", url.url);
+}
+
+std::unique_ptr<Input> inputFromURL(const std::string & url)
+{
+    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);
+        if (res) {
+            if (auto narHash = maybeGetStrAttr(attrs, "narHash"))
+                // FIXME: require SRI hash.
+                res->narHash = Hash(*narHash);
+            return res;
+        }
+    }
+    throw Error("input '%s' is unsupported", attrsToJson(attrs));
+}
+
+Attrs Input::toAttrs() const
+{
+    auto attrs = toAttrsInternal();
+    if (narHash)
+        attrs.emplace("narHash", narHash->to_string(SRI));
+    attrs.emplace("type", type());
+    return attrs;
+}
+
+std::pair<Tree, std::shared_ptr<const Input>> Input::fetchTree(ref<Store> store) const
+{
+    auto [tree, input] = fetchTreeInternal(store);
+
+    if (tree.actualPath == "")
+        tree.actualPath = store->toRealPath(tree.storePath);
+
+    if (!tree.info.narHash)
+        tree.info.narHash = store->queryPathInfo(tree.storePath)->narHash;
+
+    if (input->narHash)
+        assert(input->narHash == tree.info.narHash);
+
+    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), input->narHash->to_string(SRI));
+
+    return {std::move(tree), input};
+}
+
+}

src/libfetchers/fetchers.hh

@@ -0,0 +1,103 @@
+#pragma once
+
+#include "types.hh"
+#include "hash.hh"
+#include "path.hh"
+#include "tree-info.hh"
+#include "attrs.hh"
+#include "url.hh"
+
+#include <memory>
+
+namespace nix { class Store; }
+
+namespace nix::fetchers {
+
+struct Input;
+
+struct Tree
+{
+    Path actualPath;
+    StorePath storePath;
+    TreeInfo info;
+};
+
+struct Input : std::enable_shared_from_this<Input>
+{
+    std::optional<Hash> narHash; // FIXME: implement
+
+    virtual std::string type() const = 0;
+
+    virtual ~Input() { }
+
+    virtual bool operator ==(const Input & other) const { return false; }
+
+    /* Check whether this is a "direct" input, that is, not
+       one that goes through a registry. */
+    virtual bool isDirect() const { return true; }
+
+    /* 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; }
+
+    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();
+    }
+
+    Attrs toAttrs() const;
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTree(ref<Store> store) const;
+
+private:
+
+    virtual std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(ref<Store> store) const = 0;
+
+    virtual Attrs toAttrsInternal() const = 0;
+};
+
+struct InputScheme
+{
+    virtual ~InputScheme() { }
+
+    virtual std::unique_ptr<Input> inputFromURL(const ParsedURL & url) = 0;
+
+    virtual std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) = 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);
+
+struct DownloadFileResult
+{
+    StorePath storePath;
+    std::string etag;
+    std::string effectiveUrl;
+};
+
+DownloadFileResult downloadFile(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable);
+
+Tree downloadTarball(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable);
+
+}

src/libfetchers/git.cc

@@ -0,0 +1,438 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "globals.hh"
+#include "tarfile.hh"
+#include "store-api.hh"
+
+#include <sys/time.h>
+
+using namespace std::string_literals;
+
+namespace nix::fetchers {
+
+static std::string readHead(const Path & path)
+{
+    return chomp(runProgram("git", true, { "-C", path, "rev-parse", "--abbrev-ref", "HEAD" }));
+}
+
+static bool isNotDotGitDirectory(const Path & path)
+{
+    static const std::regex gitDirRegex("^(?:.*/)?\\.git$");
+
+    return not std::regex_match(path, gitDirRegex);
+}
+
+struct GitInput : Input
+{
+    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
+    {
+        auto other2 = dynamic_cast<const GitInput *>(&other);
+        return
+            other2
+            && url == other2->url
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev || narHash;
+    }
+
+    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;
+    }
+
+    std::pair<bool, std::string> getActualUrl() 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
+        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
+    {
+        auto name = "source";
+
+        auto input = std::make_shared<GitInput>(*this);
+
+        assert(!rev || rev->type == htSHA1);
+
+        std::string cacheType = "git";
+        if (shallow) cacheType += "-shallow";
+        if (submodules) cacheType += "-submodules";
+
+        auto getImmutableAttrs = [&]()
+        {
+            return Attrs({
+                {"type", cacheType},
+                {"name", name},
+                {"rev", input->rev->gitRev()},
+            });
+        };
+
+        auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
+            -> std::pair<Tree, std::shared_ptr<const Input>>
+        {
+            assert(input->rev);
+            assert(!rev || rev == input->rev);
+            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"),
+                    },
+                },
+                input
+            };
+        };
+
+        if (rev) {
+            if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+                return makeResult(res->first, std::move(res->second));
+        }
+
+        auto [isLocal, actualUrl_] = getActualUrl();
+        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) {
+            bool clean = false;
+
+            /* Check whether this repo has any commits. There are
+               probably better ways to do this. */
+            auto gitDir = actualUrl + "/.git";
+            auto commonGitDir = chomp(runProgram(
+                "git",
+                true,
+                { "-C", actualUrl, "rev-parse", "--git-common-dir" }
+            ));
+            if (commonGitDir != ".git")
+                    gitDir = commonGitDir;
+
+            bool haveCommits = !readDirectory(gitDir + "/refs/heads").empty();
+
+            try {
+                if (haveCommits) {
+                    runProgram("git", true, { "-C", actualUrl, "diff-index", "--quiet", "HEAD", "--" });
+                    clean = true;
+                }
+            } catch (ExecError & e) {
+                if (!WIFEXITED(e.status) || WEXITSTATUS(e.status) != 1) throw;
+            }
+
+            if (!clean) {
+
+                /* This is an unclean working tree. So copy all tracked files. */
+
+                if (!settings.allowDirty)
+                    throw Error("Git tree '%s' is dirty", actualUrl);
+
+                if (settings.warnDirty)
+                    warn("Git tree '%s' is dirty", actualUrl);
+
+                auto gitOpts = Strings({ "-C", actualUrl, "ls-files", "-z" });
+                if (submodules)
+                    gitOpts.emplace_back("--recurse-submodules");
+
+                auto files = tokenizeString<std::set<std::string>>(
+                    runProgram("git", true, gitOpts), "\0"s);
+
+                PathFilter filter = [&](const Path & p) -> bool {
+                    assert(hasPrefix(p, actualUrl));
+                    std::string file(p, actualUrl.size() + 1);
+
+                    auto st = lstat(p);
+
+                    if (S_ISDIR(st.st_mode)) {
+                        auto prefix = file + "/";
+                        auto i = files.lower_bound(prefix);
+                        return i != files.end() && hasPrefix(*i, prefix);
+                    }
+
+                    return files.count(file);
+                };
+
+                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,
+                    }
+                };
+
+                return {std::move(tree), input};
+            }
+        }
+
+        if (!input->ref) input->ref = isLocal ? readHead(actualUrl) : "master";
+
+        Attrs mutableAttrs({
+            {"type", cacheType},
+            {"name", name},
+            {"url", actualUrl},
+            {"ref", *input->ref},
+        });
+
+        Path repoDir;
+
+        if (isLocal) {
+
+            if (!input->rev)
+                input->rev = Hash(chomp(runProgram("git", true, { "-C", actualUrl, "rev-parse", *input->ref })), htSHA1);
+
+            repoDir = actualUrl;
+
+        } else {
+
+            if (auto res = getCache()->lookup(store, mutableAttrs)) {
+                auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
+                if (!rev || rev == rev2) {
+                    input->rev = rev2;
+                    return makeResult(res->first, std::move(res->second));
+                }
+            }
+
+            Path cacheDir = getCacheDir() + "/nix/gitv3/" + hashString(htSHA256, actualUrl).to_string(Base32, false);
+            repoDir = cacheDir;
+
+            if (!pathExists(cacheDir)) {
+                createDirs(dirOf(cacheDir));
+                runProgram("git", true, { "init", "--bare", repoDir });
+            }
+
+            Path localRefFile =
+                input->ref->compare(0, 5, "refs/") == 0
+                ? cacheDir + "/" + *input->ref
+                : cacheDir + "/refs/heads/" + *input->ref;
+
+            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) {
+                try {
+                    runProgram("git", true, { "-C", repoDir, "cat-file", "-e", input->rev->gitRev() });
+                    doFetch = false;
+                } catch (ExecError & e) {
+                    if (WIFEXITED(e.status)) {
+                        doFetch = true;
+                    } else {
+                        throw;
+                    }
+                }
+            } else {
+                /* If the local ref is older than ‘tarball-ttl’ seconds, do a
+                   git fetch to update the local ref to the remote ref. */
+                struct stat st;
+                doFetch = stat(localRefFile.c_str(), &st) != 0 ||
+                    (uint64_t) st.st_mtime + settings.tarballTtl <= (uint64_t) now;
+            }
+
+            if (doFetch) {
+                Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Git repository '%s'", actualUrl));
+
+                // FIXME: git stderr messes up our progress indicator, so
+                // we're using --quiet for now. Should process its stderr.
+                try {
+                    runProgram("git", true, { "-C", repoDir, "fetch", "--quiet", "--force", "--", actualUrl, fmt("%s:%s", *input->ref, *input->ref) });
+                } catch (Error & e) {
+                    if (!pathExists(localRefFile)) throw;
+                    warn("could not update local clone of Git repository '%s'; continuing with the most recent version", actualUrl);
+                }
+
+                struct timeval times[2];
+                times[0].tv_sec = now;
+                times[0].tv_usec = 0;
+                times[1].tv_sec = now;
+                times[1].tv_usec = 0;
+
+                utimes(localRefFile.c_str(), times);
+            }
+
+            if (!input->rev)
+                input->rev = Hash(chomp(readFile(localRefFile)), htSHA1);
+        }
+
+        bool isShallow = chomp(runProgram("git", true, { "-C", repoDir, "rev-parse", "--is-shallow-repository" })) == "true";
+
+        if (isShallow && !shallow)
+            throw Error("'%s' is a shallow Git repository, but a non-shallow repository is needed", actualUrl);
+
+        // FIXME: check whether rev is an ancestor of ref.
+
+        printTalkative("using revision %s of repo '%s'", input->rev->gitRev(), actualUrl);
+
+        /* Now that we know the ref, check again whether we have it in
+           the store. */
+        if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+            return makeResult(res->first, std::move(res->second));
+
+        Path tmpDir = createTempDir();
+        AutoDelete delTmpDir(tmpDir, true);
+        PathFilter filter = defaultPathFilter;
+
+        if (submodules) {
+            Path tmpGitDir = createTempDir();
+            AutoDelete delTmpGitDir(tmpGitDir, true);
+
+            runProgram("git", true, { "init", tmpDir, "--separate-git-dir", tmpGitDir });
+            // TODO: repoDir might lack the ref (it only checks if rev
+            // exists, see FIXME above) so use a big hammer and fetch
+            // everything to ensure we get the rev.
+            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, "remote", "add", "origin", actualUrl });
+            runProgram("git", true, { "-C", tmpDir, "submodule", "--quiet", "update", "--init", "--recursive" });
+
+            filter = isNotDotGitDirectory;
+        } else {
+            // 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() });
+                gitOptions.standardOut = &sink;
+                runProgram2(gitOptions);
+            });
+
+            unpackTarfile(*source, tmpDir);
+        }
+
+        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() }));
+
+        Attrs infoAttrs({
+            {"rev", input->rev->gitRev()},
+            {"lastModified", lastModified},
+        });
+
+        if (!shallow)
+            infoAttrs.insert_or_assign("revCount",
+                std::stoull(runProgram("git", true, { "-C", repoDir, "rev-list", "--count", input->rev->gitRev() })));
+
+        if (!this->rev)
+            getCache()->add(
+                store,
+                mutableAttrs,
+                infoAttrs,
+                storePath,
+                false);
+
+        getCache()->add(
+            store,
+            getImmutableAttrs(),
+            infoAttrs,
+            storePath,
+            true);
+
+        return makeResult(infoAttrs, std::move(storePath));
+    }
+};
+
+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_match(*ref, refRegex))
+                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

@@ -0,0 +1,195 @@
+#include "filetransfer.hh"
+#include "cache.hh"
+#include "fetchers.hh"
+#include "globals.hh"
+#include "store-api.hh"
+
+#include <nlohmann/json.hpp>
+
+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);
+
+struct GitHubInput : Input
+{
+    std::string owner;
+    std::string repo;
+    std::optional<std::string> ref;
+    std::optional<Hash> rev;
+
+    std::string type() const override { return "github"; }
+
+    bool operator ==(const Input & other) const override
+    {
+        auto other2 = dynamic_cast<const GitHubInput *>(&other);
+        return
+            other2
+            && owner == other2->owner
+            && repo == other2->repo
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev || narHash;
+    }
+
+    std::optional<std::string> getRef() const override { return ref; }
+
+    std::optional<Hash> getRev() const override { return rev; }
+
+    ParsedURL toURL() const override
+    {
+        auto path = owner + "/" + repo;
+        assert(!(ref && rev));
+        if (ref) path += "/" + *ref;
+        if (rev) path += "/" + rev->to_string(Base16, false);
+        return ParsedURL {
+            .scheme = "github",
+            .path = path,
+        };
+    }
+
+    Attrs toAttrsInternal() const 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;
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const 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(json["sha"], htSHA1);
+            debug("HEAD revision for '%s' is %s", url, rev->gitRev());
+        }
+
+        auto input = std::make_shared<GitHubInput>(*this);
+        input->ref = {};
+        input->rev = *rev;
+
+        Attrs immutableAttrs({
+            {"type", "git-tarball"},
+            {"rev", rev->gitRev()},
+        });
+
+        if (auto res = getCache()->lookup(store, immutableAttrs)) {
+            return {
+                Tree{
+                    .actualPath = store->toRealPath(res->second),
+                    .storePath = std::move(res->second),
+                    .info = TreeInfo {
+                        .lastModified = getIntAttr(res->first, "lastModified"),
+                    },
+                },
+                input
+            };
+        }
+
+        // FIXME: use regular /archive URLs instead? api.github.com
+        // might have stricter rate limits.
+
+        auto url = fmt("https://api.github.com/repos/%s/%s/tarball/%s",
+            owner, repo, rev->to_string(Base16, false));
+
+        std::string accessToken = settings.githubAccessToken.get();
+        if (accessToken != "")
+            url += "?access_token=" + accessToken;
+
+        auto tree = downloadTarball(store, url, "source", true);
+
+        getCache()->add(
+            store,
+            immutableAttrs,
+            {
+                {"rev", rev->gitRev()},
+                {"lastModified", *tree.info.lastModified}
+            },
+            tree.storePath,
+            true);
+
+        return {std::move(tree), input};
+    }
+};
+
+struct GitHubInputScheme : InputScheme
+{
+    std::unique_ptr<Input> inputFromURL(const ParsedURL & url) 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;
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "github") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "owner" && name != "repo" && name != "ref" && name != "rev")
+                throw Error("unsupported GitHub input attribute '%s'", name);
+
+        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;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<GitHubInputScheme>()); });
+
+}

src/libfetchers/local.mk

@@ -0,0 +1,11 @@
+libraries += libfetchers
+
+libfetchers_NAME = libnixfetchers
+
+libfetchers_DIR := $(d)
+
+libfetchers_SOURCES := $(wildcard $(d)/*.cc)
+
+libfetchers_CXXFLAGS += -I src/libutil -I src/libstore
+
+libfetchers_LIBS = libutil libstore libnixrust

src/libfetchers/mercurial.cc

@@ -0,0 +1,303 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "globals.hh"
+#include "tarfile.hh"
+#include "store-api.hh"
+
+#include <sys/time.h>
+
+using namespace std::string_literals;
+
+namespace nix::fetchers {
+
+struct MercurialInput : Input
+{
+    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
+    {
+        auto other2 = dynamic_cast<const MercurialInput *>(&other);
+        return
+            other2
+            && url == other2->url
+            && rev == other2->rev
+            && ref == other2->ref;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) rev || narHash;
+    }
+
+    std::optional<std::string> getRef() const override { return ref; }
+
+    std::optional<Hash> getRev() const override { return rev; }
+
+    ParsedURL toURL() const 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);
+        return url;
+    }
+
+    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());
+        return attrs;
+    }
+
+    std::pair<bool, std::string> getActualUrl() const
+    {
+        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
+    {
+        auto name = "source";
+
+        auto input = std::make_shared<MercurialInput>(*this);
+
+        auto [isLocal, actualUrl_] = getActualUrl();
+        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")) {
+
+            bool clean = runProgram("hg", true, { "status", "-R", actualUrl, "--modified", "--added", "--removed" }) == "";
+
+            if (!clean) {
+
+                /* This is an unclean working tree. So copy all tracked
+                   files. */
+
+                if (!settings.allowDirty)
+                    throw Error("Mercurial tree '%s' is unclean", actualUrl);
+
+                if (settings.warnDirty)
+                    warn("Mercurial tree '%s' is unclean", actualUrl);
+
+                input->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);
+
+                PathFilter filter = [&](const Path & p) -> bool {
+                    assert(hasPrefix(p, actualUrl));
+                    std::string file(p, actualUrl.size() + 1);
+
+                    auto st = lstat(p);
+
+                    if (S_ISDIR(st.st_mode)) {
+                        auto prefix = file + "/";
+                        auto i = files.lower_bound(prefix);
+                        return i != files.end() && hasPrefix(*i, prefix);
+                    }
+
+                    return files.count(file);
+                };
+
+                auto storePath = store->addToStore("source", actualUrl, FileIngestionMethod::Recursive, htSHA256, filter);
+
+                return {Tree {
+                    .actualPath = store->printStorePath(storePath),
+                    .storePath = std::move(storePath),
+                }, input};
+            }
+        }
+
+        if (!input->ref) input->ref = "default";
+
+        auto getImmutableAttrs = [&]()
+        {
+            return Attrs({
+                {"type", "hg"},
+                {"name", name},
+                {"rev", input->rev->gitRev()},
+            });
+        };
+
+        auto makeResult = [&](const Attrs & infoAttrs, StorePath && storePath)
+            -> std::pair<Tree, std::shared_ptr<const Input>>
+        {
+            assert(input->rev);
+            assert(!rev || rev == input->rev);
+            return {
+                Tree{
+                    .actualPath = store->toRealPath(storePath),
+                    .storePath = std::move(storePath),
+                    .info = TreeInfo {
+                        .revCount = getIntAttr(infoAttrs, "revCount"),
+                    },
+                },
+                input
+            };
+        };
+
+        if (input->rev) {
+            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;
+
+        Attrs mutableAttrs({
+            {"type", "hg"},
+            {"name", name},
+            {"url", actualUrl},
+            {"ref", *input->ref},
+        });
+
+        if (auto res = getCache()->lookup(store, mutableAttrs)) {
+            auto rev2 = Hash(getStrAttr(res->first, "rev"), htSHA1);
+            if (!rev || rev == rev2) {
+                input->rev = rev2;
+                return makeResult(res->first, std::move(res->second));
+            }
+        }
+
+        Path cacheDir = fmt("%s/nix/hg/%s", getCacheDir(), hashString(htSHA256, actualUrl).to_string(Base32, false));
+
+        /* If this is a commit hash that we already have, we don't
+           have to pull again. */
+        if (!(input->rev
+                && pathExists(cacheDir)
+                && runProgram(
+                    RunOptions("hg", { "log", "-R", cacheDir, "-r", input->rev->gitRev(), "--template", "1" })
+                    .killStderr(true)).second == "1"))
+        {
+            Activity act(*logger, lvlTalkative, actUnknown, fmt("fetching Mercurial repository '%s'", actualUrl));
+
+            if (pathExists(cacheDir)) {
+                try {
+                    runProgram("hg", true, { "pull", "-R", cacheDir, "--", actualUrl });
+                }
+                catch (ExecError & e) {
+                    string transJournal = cacheDir + "/.hg/store/journal";
+                    /* hg throws "abandoned transaction" error only if this file exists */
+                    if (pathExists(transJournal)) {
+                        runProgram("hg", true, { "recover", "-R", cacheDir });
+                        runProgram("hg", true, { "pull", "-R", cacheDir, "--", actualUrl });
+                    } else {
+                        throw ExecError(e.status, fmt("'hg pull' %s", statusToString(e.status)));
+                    }
+                }
+            } else {
+                createDirs(dirOf(cacheDir));
+                runProgram("hg", true, { "clone", "--noupdate", "--", actualUrl, cacheDir });
+            }
+        }
+
+        auto tokens = tokenizeString<std::vector<std::string>>(
+            runProgram("hg", true, { "log", "-R", cacheDir, "-r", revOrRef, "--template", "{node} {rev} {branch}" }));
+        assert(tokens.size() == 3);
+
+        input->rev = Hash(tokens[0], htSHA1);
+        auto revCount = std::stoull(tokens[1]);
+        input->ref = tokens[2];
+
+        if (auto res = getCache()->lookup(store, getImmutableAttrs()))
+            return makeResult(res->first, std::move(res->second));
+
+        Path tmpDir = createTempDir();
+        AutoDelete delTmpDir(tmpDir, true);
+
+        runProgram("hg", true, { "archive", "-R", cacheDir, "-r", input->rev->gitRev(), tmpDir });
+
+        deletePath(tmpDir + "/.hg_archival.txt");
+
+        auto storePath = store->addToStore(name, tmpDir);
+
+        Attrs infoAttrs({
+            {"rev", input->rev->gitRev()},
+            {"revCount", (int64_t) revCount},
+        });
+
+        if (!this->rev)
+            getCache()->add(
+                store,
+                mutableAttrs,
+                infoAttrs,
+                storePath,
+                false);
+
+        getCache()->add(
+            store,
+            getImmutableAttrs(),
+            infoAttrs,
+            storePath,
+            true);
+
+        return makeResult(infoAttrs, std::move(storePath));
+    }
+};
+
+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

@@ -0,0 +1,148 @@
+#include "fetchers.hh"
+#include "store-api.hh"
+
+namespace nix::fetchers {
+
+struct PathInput : Input
+{
+    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
+    {
+        auto other2 = dynamic_cast<const PathInput *>(&other);
+        return
+            other2
+            && path == other2->path
+            && rev == other2->rev
+            && revCount == other2->revCount
+            && lastModified == other2->lastModified;
+    }
+
+    bool isImmutable() const override
+    {
+        return (bool) narHash;
+    }
+
+    ParsedURL toURL() const override
+    {
+        auto query = attrsToQuery(toAttrsInternal());
+        query.erase("path");
+        return ParsedURL {
+            .scheme = "path",
+            .path = path,
+            .query = query,
+        };
+    }
+
+    Attrs toAttrsInternal() const 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;
+    }
+
+    std::pair<Tree, std::shared_ptr<const Input>> fetchTreeInternal(nix::ref<Store> store) const override
+    {
+        auto input = std::make_shared<PathInput>(*this);
+
+        // FIXME: check whether access to 'path' is allowed.
+
+        auto storePath = store->maybeParseStorePath(path);
+
+        if (storePath)
+            store->addTempRoot(*storePath);
+
+        if (!storePath || storePath->name() != "source" || !store->isValidPath(*storePath))
+            // 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;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<PathInputScheme>()); });
+
+}

src/libfetchers/tarball.cc

@@ -0,0 +1,275 @@
+#include "fetchers.hh"
+#include "cache.hh"
+#include "filetransfer.hh"
+#include "globals.hh"
+#include "store-api.hh"
+#include "archive.hh"
+#include "tarfile.hh"
+
+namespace nix::fetchers {
+
+DownloadFileResult downloadFile(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable)
+{
+    // FIXME: check store
+
+    Attrs inAttrs({
+        {"type", "file"},
+        {"url", url},
+        {"name", name},
+    });
+
+    auto cached = getCache()->lookupExpired(store, inAttrs);
+
+    auto useCached = [&]() -> DownloadFileResult
+    {
+        return {
+            .storePath = std::move(cached->storePath),
+            .etag = getStrAttr(cached->infoAttrs, "etag"),
+            .effectiveUrl = getStrAttr(cached->infoAttrs, "url")
+        };
+    };
+
+    if (cached && !cached->expired)
+        return useCached();
+
+    FileTransferRequest request(url);
+    if (cached)
+        request.expectedETag = getStrAttr(cached->infoAttrs, "etag");
+    FileTransferResult res;
+    try {
+        res = getFileTransfer()->download(request);
+    } catch (FileTransferError & e) {
+        if (cached) {
+            warn("%s; using cached version", e.msg());
+            return useCached();
+        } else
+            throw;
+    }
+
+    // FIXME: write to temporary file.
+
+    Attrs infoAttrs({
+        {"etag", res.etag},
+        {"url", res.effectiveUri},
+    });
+
+    std::optional<StorePath> storePath;
+
+    if (res.cached) {
+        assert(cached);
+        assert(request.expectedETag == res.etag);
+        storePath = std::move(cached->storePath);
+    } else {
+        StringSink sink;
+        dumpString(*res.data, sink);
+        auto hash = hashString(htSHA256, *res.data);
+        ValidPathInfo info(store->makeFixedOutputPath(FileIngestionMethod::Flat, hash, name));
+        info.narHash = hashString(htSHA256, *sink.s);
+        info.narSize = sink.s->size();
+        info.ca = makeFixedOutputCA(FileIngestionMethod::Flat, hash);
+        store->addToStore(info, sink.s, NoRepair, NoCheckSigs);
+        storePath = std::move(info.path);
+    }
+
+    getCache()->add(
+        store,
+        inAttrs,
+        infoAttrs,
+        *storePath,
+        immutable);
+
+    if (url != res.effectiveUri)
+        getCache()->add(
+            store,
+            {
+                {"type", "file"},
+                {"url", res.effectiveUri},
+                {"name", name},
+            },
+            infoAttrs,
+            *storePath,
+            immutable);
+
+    return {
+        .storePath = std::move(*storePath),
+        .etag = res.etag,
+        .effectiveUrl = res.effectiveUri,
+    };
+}
+
+Tree downloadTarball(
+    ref<Store> store,
+    const std::string & url,
+    const std::string & name,
+    bool immutable)
+{
+    Attrs inAttrs({
+        {"type", "tarball"},
+        {"url", url},
+        {"name", name},
+    });
+
+    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"),
+            },
+        };
+
+    auto res = downloadFile(store, url, name, immutable);
+
+    std::optional<StorePath> unpackedStorePath;
+    time_t lastModified;
+
+    if (cached && res.etag != "" && getStrAttr(cached->infoAttrs, "etag") == res.etag) {
+        unpackedStorePath = std::move(cached->storePath);
+        lastModified = getIntAttr(cached->infoAttrs, "lastModified");
+    } else {
+        Path tmpDir = createTempDir();
+        AutoDelete autoDelete(tmpDir, true);
+        unpackTarfile(store->toRealPath(res.storePath), tmpDir);
+        auto members = readDirectory(tmpDir);
+        if (members.size() != 1)
+            throw nix::Error("tarball '%s' contains an unexpected number of top-level files", url);
+        auto topDir = tmpDir + "/" + members.begin()->name;
+        lastModified = lstat(topDir).st_mtime;
+        unpackedStorePath = store->addToStore(name, topDir, FileIngestionMethod::Recursive, htSHA256, defaultPathFilter, NoRepair);
+    }
+
+    Attrs infoAttrs({
+        {"lastModified", lastModified},
+        {"etag", res.etag},
+    });
+
+    getCache()->add(
+        store,
+        inAttrs,
+        infoAttrs,
+        *unpackedStorePath,
+        immutable);
+
+    return Tree {
+        .actualPath = store->toRealPath(*unpackedStorePath),
+        .storePath = std::move(*unpackedStorePath),
+        .info = TreeInfo {
+            .lastModified = 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));
+        else if (hash)
+            url2.query.insert_or_assign("hash", hash->to_string(SRI));
+        return url2;
+    }
+
+    Attrs toAttrsInternal() const override
+    {
+        Attrs attrs;
+        attrs.emplace("url", url.to_string());
+        if (hash)
+            attrs.emplace("hash", hash->to_string(SRI));
+        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
+    {
+        if (url.scheme != "file" && url.scheme != "http" && url.scheme != "https") return nullptr;
+
+        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;
+    }
+
+    std::unique_ptr<Input> inputFromAttrs(const Attrs & attrs) override
+    {
+        if (maybeGetStrAttr(attrs, "type") != "tarball") return {};
+
+        for (auto & [name, value] : attrs)
+            if (name != "type" && name != "url" && name != "hash")
+                throw Error("unsupported tarball input attribute '%s'", name);
+
+        auto input = std::make_unique<TarballInput>(parseURL(getStrAttr(attrs, "url")));
+        if (auto hash = maybeGetStrAttr(attrs, "hash"))
+            // FIXME: require SRI hash.
+            input->hash = Hash(*hash);
+
+        return input;
+    }
+};
+
+static auto r1 = OnStartup([] { registerInputScheme(std::make_unique<TarballInputScheme>()); });
+
+}

src/libfetchers/tree-info.cc

@@ -0,0 +1,14 @@
+#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

@@ -0,0 +1,29 @@
+#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

@@ -6,43 +6,47 @@ namespace nix {
 MixCommonArgs::MixCommonArgs(const string & programName)
     : programName(programName)
 {
-    mkFlag()
-        .longName("verbose")
-        .shortName('v')
-        .description("increase verbosity level")
-        .handler([]() { verbosity = (Verbosity) (verbosity + 1); });
+    addFlag({
+        .longName = "verbose",
+        .shortName = 'v',
+        .description = "increase verbosity level",
+        .handler = {[]() { verbosity = (Verbosity) (verbosity + 1); }},
+    });
 
-    mkFlag()
-        .longName("quiet")
-        .description("decrease verbosity level")
-        .handler([]() { verbosity = verbosity > lvlError ? (Verbosity) (verbosity - 1) : lvlError; });
+    addFlag({
+        .longName = "quiet",
+        .description = "decrease verbosity level",
+        .handler = {[]() { verbosity = verbosity > lvlError ? (Verbosity) (verbosity - 1) : lvlError; }},
+    });
 
-    mkFlag()
-        .longName("debug")
-        .description("enable debug output")
-        .handler([]() { verbosity = lvlDebug; });
+    addFlag({
+        .longName = "debug",
+        .description = "enable debug output",
+        .handler = {[]() { verbosity = lvlDebug; }},
+    });
 
-    mkFlag()
-        .longName("option")
-        .labels({"name", "value"})
-        .description("set a Nix configuration option (overriding nix.conf)")
-        .arity(2)
-        .handler([](std::vector<std::string> ss) {
+    addFlag({
+        .longName = "option",
+        .description = "set a Nix configuration option (overriding nix.conf)",
+        .labels = {"name", "value"},
+        .handler = {[](std::string name, std::string value) {
             try {
-                globalConfig.set(ss[0], ss[1]);
+                globalConfig.set(name, value);
             } catch (UsageError & e) {
                 warn(e.what());
             }
-        });
+        }},
+    });
 
-    mkFlag()
-        .longName("max-jobs")
-        .shortName('j')
-        .label("jobs")
-        .description("maximum number of parallel builds")
-        .handler([=](std::string s) {
+    addFlag({
+        .longName = "max-jobs",
+        .shortName = 'j',
+        .description = "maximum number of parallel builds",
+        .labels = Strings{"jobs"},
+        .handler = {[=](std::string s) {
             settings.set("max-jobs", s);
-        });
+        }}
+    });
 
     std::string cat = "config";
     globalConfig.convertToArgs(*this, cat);

src/libmain/local.mk

@@ -6,6 +6,8 @@ libmain_DIR := $(d)
 
 libmain_SOURCES := $(wildcard $(d)/*.cc)
 
+libmain_CXXFLAGS += -I src/libutil -I src/libstore
+
 libmain_LDFLAGS = $(OPENSSL_LIBS)
 
 libmain_LIBS = libstore libutil

src/libmain/shared.cc

@@ -155,7 +155,7 @@ void initNix()
        sshd). This breaks build users because they don't have access
        to the TMPDIR, in particular in ‘nix-store --serve’. */
 #if __APPLE__
-    if (getuid() == 0 && hasPrefix(getEnv("TMPDIR").value_or("/tmp"), "/var/folders/"))
+    if (hasPrefix(getEnv("TMPDIR").value_or("/tmp"), "/var/folders/"))
         unsetenv("TMPDIR");
 #endif
 }
@@ -165,28 +165,32 @@ LegacyArgs::LegacyArgs(const std::string & programName,
     std::function<bool(Strings::iterator & arg, const Strings::iterator & end)> parseArg)
     : MixCommonArgs(programName), parseArg(parseArg)
 {
-    mkFlag()
-        .longName("no-build-output")
-        .shortName('Q')
-        .description("do not show build output")
-        .set(&settings.verboseBuild, false);
+    addFlag({
+        .longName = "no-build-output",
+        .shortName = 'Q',
+        .description = "do not show build output",
+        .handler = {&settings.verboseBuild, false},
+    });
 
-    mkFlag()
-        .longName("keep-failed")
-        .shortName('K')
-        .description("keep temporary directories of failed builds")
-        .set(&(bool&) settings.keepFailed, true);
+    addFlag({
+        .longName = "keep-failed",
+        .shortName ='K',
+        .description = "keep temporary directories of failed builds",
+        .handler = {&(bool&) settings.keepFailed, true},
+    });
 
-    mkFlag()
-        .longName("keep-going")
-        .shortName('k')
-        .description("keep going after a build fails")
-        .set(&(bool&) settings.keepGoing, true);
+    addFlag({
+        .longName = "keep-going",
+        .shortName ='k',
+        .description = "keep going after a build fails",
+        .handler = {&(bool&) settings.keepGoing, true},
+    });
 
-    mkFlag()
-        .longName("fallback")
-        .description("build from source if substitution fails")
-        .set(&(bool&) settings.tryFallback, true);
+    addFlag({
+        .longName = "fallback",
+        .description = "build from source if substitution fails",
+        .handler = {&(bool&) settings.tryFallback, true},
+    });
 
     auto intSettingAlias = [&](char shortName, const std::string & longName,
         const std::string & description, const std::string & dest) {
@@ -205,11 +209,12 @@ LegacyArgs::LegacyArgs(const std::string & programName,
     mkFlag(0, "no-gc-warning", "disable warning about not using '--add-root'",
         &gcWarning, false);
 
-    mkFlag()
-        .longName("store")
-        .label("store-uri")
-        .description("URI of the Nix store to use")
-        .dest(&(std::string&) settings.storeUri);
+    addFlag({
+        .longName = "store",
+        .description = "URI of the Nix store to use",
+        .labels = {"store-uri"},
+        .handler = {&(std::string&) settings.storeUri},
+    });
 }
 
 
@@ -260,7 +265,10 @@ void printVersion(const string & programName)
         cfg.push_back("signed-caches");
 #endif
         std::cout << "Features: " << concatStringsSep(", ", cfg) << "\n";
-        std::cout << "Configuration file: " << settings.nixConfDir + "/nix.conf" << "\n";
+        std::cout << "System configuration file: " << settings.nixConfDir + "/nix.conf" << "\n";
+        std::cout << "User configuration files: " <<
+            concatStringsSep(":", settings.nixUserConfFiles)
+            << "\n";
         std::cout << "Store directory: " << settings.nixStore << "\n";
         std::cout << "State directory: " << settings.nixStateDir << "\n";
     }

src/libstore/build.cc

@@ -7,7 +7,7 @@
 #include "affinity.hh"
 #include "builtins.hh"
 #include "builtins/buildenv.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "finally.hh"
 #include "compression.hh"
 #include "json.hh"
@@ -33,7 +33,6 @@
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <sys/utsname.h>
-#include <sys/select.h>
 #include <sys/resource.h>
 #include <sys/socket.h>
 #include <sys/un.h>
@@ -43,6 +42,7 @@
 #include <errno.h>
 #include <cstring>
 #include <termios.h>
+#include <poll.h>
 
 #include <pwd.h>
 #include <grp.h>
@@ -361,7 +361,7 @@ public:
     {
         actDerivations.progress(doneBuilds, expectedBuilds + doneBuilds, runningBuilds, failedBuilds);
         actSubstitutions.progress(doneSubstitutions, expectedSubstitutions + doneSubstitutions, runningSubstitutions, failedSubstitutions);
-        act.setExpected(actDownload, expectedDownloadSize + doneDownloadSize);
+        act.setExpected(actFileTransfer, expectedDownloadSize + doneDownloadSize);
         act.setExpected(actCopyPath, expectedNarSize + doneNarSize);
     }
 };
@@ -507,9 +507,10 @@ private:
     Path fnUserLock;
     AutoCloseFD fdUserLock;
 
+    bool isEnabled = false;
     string user;
-    uid_t uid;
-    gid_t gid;
+    uid_t uid = 0;
+    gid_t gid = 0;
     std::vector<gid_t> supplementaryGIDs;
 
 public:
@@ -522,7 +523,9 @@ public:
     uid_t getGID() { assert(gid); return gid; }
     std::vector<gid_t> getSupplementaryGIDs() { return supplementaryGIDs; }
 
-    bool enabled() { return uid != 0; }
+    bool findFreeUser();
+
+    bool enabled() { return isEnabled; }
 
 };
 
@@ -530,6 +533,11 @@ public:
 UserLock::UserLock()
 {
     assert(settings.buildUsersGroup != "");
+    createDirs(settings.nixStateDir + "/userpool");
+}
+
+bool UserLock::findFreeUser() {
+    if (enabled()) return true;
 
     /* Get the members of the build-users-group. */
     struct group * gr = getgrnam(settings.buildUsersGroup.get().c_str());
@@ -559,7 +567,6 @@ UserLock::UserLock()
             throw Error(format("the user '%1%' in the group '%2%' does not exist")
                 % i % settings.buildUsersGroup);
 
-        createDirs(settings.nixStateDir + "/userpool");
 
         fnUserLock = (format("%1%/userpool/%2%") % settings.nixStateDir % pw->pw_uid).str();
 
@@ -590,16 +597,13 @@ UserLock::UserLock()
             supplementaryGIDs.resize(ngroups);
 #endif
 
-            return;
+            isEnabled = true;
+            return true;
         }
     }
-
-    throw Error(format("all build users are currently in use; "
-        "consider creating additional users and adding them to the '%1%' group")
-        % settings.buildUsersGroup);
+    return false;
 }
 
-
 void UserLock::kill()
 {
     killUser(uid);
@@ -928,6 +932,7 @@ private:
     void closureRepaired();
     void inputsRealised();
     void tryToBuild();
+    void tryLocalBuild();
     void buildDone();
 
     /* Is the build hook willing to perform the build? */
@@ -999,6 +1004,8 @@ private:
         Goal::amDone(result);
     }
 
+    void started();
+
     void done(BuildResult::Status status, const string & msg = "");
 
     StorePathSet exportReferences(const StorePathSet & storePaths);
@@ -1386,6 +1393,19 @@ void DerivationGoal::inputsRealised()
     result = BuildResult();
 }
 
+void DerivationGoal::started() {
+    auto msg = fmt(
+        buildMode == bmRepair ? "repairing outputs of '%s'" :
+        buildMode == bmCheck ? "checking outputs of '%s'" :
+        nrRounds > 1 ? "building '%s' (round %d/%d)" :
+        "building '%s'", worker.store.printStorePath(drvPath), curRound, nrRounds);
+    fmt("building '%s'", worker.store.printStorePath(drvPath));
+    if (hook) msg += fmt(" on '%s'", machineName);
+    act = std::make_unique<Activity>(*logger, lvlInfo, actBuild, msg,
+        Logger::Fields{worker.store.printStorePath(drvPath), hook ? machineName : "", curRound, nrRounds});
+    mcRunningBuilds = std::make_unique<MaintainCount<uint64_t>>(worker.runningBuilds);
+    worker.updateProgress();
+}
 
 void DerivationGoal::tryToBuild()
 {
@@ -1437,20 +1457,6 @@ void DerivationGoal::tryToBuild()
        supported for local builds. */
     bool buildLocally = buildMode != bmNormal || parsedDrv->willBuildLocally();
 
-    auto started = [&]() {
-        auto msg = fmt(
-            buildMode == bmRepair ? "repairing outputs of '%s'" :
-            buildMode == bmCheck ? "checking outputs of '%s'" :
-            nrRounds > 1 ? "building '%s' (round %d/%d)" :
-            "building '%s'", worker.store.printStorePath(drvPath), curRound, nrRounds);
-        fmt("building '%s'", worker.store.printStorePath(drvPath));
-        if (hook) msg += fmt(" on '%s'", machineName);
-        act = std::make_unique<Activity>(*logger, lvlInfo, actBuild, msg,
-            Logger::Fields{worker.store.printStorePath(drvPath), hook ? machineName : "", curRound, nrRounds});
-        mcRunningBuilds = std::make_unique<MaintainCount<uint64_t>>(worker.runningBuilds);
-        worker.updateProgress();
-    };
-
     /* Is the build hook willing to accept this job? */
     if (!buildLocally) {
         switch (tryBuildHook()) {
@@ -1483,6 +1489,34 @@ void DerivationGoal::tryToBuild()
         return;
     }
 
+    state = &DerivationGoal::tryLocalBuild;
+    worker.wakeUp(shared_from_this());
+}
+
+void DerivationGoal::tryLocalBuild() {
+
+    /* If `build-users-group' is not empty, then we have to build as
+       one of the members of that group. */
+    if (settings.buildUsersGroup != "" && getuid() == 0) {
+#if defined(__linux__) || defined(__APPLE__)
+        if (!buildUser) buildUser = std::make_unique<UserLock>();
+
+        if (buildUser->findFreeUser()) {
+            /* Make sure that no other processes are executing under this
+               uid. */
+            buildUser->kill();
+        } else {
+            debug("waiting for build users");
+            worker.waitForAWhile(shared_from_this());
+            return;
+        }
+#else
+        /* Don't know how to block the creation of setuid/setgid
+           binaries on this platform. */
+        throw Error("build users are not supported on this platform for security reasons");
+#endif
+    }
+
     try {
 
         /* Okay, we have to build. */
@@ -1680,6 +1714,7 @@ void DerivationGoal::buildDone()
         }
 
         if (buildMode == bmCheck) {
+            deleteTmpDir(true);
             done(BuildResult::Built);
             return;
         }
@@ -1942,22 +1977,6 @@ void DerivationGoal::startBuilder()
         #endif
     }
 
-    /* If `build-users-group' is not empty, then we have to build as
-       one of the members of that group. */
-    if (settings.buildUsersGroup != "" && getuid() == 0) {
-#if defined(__linux__) || defined(__APPLE__)
-        buildUser = std::make_unique<UserLock>();
-
-        /* Make sure that no other processes are executing under this
-           uid. */
-        buildUser->kill();
-#else
-        /* Don't know how to block the creation of setuid/setgid
-           binaries on this platform. */
-        throw Error("build users are not supported on this platform for security reasons");
-#endif
-    }
-
     /* Create a temporary directory where the build will take
        place. */
     tmpDir = createTempDir("", "nix-build-" + std::string(drvPath.name()), false, false, 0700);
@@ -2161,7 +2180,7 @@ void DerivationGoal::startBuilder()
     if (needsHashRewrite()) {
 
         if (pathExists(homeDir))
-            throw Error(format("directory '%1%' exists; please remove it") % homeDir);
+            throw Error(format("home directory '%1%' exists; please remove it to assure purity of builds without sandboxing") % homeDir);
 
         /* We're not doing a chroot build, but we have some valid
            output paths.  Since we can't just overwrite or delete
@@ -2249,10 +2268,13 @@ void DerivationGoal::startBuilder()
 
         if (chown(slaveName.c_str(), buildUser->getUID(), 0))
             throw SysError("changing owner of pseudoterminal slave");
-    } else {
+    }
+#if __APPLE__
+    else {
         if (grantpt(builderOut.readSide.get()))
             throw SysError("granting access to pseudoterminal slave");
     }
+#endif
 
     #if 0
     // Mount the pt in the sandbox so that the "tty" command works.
@@ -2465,7 +2487,7 @@ void DerivationGoal::initTmpDir() {
                 auto hash = hashString(htSHA256, i.first);
                 string fn = ".attr-" + hash.to_string(Base32, false);
                 Path p = tmpDir + "/" + fn;
-                writeFile(p, i.second);
+                writeFile(p, rewriteStrings(i.second, inputRewrites));
                 chownToBuilder(p);
                 env[i.first + "Path"] = tmpDirInSandbox + "/" + fn;
             }
@@ -3151,7 +3173,7 @@ void DerivationGoal::runChild()
                 // Only use nss functions to resolve hosts and
                 // services. Don’t use it for anything else that may
                 // be configured for this system. This limits the
-                // potential impurities introduced in fixed outputs.
+                // potential impurities introduced in fixed-outputs.
                 writeFile(chrootRootDir + "/etc/nsswitch.conf", "hosts: files dns\nservices: files\n");
 
                 ss.push_back("/etc/services");
@@ -3536,6 +3558,29 @@ StorePathSet parseReferenceSpecifiers(Store & store, const BasicDerivation & drv
 }
 
 
+static void moveCheckToStore(const Path & src, const Path & dst)
+{
+    /* For the rename of directory to succeed, we must be running as root or
+       the directory must be made temporarily writable (to update the
+       directory's parent link ".."). */
+    struct stat st;
+    if (lstat(src.c_str(), &st) == -1) {
+        throw SysError(format("getting attributes of path '%1%'") % src);
+    }
+
+    bool changePerm = (geteuid() && S_ISDIR(st.st_mode) && !(st.st_mode & S_IWUSR));
+
+    if (changePerm)
+        chmod_(src, st.st_mode | S_IWUSR);
+
+    if (rename(src.c_str(), dst.c_str()))
+        throw SysError(format("renaming '%1%' to '%2%'") % src % dst);
+
+    if (changePerm)
+        chmod_(dst, st.st_mode);
+}
+
+
 void DerivationGoal::registerOutputs()
 {
     /* When using a build hook, the build hook can register the output
@@ -3651,7 +3696,8 @@ void DerivationGoal::registerOutputs()
                 /* 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(
-                        format("output path '%1%' should be a non-executable regular file") % path);
+                        format("output path '%1%' should be a non-executable regular file "
+                               "since recursive hashing is not enabled (outputHashMode=flat)") % path);
             }
 
             /* Check the hash. In hash mode, move the path produced by
@@ -3715,8 +3761,7 @@ void DerivationGoal::registerOutputs()
                 if (settings.runDiffHook || settings.keepFailed) {
                     Path dst = worker.store.toRealPath(path + checkSuffix);
                     deletePath(dst);
-                    if (rename(actualPath.c_str(), dst.c_str()))
-                        throw SysError(format("renaming '%1%' to '%2%'") % actualPath % dst);
+                    moveCheckToStore(actualPath, dst);
 
                     handleDiffHook(
                         buildUser ? buildUser->getUID() : getuid(),
@@ -3724,10 +3769,10 @@ void DerivationGoal::registerOutputs()
                         path, dst, worker.store.printStorePath(drvPath), tmpDir);
 
                     throw NotDeterministic("derivation '%s' may not be deterministic: output '%s' differs from '%s'",
-                        worker.store.printStorePath(drvPath), path, dst);
+                        worker.store.printStorePath(drvPath), worker.store.toRealPath(path), dst);
                 } else
                     throw NotDeterministic("derivation '%s' may not be deterministic: output '%s' differs",
-                        worker.store.printStorePath(drvPath), path);
+                        worker.store.printStorePath(drvPath), worker.store.toRealPath(path));
             }
 
             /* Since we verified the build, it's now ultimately trusted. */
@@ -3913,7 +3958,9 @@ void DerivationGoal::checkOutputs(const std::map<Path, ValidPathInfo> & outputs)
 
                 auto spec = parseReferenceSpecifiers(worker.store, *drv, *value);
 
-                auto used = static_cast<bool>(recursive) ? cloneStorePathSet(getClosure(info.path).first) : cloneStorePathSet(info.references);
+                auto used = recursive
+                    ? cloneStorePathSet(getClosure(info.path).first)
+                    : cloneStorePathSet(info.references);
 
                 if (recursive && checks.ignoreSelfRefs)
                     used.erase(info.path);
@@ -4767,8 +4814,7 @@ void Worker::waitForInput()
        terminated. */
 
     bool useTimeout = false;
-    struct timeval timeout;
-    timeout.tv_usec = 0;
+    long timeout = 0;
     auto before = steady_time_point::clock::now();
 
     /* If we're monitoring for silence on stdout/stderr, or if there
@@ -4786,7 +4832,7 @@ void Worker::waitForInput()
             nearest = std::min(nearest, i.timeStarted + std::chrono::seconds(settings.buildTimeout));
     }
     if (nearest != steady_time_point::max()) {
-        timeout.tv_sec = std::max(1L, (long) std::chrono::duration_cast<std::chrono::seconds>(nearest - before).count());
+        timeout = std::max(1L, (long) std::chrono::duration_cast<std::chrono::seconds>(nearest - before).count());
         useTimeout = true;
     }
 
@@ -4795,32 +4841,30 @@ void Worker::waitForInput()
     if (!waitingForAWhile.empty()) {
         useTimeout = true;
         if (lastWokenUp == steady_time_point::min())
-            printError("waiting for locks or build slots...");
+            printError("waiting for locks, build slots or build users...");
         if (lastWokenUp == steady_time_point::min() || lastWokenUp > before) lastWokenUp = before;
-        timeout.tv_sec = std::max(1L,
+        timeout = std::max(1L,
             (long) std::chrono::duration_cast<std::chrono::seconds>(
                 lastWokenUp + std::chrono::seconds(settings.pollInterval) - before).count());
     } else lastWokenUp = steady_time_point::min();
 
     if (useTimeout)
-        vomit("sleeping %d seconds", timeout.tv_sec);
+        vomit("sleeping %d seconds", timeout);
 
     /* Use select() to wait for the input side of any logger pipe to
        become `available'.  Note that `available' (i.e., non-blocking)
        includes EOF. */
-    fd_set fds;
-    FD_ZERO(&fds);
-    int fdMax = 0;
+    std::vector<struct pollfd> pollStatus;
+    std::map <int, int> fdToPollStatus;
     for (auto & i : children) {
         for (auto & j : i.fds) {
-            if (j >= FD_SETSIZE)
-                throw Error("reached FD_SETSIZE limit");
-            FD_SET(j, &fds);
-            if (j >= fdMax) fdMax = j + 1;
+            pollStatus.push_back((struct pollfd) { .fd = j, .events = POLLIN });
+            fdToPollStatus[j] = pollStatus.size() - 1;
         }
     }
 
-    if (select(fdMax, &fds, 0, 0, useTimeout ? &timeout : 0) == -1) {
+    if (poll(pollStatus.data(), pollStatus.size(),
+            useTimeout ? timeout * 1000 : -1) == -1) {
         if (errno == EINTR) return;
         throw SysError("waiting for input");
     }
@@ -4841,7 +4885,7 @@ void Worker::waitForInput()
         set<int> fds2(j->fds);
         std::vector<unsigned char> buffer(4096);
         for (auto & k : fds2) {
-            if (FD_ISSET(k, &fds)) {
+            if (pollStatus.at(fdToPollStatus.at(k)).revents) {
                 ssize_t rd = read(k, buffer.data(), buffer.size());
                 // FIXME: is there a cleaner way to handle pt close
                 // than EIO? Is this even standard?

src/libstore/builtins/fetchurl.cc

@@ -1,5 +1,5 @@
 #include "builtins.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "store-api.hh"
 #include "archive.hh"
 #include "compression.hh"
@@ -26,9 +26,9 @@ void builtinFetchurl(const BasicDerivation & drv, const std::string & netrcData)
     auto mainUrl = getAttr("url");
     bool unpack = get(drv.env, "unpack").value_or("") == "1";
 
-    /* Note: have to use a fresh downloader here because we're in
+    /* Note: have to use a fresh fileTransfer here because we're in
        a forked process. */
-    auto downloader = makeDownloader();
+    auto fileTransfer = makeFileTransfer();
 
     auto fetch = [&](const std::string & url) {
 
@@ -36,13 +36,13 @@ void builtinFetchurl(const BasicDerivation & drv, const std::string & netrcData)
 
             /* No need to do TLS verification, because we check the hash of
                the result anyway. */
-            DownloadRequest request(url);
+            FileTransferRequest request(url);
             request.verifyTLS = false;
             request.decompress = false;
 
             auto decompressor = makeDecompressionSink(
                 unpack && hasSuffix(mainUrl, ".xz") ? "xz" : "none", sink);
-            downloader->download(std::move(request), *decompressor);
+            fileTransfer->download(std::move(request), *decompressor);
             decompressor->finish();
         });
 

src/libstore/derivations.cc

@@ -388,7 +388,7 @@ Hash hashDerivationModulo(Store & store, const Derivation & drv, bool maskOutput
         if (h == drvHashes.end()) {
             assert(store.isValidPath(i.first));
             h = drvHashes.insert_or_assign(i.first.clone(), hashDerivationModulo(store,
-                readDerivation(store, store.toRealPath(store.printStorePath(i.first))), false)).first;
+                readDerivation(store, store.toRealPath(i.first)), false)).first;
         }
         inputs2.insert_or_assign(h->second.to_string(Base16, false), i.second);
     }

src/libstore/filetransfer.cc

@@ -1,14 +1,10 @@
-#include "download.hh"
+#include "filetransfer.hh"
 #include "util.hh"
 #include "globals.hh"
-#include "hash.hh"
 #include "store-api.hh"
-#include "archive.hh"
 #include "s3.hh"
 #include "compression.hh"
-#include "pathlocks.hh"
 #include "finally.hh"
-#include "tarfile.hh"
 
 #ifdef ENABLE_S3
 #include <aws/core/client/ClientConfiguration.h>
@@ -31,13 +27,9 @@ using namespace std::string_literals;
 
 namespace nix {
 
-DownloadSettings downloadSettings;
+FileTransferSettings fileTransferSettings;
 
-static GlobalConfig::Register r1(&downloadSettings);
-
-CachedDownloadRequest::CachedDownloadRequest(const std::string & uri)
-    : uri(uri), ttl(settings.tarballTtl)
-{ }
+static GlobalConfig::Register r1(&fileTransferSettings);
 
 std::string resolveUri(const std::string & uri)
 {
@@ -47,21 +39,21 @@ std::string resolveUri(const std::string & uri)
         return uri;
 }
 
-struct CurlDownloader : public Downloader
+struct curlFileTransfer : public FileTransfer
 {
     CURLM * curlm = 0;
 
     std::random_device rd;
     std::mt19937 mt19937;
 
-    struct DownloadItem : public std::enable_shared_from_this<DownloadItem>
+    struct TransferItem : public std::enable_shared_from_this<TransferItem>
     {
-        CurlDownloader & downloader;
-        DownloadRequest request;
-        DownloadResult result;
+        curlFileTransfer & fileTransfer;
+        FileTransferRequest request;
+        FileTransferResult result;
         Activity act;
         bool done = false; // whether either the success or failure function has been called
-        Callback<DownloadResult> callback;
+        Callback<FileTransferResult> callback;
         CURL * req = 0;
         bool active = false; // whether the handle has been added to the multi object
         std::string status;
@@ -80,19 +72,26 @@ struct CurlDownloader : public Downloader
 
         curl_off_t writtenToSink = 0;
 
-        DownloadItem(CurlDownloader & downloader,
-            const DownloadRequest & request,
-            Callback<DownloadResult> && callback)
-            : downloader(downloader)
+        TransferItem(curlFileTransfer & fileTransfer,
+            const FileTransferRequest & request,
+            Callback<FileTransferResult> && callback)
+            : fileTransfer(fileTransfer)
             , request(request)
-            , act(*logger, lvlTalkative, actDownload,
+            , act(*logger, lvlTalkative, actFileTransfer,
                 fmt(request.data ? "uploading '%s'" : "downloading '%s'", request.uri),
                 {request.uri}, request.parentAct)
             , callback(std::move(callback))
             , finalSink([this](const unsigned char * data, size_t len) {
                 if (this->request.dataCallback) {
-                    writtenToSink += len;
-                    this->request.dataCallback((char *) data, len);
+                    long httpStatus = 0;
+                    curl_easy_getinfo(req, CURLINFO_RESPONSE_CODE, &httpStatus);
+
+                    /* Only write data to the sink if this is a
+                       successful response. */
+                    if (httpStatus == 0 || httpStatus == 200 || httpStatus == 201 || httpStatus == 206) {
+                        writtenToSink += len;
+                        this->request.dataCallback((char *) data, len);
+                    }
                 } else
                     this->result.data->append((char *) data, len);
               })
@@ -103,17 +102,17 @@ struct CurlDownloader : public Downloader
                 requestHeaders = curl_slist_append(requestHeaders, ("Content-Type: " + request.mimeType).c_str());
         }
 
-        ~DownloadItem()
+        ~TransferItem()
         {
             if (req) {
                 if (active)
-                    curl_multi_remove_handle(downloader.curlm, req);
+                    curl_multi_remove_handle(fileTransfer.curlm, req);
                 curl_easy_cleanup(req);
             }
             if (requestHeaders) curl_slist_free_all(requestHeaders);
             try {
                 if (!done)
-                    fail(DownloadError(Interrupted, format("download of '%s' was interrupted") % request.uri));
+                    fail(FileTransferError(Interrupted, format("download of '%s' was interrupted") % request.uri));
             } catch (...) {
                 ignoreException();
             }
@@ -157,7 +156,7 @@ struct CurlDownloader : public Downloader
 
         static size_t writeCallbackWrapper(void * contents, size_t size, size_t nmemb, void * userp)
         {
-            return ((DownloadItem *) userp)->writeCallback(contents, size, nmemb);
+            return ((TransferItem *) userp)->writeCallback(contents, size, nmemb);
         }
 
         size_t headerCallback(void * contents, size_t size, size_t nmemb)
@@ -199,7 +198,7 @@ struct CurlDownloader : public Downloader
 
         static size_t headerCallbackWrapper(void * contents, size_t size, size_t nmemb, void * userp)
         {
-            return ((DownloadItem *) userp)->headerCallback(contents, size, nmemb);
+            return ((TransferItem *) userp)->headerCallback(contents, size, nmemb);
         }
 
         int progressCallback(double dltotal, double dlnow)
@@ -214,7 +213,7 @@ struct CurlDownloader : public Downloader
 
         static int progressCallbackWrapper(void * userp, double dltotal, double dlnow, double ultotal, double ulnow)
         {
-            return ((DownloadItem *) userp)->progressCallback(dltotal, dlnow);
+            return ((TransferItem *) userp)->progressCallback(dltotal, dlnow);
         }
 
         static int debugCallback(CURL * handle, curl_infotype type, char * data, size_t size, void * userptr)
@@ -238,7 +237,7 @@ struct CurlDownloader : public Downloader
 
         static size_t readCallbackWrapper(char *buffer, size_t size, size_t nitems, void * userp)
         {
-            return ((DownloadItem *) userp)->readCallback(buffer, size, nitems);
+            return ((TransferItem *) userp)->readCallback(buffer, size, nitems);
         }
 
         void init()
@@ -249,7 +248,7 @@ struct CurlDownloader : public Downloader
 
             if (verbosity >= lvlVomit) {
                 curl_easy_setopt(req, CURLOPT_VERBOSE, 1);
-                curl_easy_setopt(req, CURLOPT_DEBUGFUNCTION, DownloadItem::debugCallback);
+                curl_easy_setopt(req, CURLOPT_DEBUGFUNCTION, TransferItem::debugCallback);
             }
 
             curl_easy_setopt(req, CURLOPT_URL, request.uri.c_str());
@@ -258,19 +257,19 @@ struct CurlDownloader : public Downloader
             curl_easy_setopt(req, CURLOPT_NOSIGNAL, 1);
             curl_easy_setopt(req, CURLOPT_USERAGENT,
                 ("curl/" LIBCURL_VERSION " Nix/" + nixVersion +
-                    (downloadSettings.userAgentSuffix != "" ? " " + downloadSettings.userAgentSuffix.get() : "")).c_str());
+                    (fileTransferSettings.userAgentSuffix != "" ? " " + fileTransferSettings.userAgentSuffix.get() : "")).c_str());
             #if LIBCURL_VERSION_NUM >= 0x072b00
             curl_easy_setopt(req, CURLOPT_PIPEWAIT, 1);
             #endif
             #if LIBCURL_VERSION_NUM >= 0x072f00
-            if (downloadSettings.enableHttp2)
+            if (fileTransferSettings.enableHttp2)
                 curl_easy_setopt(req, CURLOPT_HTTP_VERSION, CURL_HTTP_VERSION_2TLS);
             else
                 curl_easy_setopt(req, CURLOPT_HTTP_VERSION, CURL_HTTP_VERSION_1_1);
             #endif
-            curl_easy_setopt(req, CURLOPT_WRITEFUNCTION, DownloadItem::writeCallbackWrapper);
+            curl_easy_setopt(req, CURLOPT_WRITEFUNCTION, TransferItem::writeCallbackWrapper);
             curl_easy_setopt(req, CURLOPT_WRITEDATA, this);
-            curl_easy_setopt(req, CURLOPT_HEADERFUNCTION, DownloadItem::headerCallbackWrapper);
+            curl_easy_setopt(req, CURLOPT_HEADERFUNCTION, TransferItem::headerCallbackWrapper);
             curl_easy_setopt(req, CURLOPT_HEADERDATA, this);
 
             curl_easy_setopt(req, CURLOPT_PROGRESSFUNCTION, progressCallbackWrapper);
@@ -298,10 +297,10 @@ struct CurlDownloader : public Downloader
                 curl_easy_setopt(req, CURLOPT_SSL_VERIFYHOST, 0);
             }
 
-            curl_easy_setopt(req, CURLOPT_CONNECTTIMEOUT, downloadSettings.connectTimeout.get());
+            curl_easy_setopt(req, CURLOPT_CONNECTTIMEOUT, fileTransferSettings.connectTimeout.get());
 
             curl_easy_setopt(req, CURLOPT_LOW_SPEED_LIMIT, 1L);
-            curl_easy_setopt(req, CURLOPT_LOW_SPEED_TIME, downloadSettings.stalledDownloadTimeout.get());
+            curl_easy_setopt(req, CURLOPT_LOW_SPEED_TIME, fileTransferSettings.stalledDownloadTimeout.get());
 
             /* If no file exist in the specified path, curl continues to work
                anyway as if netrc support was disabled. */
@@ -390,6 +389,7 @@ struct CurlDownloader : public Downloader
                         case CURLE_SSL_CACERT_BADFILE:
                         case CURLE_TOO_MANY_REDIRECTS:
                         case CURLE_WRITE_ERROR:
+                        case CURLE_UNSUPPORTED_PROTOCOL:
                             err = Misc;
                             break;
                         default: // Shut up warnings
@@ -401,14 +401,14 @@ struct CurlDownloader : public Downloader
 
                 auto exc =
                     code == CURLE_ABORTED_BY_CALLBACK && _isInterrupted
-                    ? DownloadError(Interrupted, fmt("%s of '%s' was interrupted", request.verb(), request.uri))
+                    ? FileTransferError(Interrupted, fmt("%s of '%s' was interrupted", request.verb(), request.uri))
                     : httpStatus != 0
-                    ? DownloadError(err,
+                    ? FileTransferError(err,
                         fmt("unable to %s '%s': HTTP error %d",
                             request.verb(), request.uri, httpStatus)
                         + (code == CURLE_OK ? "" : fmt(" (curl error: %s)", curl_easy_strerror(code)))
                         )
-                    : DownloadError(err,
+                    : FileTransferError(err,
                         fmt("unable to %s '%s': %s (%d)",
                             request.verb(), request.uri, curl_easy_strerror(code), code));
 
@@ -422,13 +422,13 @@ struct CurlDownloader : public Downloader
                         || writtenToSink == 0
                         || (acceptRanges && encoding.empty())))
                 {
-                    int ms = request.baseRetryTimeMs * std::pow(2.0f, attempt - 1 + std::uniform_real_distribution<>(0.0, 0.5)(downloader.mt19937));
+                    int ms = request.baseRetryTimeMs * std::pow(2.0f, attempt - 1 + std::uniform_real_distribution<>(0.0, 0.5)(fileTransfer.mt19937));
                     if (writtenToSink)
                         warn("%s; retrying from offset %d in %d ms", exc.what(), writtenToSink, ms);
                     else
                         warn("%s; retrying in %d ms", exc.what(), ms);
                     embargo = std::chrono::steady_clock::now() + std::chrono::milliseconds(ms);
-                    downloader.enqueueItem(shared_from_this());
+                    fileTransfer.enqueueItem(shared_from_this());
                 }
                 else
                     fail(exc);
@@ -439,12 +439,12 @@ struct CurlDownloader : public Downloader
     struct State
     {
         struct EmbargoComparator {
-            bool operator() (const std::shared_ptr<DownloadItem> & i1, const std::shared_ptr<DownloadItem> & i2) {
+            bool operator() (const std::shared_ptr<TransferItem> & i1, const std::shared_ptr<TransferItem> & i2) {
                 return i1->embargo > i2->embargo;
             }
         };
         bool quit = false;
-        std::priority_queue<std::shared_ptr<DownloadItem>, std::vector<std::shared_ptr<DownloadItem>>, EmbargoComparator> incoming;
+        std::priority_queue<std::shared_ptr<TransferItem>, std::vector<std::shared_ptr<TransferItem>>, EmbargoComparator> incoming;
     };
 
     Sync<State> state_;
@@ -456,7 +456,7 @@ struct CurlDownloader : public Downloader
 
     std::thread workerThread;
 
-    CurlDownloader()
+    curlFileTransfer()
         : mt19937(rd())
     {
         static std::once_flag globalInit;
@@ -469,7 +469,7 @@ struct CurlDownloader : public Downloader
         #endif
         #if LIBCURL_VERSION_NUM >= 0x071e00 // Max connections requires >= 7.30.0
         curl_multi_setopt(curlm, CURLMOPT_MAX_TOTAL_CONNECTIONS,
-            downloadSettings.httpConnections.get());
+            fileTransferSettings.httpConnections.get());
         #endif
 
         wakeupPipe.create();
@@ -478,7 +478,7 @@ struct CurlDownloader : public Downloader
         workerThread = std::thread([&]() { workerThreadEntry(); });
     }
 
-    ~CurlDownloader()
+    ~curlFileTransfer()
     {
         stopWorkerThread();
 
@@ -504,7 +504,7 @@ struct CurlDownloader : public Downloader
             stopWorkerThread();
         });
 
-        std::map<CURL *, std::shared_ptr<DownloadItem>> items;
+        std::map<CURL *, std::shared_ptr<TransferItem>> items;
 
         bool quit = false;
 
@@ -561,7 +561,7 @@ struct CurlDownloader : public Downloader
                     throw SysError("reading curl wakeup socket");
             }
 
-            std::vector<std::shared_ptr<DownloadItem>> incoming;
+            std::vector<std::shared_ptr<TransferItem>> incoming;
             auto now = std::chrono::steady_clock::now();
 
             {
@@ -609,7 +609,7 @@ struct CurlDownloader : public Downloader
         }
     }
 
-    void enqueueItem(std::shared_ptr<DownloadItem> item)
+    void enqueueItem(std::shared_ptr<TransferItem> item)
     {
         if (item->request.data
             && !hasPrefix(item->request.uri, "http://")
@@ -641,8 +641,8 @@ struct CurlDownloader : public Downloader
     }
 #endif
 
-    void enqueueDownload(const DownloadRequest & request,
-        Callback<DownloadResult> callback) override
+    void enqueueFileTransfer(const FileTransferRequest & request,
+        Callback<FileTransferResult> callback) override
     {
         /* Ugly hack to support s3:// URIs. */
         if (hasPrefix(request.uri, "s3://")) {
@@ -660,9 +660,9 @@ struct CurlDownloader : public Downloader
 
                 // FIXME: implement ETag
                 auto s3Res = s3Helper.getObject(bucketName, key);
-                DownloadResult res;
+                FileTransferResult res;
                 if (!s3Res.data)
-                    throw DownloadError(NotFound, fmt("S3 object '%s' does not exist", request.uri));
+                    throw FileTransferError(NotFound, fmt("S3 object '%s' does not exist", request.uri));
                 res.data = s3Res.data;
                 callback(std::move(res));
 #else
@@ -672,26 +672,26 @@ struct CurlDownloader : public Downloader
             return;
         }
 
-        enqueueItem(std::make_shared<DownloadItem>(*this, request, std::move(callback)));
+        enqueueItem(std::make_shared<TransferItem>(*this, request, std::move(callback)));
     }
 };
 
-ref<Downloader> getDownloader()
+ref<FileTransfer> getFileTransfer()
 {
-    static ref<Downloader> downloader = makeDownloader();
-    return downloader;
+    static ref<FileTransfer> fileTransfer = makeFileTransfer();
+    return fileTransfer;
 }
 
-ref<Downloader> makeDownloader()
+ref<FileTransfer> makeFileTransfer()
 {
-    return make_ref<CurlDownloader>();
+    return make_ref<curlFileTransfer>();
 }
 
-std::future<DownloadResult> Downloader::enqueueDownload(const DownloadRequest & request)
+std::future<FileTransferResult> FileTransfer::enqueueFileTransfer(const FileTransferRequest & request)
 {
-    auto promise = std::make_shared<std::promise<DownloadResult>>();
-    enqueueDownload(request,
-        {[promise](std::future<DownloadResult> fut) {
+    auto promise = std::make_shared<std::promise<FileTransferResult>>();
+    enqueueFileTransfer(request,
+        {[promise](std::future<FileTransferResult> fut) {
             try {
                 promise->set_value(fut.get());
             } catch (...) {
@@ -701,15 +701,21 @@ std::future<DownloadResult> Downloader::enqueueDownload(const DownloadRequest &
     return promise->get_future();
 }
 
-DownloadResult Downloader::download(const DownloadRequest & request)
+FileTransferResult FileTransfer::download(const FileTransferRequest & request)
 {
-    return enqueueDownload(request).get();
+    return enqueueFileTransfer(request).get();
 }
 
-void Downloader::download(DownloadRequest && request, Sink & sink)
+FileTransferResult FileTransfer::upload(const FileTransferRequest & request)
+{
+    /* Note: this method is the same as download, but helps in readability */
+    return enqueueFileTransfer(request).get();
+}
+
+void FileTransfer::download(FileTransferRequest && request, Sink & sink)
 {
     /* Note: we can't call 'sink' via request.dataCallback, because
-       that would cause the sink to execute on the downloader
+       that would cause the sink to execute on the fileTransfer
        thread. If 'sink' is a coroutine, this will fail. Also, if the
        sink is expensive (e.g. one that does decompression and writing
        to the Nix store), it would stall the download thread too much.
@@ -755,8 +761,8 @@ void Downloader::download(DownloadRequest && request, Sink & sink)
         state->avail.notify_one();
     };
 
-    enqueueDownload(request,
-        {[_state](std::future<DownloadResult> fut) {
+    enqueueFileTransfer(request,
+        {[_state](std::future<FileTransferResult> fut) {
             auto state(_state->lock());
             state->quit = true;
             try {
@@ -801,141 +807,6 @@ void Downloader::download(DownloadRequest && request, Sink & sink)
     }
 }
 
-CachedDownloadResult Downloader::downloadCached(
-    ref<Store> store, const CachedDownloadRequest & request)
-{
-    auto url = resolveUri(request.uri);
-
-    auto name = request.name;
-    if (name == "") {
-        auto p = url.rfind('/');
-        if (p != string::npos) name = string(url, p + 1);
-    }
-
-    std::optional<StorePath> expectedStorePath;
-    if (request.expectedHash) {
-        auto method = request.unpack ? FileIngestionMethod::Recursive : FileIngestionMethod::Flat;
-        expectedStorePath = store->makeFixedOutputPath(method, request.expectedHash, name);
-        if (store->isValidPath(*expectedStorePath)) {
-            CachedDownloadResult result;
-            result.storePath = store->printStorePath(*expectedStorePath);
-            result.path = store->toRealPath(result.storePath);
-            return result;
-        }
-    }
-
-    Path cacheDir = getCacheDir() + "/nix/tarballs";
-    createDirs(cacheDir);
-
-    string urlHash = hashString(htSHA256, name + std::string("\0"s) + url).to_string(Base32, false);
-
-    Path dataFile = cacheDir + "/" + urlHash + ".info";
-    Path fileLink = cacheDir + "/" + urlHash + "-file";
-
-    PathLocks lock({fileLink}, fmt("waiting for lock on '%1%'...", fileLink));
-
-    std::optional<StorePath> storePath;
-
-    string expectedETag;
-
-    bool skip = false;
-
-    CachedDownloadResult result;
-
-    if (pathExists(fileLink) && pathExists(dataFile)) {
-        storePath = store->parseStorePath(readLink(fileLink));
-        // FIXME
-        store->addTempRoot(*storePath);
-        if (store->isValidPath(*storePath)) {
-            auto ss = tokenizeString<vector<string>>(readFile(dataFile), "\n");
-            if (ss.size() >= 3 && ss[0] == url) {
-                time_t lastChecked;
-                if (string2Int(ss[2], lastChecked) && (uint64_t) lastChecked + request.ttl >= (uint64_t) time(0)) {
-                    skip = true;
-                    result.effectiveUri = request.uri;
-                    result.etag = ss[1];
-                } else if (!ss[1].empty()) {
-                    debug(format("verifying previous ETag '%1%'") % ss[1]);
-                    expectedETag = ss[1];
-                }
-            }
-        } else
-            storePath.reset();
-    }
-
-    if (!skip) {
-
-        try {
-            DownloadRequest request2(url);
-            request2.expectedETag = expectedETag;
-            auto res = download(request2);
-            result.effectiveUri = res.effectiveUri;
-            result.etag = res.etag;
-
-            if (!res.cached) {
-                StringSink sink;
-                dumpString(*res.data, sink);
-                Hash hash = hashString(request.expectedHash ? request.expectedHash.type : htSHA256, *res.data);
-                ValidPathInfo info(store->makeFixedOutputPath(FileIngestionMethod::Flat, hash, name));
-                info.narHash = hashString(htSHA256, *sink.s);
-                info.narSize = sink.s->size();
-                info.ca = makeFixedOutputCA(FileIngestionMethod::Flat, hash);
-                store->addToStore(info, sink.s, NoRepair, NoCheckSigs);
-                storePath = info.path.clone();
-            }
-
-            assert(storePath);
-            replaceSymlink(store->printStorePath(*storePath), fileLink);
-
-            writeFile(dataFile, url + "\n" + res.etag + "\n" + std::to_string(time(0)) + "\n");
-        } catch (DownloadError & e) {
-            if (!storePath) throw;
-            warn("warning: %s; using cached result", e.msg());
-            result.etag = expectedETag;
-        }
-    }
-
-    if (request.unpack) {
-        Path unpackedLink = cacheDir + "/" + ((std::string) storePath->to_string()) + "-unpacked";
-        PathLocks lock2({unpackedLink}, fmt("waiting for lock on '%1%'...", unpackedLink));
-        std::optional<StorePath> unpackedStorePath;
-        if (pathExists(unpackedLink)) {
-            unpackedStorePath = store->parseStorePath(readLink(unpackedLink));
-            // FIXME
-            store->addTempRoot(*unpackedStorePath);
-            if (!store->isValidPath(*unpackedStorePath))
-                unpackedStorePath.reset();
-        }
-        if (!unpackedStorePath) {
-            printInfo("unpacking '%s'...", url);
-            Path tmpDir = createTempDir();
-            AutoDelete autoDelete(tmpDir, true);
-            unpackTarfile(store->toRealPath(store->printStorePath(*storePath)), tmpDir);
-            auto members = readDirectory(tmpDir);
-            if (members.size() != 1)
-                throw nix::Error("tarball '%s' contains an unexpected number of top-level files", url);
-            auto topDir = tmpDir + "/" + members.begin()->name;
-            unpackedStorePath = store->addToStore(name, topDir, FileIngestionMethod::Recursive, htSHA256, defaultPathFilter, NoRepair);
-        }
-        replaceSymlink(store->printStorePath(*unpackedStorePath), unpackedLink);
-        storePath = std::move(*unpackedStorePath);
-    }
-
-    if (expectedStorePath && *storePath != *expectedStorePath) {
-        unsigned int statusCode = 102;
-        Hash gotHash = request.unpack
-            ? hashPath(request.expectedHash.type, store->toRealPath(store->printStorePath(*storePath))).first
-            : hashFile(request.expectedHash.type, store->toRealPath(store->printStorePath(*storePath)));
-        throw nix::Error(statusCode, "hash mismatch in file downloaded from '%s':\n  wanted: %s\n  got:    %s",
-            url, request.expectedHash.to_string(), gotHash.to_string());
-    }
-
-    result.storePath = store->printStorePath(*storePath);
-    result.path = store->toRealPath(result.storePath);
-    return result;
-}
-
-
 bool isUri(const string & s)
 {
     if (s.compare(0, 8, "channel:") == 0) return true;

src/libstore/filetransfer.hh

@@ -9,7 +9,7 @@
 
 namespace nix {
 
-struct DownloadSettings : Config
+struct FileTransferSettings : Config
 {
     Setting<bool> enableHttp2{this, true, "http2",
         "Whether to enable HTTP/2 support."};
@@ -31,15 +31,15 @@ struct DownloadSettings : Config
         "How often Nix will attempt to download a file before giving up."};
 };
 
-extern DownloadSettings downloadSettings;
+extern FileTransferSettings fileTransferSettings;
 
-struct DownloadRequest
+struct FileTransferRequest
 {
     std::string uri;
     std::string expectedETag;
     bool verifyTLS = true;
     bool head = false;
-    size_t tries = downloadSettings.tries;
+    size_t tries = fileTransferSettings.tries;
     unsigned int baseRetryTimeMs = 250;
     ActivityId parentAct;
     bool decompress = true;
@@ -47,7 +47,7 @@ struct DownloadRequest
     std::string mimeType;
     std::function<void(char *, size_t)> dataCallback;
 
-    DownloadRequest(const std::string & uri)
+    FileTransferRequest(const std::string & uri)
         : uri(uri), parentAct(getCurActivity()) { }
 
     std::string verb()
@@ -56,7 +56,7 @@ struct DownloadRequest
     }
 };
 
-struct DownloadResult
+struct FileTransferResult
 {
     bool cached = false;
     std::string etag;
@@ -65,74 +65,52 @@ struct DownloadResult
     uint64_t bodySize = 0;
 };
 
-struct CachedDownloadRequest
-{
-    std::string uri;
-    bool unpack = false;
-    std::string name;
-    Hash expectedHash;
-    unsigned int ttl;
-
-    CachedDownloadRequest(const std::string & uri);
-    CachedDownloadRequest() = delete;
-};
-
-struct CachedDownloadResult
-{
-    // Note: 'storePath' may be different from 'path' when using a
-    // chroot store.
-    Path storePath;
-    Path path;
-    std::optional<std::string> etag;
-    std::string effectiveUri;
-};
-
 class Store;
 
-struct Downloader
+struct FileTransfer
 {
-    virtual ~Downloader() { }
+    virtual ~FileTransfer() { }
 
-    /* Enqueue a download request, returning a future to the result of
-       the download. The future may throw a DownloadError
+    /* Enqueue a data transfer request, returning a future to the result of
+       the download. The future may throw a FileTransferError
        exception. */
-    virtual void enqueueDownload(const DownloadRequest & request,
-        Callback<DownloadResult> callback) = 0;
+    virtual void enqueueFileTransfer(const FileTransferRequest & request,
+        Callback<FileTransferResult> callback) = 0;
 
-    std::future<DownloadResult> enqueueDownload(const DownloadRequest & request);
+    std::future<FileTransferResult> enqueueFileTransfer(const FileTransferRequest & request);
 
     /* Synchronously download a file. */
-    DownloadResult download(const DownloadRequest & request);
+    FileTransferResult download(const FileTransferRequest & request);
+
+    /* Synchronously upload a file. */
+    FileTransferResult upload(const FileTransferRequest & request);
 
     /* Download a file, writing its data to a sink. The sink will be
        invoked on the thread of the caller. */
-    void download(DownloadRequest && request, Sink & sink);
-
-    /* Check if the specified file is already in ~/.cache/nix/tarballs
-       and is more recent than ‘tarball-ttl’ seconds. Otherwise,
-       use the recorded ETag to verify if the server has a more
-       recent version, and if so, download it to the Nix store. */
-    CachedDownloadResult downloadCached(ref<Store> store, const CachedDownloadRequest & request);
+    void download(FileTransferRequest && request, Sink & sink);
 
     enum Error { NotFound, Forbidden, Misc, Transient, Interrupted };
 };
 
-/* Return a shared Downloader object. Using this object is preferred
+/* Return a shared FileTransfer object. Using this object is preferred
    because it enables connection reuse and HTTP/2 multiplexing. */
-ref<Downloader> getDownloader();
+ref<FileTransfer> getFileTransfer();
 
-/* Return a new Downloader object. */
-ref<Downloader> makeDownloader();
+/* Return a new FileTransfer object. */
+ref<FileTransfer> makeFileTransfer();
 
-class DownloadError : public Error
+class FileTransferError : public Error
 {
 public:
-    Downloader::Error error;
-    DownloadError(Downloader::Error error, const FormatOrString & fs)
+    FileTransfer::Error error;
+    FileTransferError(FileTransfer::Error error, const FormatOrString & fs)
         : Error(fs), error(error)
     { }
 };
 
 bool isUri(const string & s);
 
+/* Resolve deprecated 'channel:<foo>' URLs. */
+std::string resolveUri(const std::string & uri);
+
 }

src/libstore/gc.cc

@@ -202,6 +202,11 @@ void LocalStore::findTempRoots(FDs & fds, Roots & tempRoots, bool censor)
     /* Read the `temproots' directory for per-process temporary root
        files. */
     for (auto & i : readDirectory(tempRootsDir)) {
+        if (i.name[0] == '.') {
+            // Ignore hidden files. Some package managers (notably portage) create
+            // those to keep the directory alive.
+            continue;
+        }
         Path path = tempRootsDir + "/" + i.name;
 
         pid_t pid = std::stoi(i.name);
@@ -414,7 +419,7 @@ void LocalStore::findRuntimeRoots(Roots & roots, bool censor)
 
                 try {
                     auto mapFile = fmt("/proc/%s/maps", ent->d_name);
-                    auto mapLines = tokenizeString<std::vector<string>>(readFile(mapFile, true), "\n");
+                    auto mapLines = tokenizeString<std::vector<string>>(readFile(mapFile), "\n");
                     for (const auto & line : mapLines) {
                         auto match = std::smatch{};
                         if (std::regex_match(line, match, mapRegex))
@@ -422,7 +427,7 @@ void LocalStore::findRuntimeRoots(Roots & roots, bool censor)
                     }
 
                     auto envFile = fmt("/proc/%s/environ", ent->d_name);
-                    auto envString = readFile(envFile, true);
+                    auto envString = readFile(envFile);
                     auto env_end = std::sregex_iterator{};
                     for (auto i = std::sregex_iterator{envString.begin(), envString.end(), storePathRegex}; i != env_end; ++i)
                         unchecked[i->str()].emplace(envFile);
@@ -884,7 +889,7 @@ void LocalStore::autoGC(bool sync)
         if (statvfs(realStoreDir.c_str(), &st))
             throw SysError("getting filesystem info about '%s'", realStoreDir);
 
-        return (uint64_t) st.f_bavail * st.f_bsize;
+        return (uint64_t) st.f_bavail * st.f_frsize;
     };
 
     std::shared_future<void> future;

src/libstore/globals.cc

@@ -31,6 +31,7 @@ Settings::Settings()
     , nixLogDir(canonPath(getEnv("NIX_LOG_DIR").value_or(NIX_LOG_DIR)))
     , nixStateDir(canonPath(getEnv("NIX_STATE_DIR").value_or(NIX_STATE_DIR)))
     , nixConfDir(canonPath(getEnv("NIX_CONF_DIR").value_or(NIX_CONF_DIR)))
+    , nixUserConfFiles(getUserConfigFiles())
     , nixLibexecDir(canonPath(getEnv("NIX_LIBEXEC_DIR").value_or(NIX_LIBEXEC_DIR)))
     , nixBinDir(canonPath(getEnv("NIX_BIN_DIR").value_or(NIX_BIN_DIR)))
     , nixManDir(canonPath(NIX_MAN_DIR))
@@ -77,13 +78,29 @@ void loadConfFile()
        ~/.nix/nix.conf or the command line. */
     globalConfig.resetOverriden();
 
-    auto dirs = getConfigDirs();
-    // Iterate over them in reverse so that the ones appearing first in the path take priority
-    for (auto dir = dirs.rbegin(); dir != dirs.rend(); dir++) {
-        globalConfig.applyConfigFile(*dir + "/nix/nix.conf");
+    auto files = settings.nixUserConfFiles;
+    for (auto file = files.rbegin(); file != files.rend(); file++) {
+        globalConfig.applyConfigFile(*file);
     }
 }
 
+std::vector<Path> getUserConfigFiles()
+{
+    // Use the paths specified in NIX_USER_CONF_FILES if it has been defined
+    auto nixConfFiles = getEnv("NIX_USER_CONF_FILES");
+    if (nixConfFiles.has_value()) {
+        return tokenizeString<std::vector<string>>(nixConfFiles.value(), ":");
+    }
+
+    // Use the paths specified by the XDG spec
+    std::vector<Path> files;
+    auto dirs = getConfigDirs();
+    for (auto & dir : dirs) {
+        files.insert(files.end(), dir + "/nix/nix.conf");
+    }
+    return files;
+}
+
 unsigned int Settings::getDefaultCores()
 {
     return std::max(1U, std::thread::hardware_concurrency());
@@ -113,7 +130,7 @@ bool Settings::isExperimentalFeatureEnabled(const std::string & name)
 void Settings::requireExperimentalFeature(const std::string & name)
 {
     if (!isExperimentalFeatureEnabled(name))
-        throw Error("experimental Nix feature '%s' is disabled", name);
+        throw Error("experimental Nix feature '%1%' is disabled; use '--experimental-features %1%' to override", name);
 }
 
 bool Settings::isWSL1()
@@ -150,21 +167,24 @@ template<> void BaseSetting<SandboxMode>::toJSON(JSONPlaceholder & out)
 
 template<> void BaseSetting<SandboxMode>::convertToArg(Args & args, const std::string & category)
 {
-    args.mkFlag()
-        .longName(name)
-        .description("Enable sandboxing.")
-        .handler([=](std::vector<std::string> ss) { override(smEnabled); })
-        .category(category);
-    args.mkFlag()
-        .longName("no-" + name)
-        .description("Disable sandboxing.")
-        .handler([=](std::vector<std::string> ss) { override(smDisabled); })
-        .category(category);
-    args.mkFlag()
-        .longName("relaxed-" + name)
-        .description("Enable sandboxing, but allow builds to disable it.")
-        .handler([=](std::vector<std::string> ss) { override(smRelaxed); })
-        .category(category);
+    args.addFlag({
+        .longName = name,
+        .description = "Enable sandboxing.",
+        .category = category,
+        .handler = {[=]() { override(smEnabled); }}
+    });
+    args.addFlag({
+        .longName = "no-" + name,
+        .description = "Disable sandboxing.",
+        .category = category,
+        .handler = {[=]() { override(smDisabled); }}
+    });
+    args.addFlag({
+        .longName = "relaxed-" + name,
+        .description = "Enable sandboxing, but allow builds to disable it.",
+        .category = category,
+        .handler = {[=]() { override(smRelaxed); }}
+    });
 }
 
 void MaxBuildJobsSetting::set(const std::string & str)

src/libstore/globals.hh

@@ -53,9 +53,12 @@ public:
     /* The directory where state is stored. */
     Path nixStateDir;
 
-    /* The directory where configuration files are stored. */
+    /* The directory where system configuration files are stored. */
     Path nixConfDir;
 
+    /* A list of user configuration files to load. */
+    std::vector<Path> nixUserConfFiles;
+
     /* The directory where internal helper programs are stored. */
     Path nixLibexecDir;
 
@@ -351,12 +354,21 @@ public:
     Setting<Paths> pluginFiles{this, {}, "plugin-files",
         "Plugins to dynamically load at nix initialization time."};
 
+    Setting<std::string> githubAccessToken{this, "", "github-access-token",
+        "GitHub access token to get access to GitHub data through the GitHub API for github:<..> flakes."};
+
     Setting<Strings> experimentalFeatures{this, {}, "experimental-features",
         "Experimental Nix features to enable."};
 
     bool isExperimentalFeatureEnabled(const std::string & name);
 
     void requireExperimentalFeature(const std::string & name);
+
+    Setting<bool> allowDirty{this, true, "allow-dirty",
+        "Whether to allow dirty Git/Mercurial trees."};
+
+    Setting<bool> warnDirty{this, true, "warn-dirty",
+        "Whether to warn about dirty Git/Mercurial trees."};
 };
 
 
@@ -369,6 +381,9 @@ void initPlugins();
 
 void loadConfFile();
 
+// Used by the Settings constructor
+std::vector<Path> getUserConfigFiles();
+
 extern const string nixVersion;
 
 }

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

@@ -1,5 +1,5 @@
 #include "binary-cache-store.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "globals.hh"
 #include "nar-info-disk-cache.hh"
 
@@ -85,14 +85,14 @@ protected:
         checkEnabled();
 
         try {
-            DownloadRequest request(cacheUri + "/" + path);
+            FileTransferRequest request(cacheUri + "/" + path);
             request.head = true;
-            getDownloader()->download(request);
+            getFileTransfer()->download(request);
             return true;
-        } catch (DownloadError & e) {
+        } catch (FileTransferError & e) {
             /* S3 buckets return 403 if a file doesn't exist and the
                bucket is unlistable, so treat 403 as 404. */
-            if (e.error == Downloader::NotFound || e.error == Downloader::Forbidden)
+            if (e.error == FileTransfer::NotFound || e.error == FileTransfer::Forbidden)
                 return false;
             maybeDisable();
             throw;
@@ -103,19 +103,19 @@ protected:
         const std::string & data,
         const std::string & mimeType) override
     {
-        auto req = DownloadRequest(cacheUri + "/" + path);
+        auto req = FileTransferRequest(cacheUri + "/" + path);
         req.data = std::make_shared<string>(data); // FIXME: inefficient
         req.mimeType = mimeType;
         try {
-            getDownloader()->download(req);
-        } catch (DownloadError & e) {
+            getFileTransfer()->upload(req);
+        } catch (FileTransferError & e) {
             throw UploadToHTTP("while uploading to HTTP binary cache at '%s': %s", cacheUri, e.msg());
         }
     }
 
-    DownloadRequest makeRequest(const std::string & path)
+    FileTransferRequest makeRequest(const std::string & path)
     {
-        DownloadRequest request(cacheUri + "/" + path);
+        FileTransferRequest request(cacheUri + "/" + path);
         return request;
     }
 
@@ -124,9 +124,9 @@ protected:
         checkEnabled();
         auto request(makeRequest(path));
         try {
-            getDownloader()->download(std::move(request), sink);
-        } catch (DownloadError & e) {
-            if (e.error == Downloader::NotFound || e.error == Downloader::Forbidden)
+            getFileTransfer()->download(std::move(request), sink);
+        } catch (FileTransferError & e) {
+            if (e.error == FileTransfer::NotFound || e.error == FileTransfer::Forbidden)
                 throw NoSuchBinaryCacheFile("file '%s' does not exist in binary cache '%s'", path, getUri());
             maybeDisable();
             throw;
@@ -142,12 +142,12 @@ protected:
 
         auto callbackPtr = std::make_shared<decltype(callback)>(std::move(callback));
 
-        getDownloader()->enqueueDownload(request,
-            {[callbackPtr, this](std::future<DownloadResult> result) {
+        getFileTransfer()->enqueueFileTransfer(request,
+            {[callbackPtr, this](std::future<FileTransferResult> result) {
                 try {
                     (*callbackPtr)(result.get().data);
-                } catch (DownloadError & e) {
-                    if (e.error == Downloader::NotFound || e.error == Downloader::Forbidden)
+                } catch (FileTransferError & e) {
+                    if (e.error == FileTransfer::NotFound || e.error == FileTransfer::Forbidden)
                         return (*callbackPtr)(std::shared_ptr<std::string>());
                     maybeDisable();
                     callbackPtr->rethrow();
@@ -174,4 +174,3 @@ static RegisterStoreImplementation regStore([](
 });
 
 }
-

src/libstore/local.mk

@@ -31,7 +31,8 @@ ifeq ($(HAVE_SECCOMP), 1)
 	libstore_LDFLAGS += -lseccomp
 endif
 
-libstore_CXXFLAGS = \
+libstore_CXXFLAGS += \
+ -I src/libutil -I src/libstore \
  -DNIX_PREFIX=\"$(prefix)\" \
  -DNIX_STORE_DIR=\"$(storedir)\" \
  -DNIX_DATA_DIR=\"$(datadir)\" \

src/libstore/path.hh

@@ -87,7 +87,7 @@ const size_t storePathHashLen = 32; // i.e. 160 bits
 /* Extension of derivations in the Nix store. */
 const std::string drvExtension = ".drv";
 
-enum struct FileIngestionMethod : bool {
+enum struct FileIngestionMethod : uint8_t {
     Flat = false,
     Recursive = true
 };

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

@@ -6,7 +6,7 @@
 #include "nar-info-disk-cache.hh"
 #include "globals.hh"
 #include "compression.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "istringstream_nocopy.hh"
 
 #include <aws/core/Aws.h>
@@ -132,7 +132,7 @@ ref<Aws::Client::ClientConfiguration> S3Helper::makeConfig(const string & region
     return res;
 }
 
-S3Helper::DownloadResult S3Helper::getObject(
+S3Helper::FileTransferResult S3Helper::getObject(
     const std::string & bucketName, const std::string & key)
 {
     debug("fetching 's3://%s/%s'...", bucketName, key);
@@ -146,7 +146,7 @@ S3Helper::DownloadResult S3Helper::getObject(
         return Aws::New<std::stringstream>("STRINGSTREAM");
     });
 
-    DownloadResult res;
+    FileTransferResult res;
 
     auto now1 = std::chrono::steady_clock::now();
 

src/libstore/s3.hh

@@ -18,13 +18,13 @@ struct S3Helper
 
     ref<Aws::Client::ClientConfiguration> makeConfig(const std::string & region, const std::string & scheme, const std::string & endpoint);
 
-    struct DownloadResult
+    struct FileTransferResult
     {
         std::shared_ptr<std::string> data;
         unsigned int durationMs;
     };
 
-    DownloadResult getObject(
+    FileTransferResult getObject(
         const std::string & bucketName, const std::string & key);
 };
 

src/libstore/store-api.cc

@@ -6,6 +6,7 @@
 #include "thread-pool.hh"
 #include "json.hh"
 #include "derivations.hh"
+#include "url.hh"
 
 #include <future>
 
@@ -40,7 +41,7 @@ Path Store::followLinksToStore(std::string_view _path) const
         path = absPath(target, dirOf(path));
     }
     if (!isInStore(path))
-        throw Error(format("path '%1%' is not in the Nix store") % path);
+        throw NotInStore("path '%1%' is not in the Nix store", path);
     return path;
 }
 
@@ -202,7 +203,9 @@ StorePath Store::makeTextPath(std::string_view name, const Hash & hash,
 std::pair<StorePath, Hash> Store::computeStorePathForPath(std::string_view name,
     const Path & srcPath, FileIngestionMethod recursive, HashType hashAlgo, PathFilter & filter) const
 {
-    Hash h = static_cast<bool>(recursive) ? hashPath(hashAlgo, srcPath, filter).first : hashFile(hashAlgo, srcPath);
+    Hash h = recursive == FileIngestionMethod::Recursive
+        ? hashPath(hashAlgo, srcPath, filter).first
+        : hashFile(hashAlgo, srcPath);
     return std::make_pair(makeFixedOutputPath(recursive, h, name), h);
 }
 
@@ -767,7 +770,7 @@ bool ValidPathInfo::isContentAddressed(const Store & store) const
 
     else if (hasPrefix(ca, "fixed:")) {
         FileIngestionMethod recursive { ca.compare(6, 2, "r:") == 0 };
-        Hash hash(std::string(ca, static_cast<bool>(recursive) ? 8 : 6));
+        Hash hash(std::string(ca, recursive == FileIngestionMethod::Recursive ? 8 : 6));
         auto refs = cloneStorePathSet(references);
         bool hasSelfReference = false;
         if (refs.count(path)) {
@@ -864,27 +867,7 @@ std::pair<std::string, Store::Params> splitUriAndParams(const std::string & uri_
     Store::Params params;
     auto q = uri.find('?');
     if (q != std::string::npos) {
-        for (auto s : tokenizeString<Strings>(uri.substr(q + 1), "&")) {
-            auto e = s.find('=');
-            if (e != std::string::npos) {
-                auto value = s.substr(e + 1);
-                std::string decoded;
-                for (size_t i = 0; i < value.size(); ) {
-                    if (value[i] == '%') {
-                        if (i + 2 >= value.size())
-                            throw Error("invalid URI parameter '%s'", value);
-                        try {
-                            decoded += std::stoul(std::string(value, i + 1, 2), 0, 16);
-                            i += 3;
-                        } catch (...) {
-                            throw Error("invalid URI parameter '%s'", value);
-                        }
-                    } else
-                        decoded += value[i++];
-                }
-                params[s.substr(0, e)] = decoded;
-            }
-        }
+        params = decodeQuery(uri.substr(q + 1));
         uri = uri_.substr(0, q);
     }
     return {uri, params};

src/libstore/store-api.hh

@@ -28,6 +28,7 @@ MakeError(InvalidPath, Error);
 MakeError(Unsupported, Error);
 MakeError(SubstituteGone, Error);
 MakeError(SubstituterDisabled, Error);
+MakeError(NotInStore, Error);
 
 
 struct BasicDerivation;

src/libutil/ansicolor.hh

@@ -0,0 +1,15 @@
+#pragma once
+
+namespace nix {
+
+/* Some ANSI escape sequences. */
+#define ANSI_NORMAL "\e[0m"
+#define ANSI_BOLD "\e[1m"
+#define ANSI_FAINT "\e[2m"
+#define ANSI_ITALIC "\e[3m"
+#define ANSI_RED "\e[31;1m"
+#define ANSI_GREEN "\e[32;1m"
+#define ANSI_YELLOW "\e[33;1m"
+#define ANSI_BLUE "\e[34;1m"
+
+}

src/libutil/args.cc

@@ -3,16 +3,14 @@
 
 namespace nix {
 
-Args::FlagMaker Args::mkFlag()
-{
-    return FlagMaker(*this);
-}
-
-Args::FlagMaker::~FlagMaker()
+void Args::addFlag(Flag && flag_)
 {
+    auto flag = std::make_shared<Flag>(std::move(flag_));
+    if (flag->handler.arity != ArityAny)
+        assert(flag->handler.arity == flag->labels.size());
     assert(flag->longName != "");
-    args.longFlags[flag->longName] = flag;
-    if (flag->shortName) args.shortFlags[flag->shortName] = flag;
+    longFlags[flag->longName] = flag;
+    if (flag->shortName) shortFlags[flag->shortName] = flag;
 }
 
 void Args::parseCmdline(const Strings & _cmdline)
@@ -61,7 +59,7 @@ void Args::parseCmdline(const Strings & _cmdline)
 
 void Args::printHelp(const string & programName, std::ostream & out)
 {
-    std::cout << "Usage: " << programName << " <FLAGS>...";
+    std::cout << fmt(ANSI_BOLD "Usage:" ANSI_NORMAL " %s " ANSI_ITALIC "FLAGS..." ANSI_NORMAL, programName);
     for (auto & exp : expectedArgs) {
         std::cout << renderLabels({exp.label});
         // FIXME: handle arity > 1
@@ -72,11 +70,11 @@ void Args::printHelp(const string & programName, std::ostream & out)
 
     auto s = description();
     if (s != "")
-        std::cout << "\nSummary: " << s << ".\n";
+        std::cout << "\n" ANSI_BOLD "Summary:" ANSI_NORMAL " " << s << ".\n";
 
     if (longFlags.size()) {
         std::cout << "\n";
-        std::cout << "Flags:\n";
+        std::cout << ANSI_BOLD "Flags:" ANSI_NORMAL "\n";
         printFlags(out);
     }
 }
@@ -101,15 +99,14 @@ 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;
-        for (size_t n = 0 ; n < flag.arity; ++n) {
+        for (size_t n = 0 ; n < flag.handler.arity; ++n) {
             if (pos == end) {
-                if (flag.arity == ArityAny) break;
-                throw UsageError(format("flag '%1%' requires %2% argument(s)")
-                    % name % flag.arity);
+                if (flag.handler.arity == ArityAny) break;
+                throw UsageError("flag '%s' requires %d argument(s)", name, flag.handler.arity);
             }
             args.push_back(*pos++);
         }
-        flag.handler(std::move(args));
+        flag.handler.fun(std::move(args));
         return true;
     };
 
@@ -157,17 +154,18 @@ bool Args::processArgs(const Strings & args, bool finish)
     return res;
 }
 
-Args::FlagMaker & Args::FlagMaker::mkHashTypeFlag(HashType * ht)
+Args::Flag Args::Flag::mkHashTypeFlag(std::string && longName, HashType * ht)
 {
-    arity(1);
-    label("type");
-    description("hash algorithm ('md5', 'sha1', 'sha256', or 'sha512')");
-    handler([ht](std::string s) {
-        *ht = parseHashType(s);
-        if (*ht == htUnknown)
-            throw UsageError("unknown hash type '%1%'", s);
-    });
-    return *this;
+    return Flag {
+        .longName = std::move(longName),
+        .description = "hash algorithm ('md5', 'sha1', 'sha256', or 'sha512')",
+        .labels = {"hash-algo"},
+        .handler = {[ht](std::string s) {
+            *ht = parseHashType(s);
+            if (*ht == htUnknown)
+                throw UsageError("unknown hash type '%1%'", s);
+        }}
+    };
 }
 
 Strings argvToStrings(int argc, char * * argv)
@@ -183,7 +181,7 @@ std::string renderLabels(const Strings & labels)
     std::string res;
     for (auto label : labels) {
         for (auto & c : label) c = std::toupper(c);
-        res += " <" + label + ">";
+        res += " " ANSI_ITALIC + label + ANSI_NORMAL;
     }
     return res;
 }
@@ -192,10 +190,10 @@ void printTable(std::ostream & out, const Table2 & table)
 {
     size_t max = 0;
     for (auto & row : table)
-        max = std::max(max, row.first.size());
+        max = std::max(max, filterANSIEscapes(row.first, true).size());
     for (auto & row : table) {
         out << "  " << row.first
-            << std::string(max - row.first.size() + 2, ' ')
+            << std::string(max - filterANSIEscapes(row.first, true).size() + 2, ' ')
             << row.second << "\n";
     }
 }
@@ -206,8 +204,7 @@ void Command::printHelp(const string & programName, std::ostream & out)
 
     auto exs = examples();
     if (!exs.empty()) {
-        out << "\n";
-        out << "Examples:\n";
+        out << "\n" ANSI_BOLD "Examples:" ANSI_NORMAL "\n";
         for (auto & ex : exs)
             out << "\n"
                 << "  " << ex.description << "\n" // FIXME: wrap
@@ -223,49 +220,55 @@ MultiCommand::MultiCommand(const Commands & commands)
         auto i = commands.find(ss[0]);
         if (i == commands.end())
             throw UsageError("'%s' is not a recognised command", ss[0]);
-        command = i->second();
-        command->_name = ss[0];
+        command = {ss[0], i->second()};
     }});
+
+    categories[Command::catDefault] = "Available commands";
 }
 
 void MultiCommand::printHelp(const string & programName, std::ostream & out)
 {
     if (command) {
-        command->printHelp(programName + " " + command->name(), out);
+        command->second->printHelp(programName + " " + command->first, out);
         return;
     }
 
-    out << "Usage: " << programName << " <COMMAND> <FLAGS>... <ARGS>...\n";
+    out << fmt(ANSI_BOLD "Usage:" ANSI_NORMAL " %s " ANSI_ITALIC "COMMAND FLAGS... ARGS..." ANSI_NORMAL "\n", programName);
 
-    out << "\n";
-    out << "Common flags:\n";
+    out << "\n" ANSI_BOLD "Common flags:" ANSI_NORMAL "\n";
     printFlags(out);
 
-    out << "\n";
-    out << "Available commands:\n";
+    std::map<Command::Category, std::map<std::string, ref<Command>>> commandsByCategory;
 
-    Table2 table;
-    for (auto & i : commands) {
-        auto command = i.second();
-        command->_name = i.first;
-        auto descr = command->description();
-        if (!descr.empty())
-            table.push_back(std::make_pair(command->name(), descr));
+    for (auto & [name, commandFun] : commands) {
+        auto command = commandFun();
+        commandsByCategory[command->category()].insert_or_assign(name, command);
+    }
+
+    for (auto & [category, commands] : commandsByCategory) {
+        out << fmt("\n" ANSI_BOLD "%s:" ANSI_NORMAL "\n", categories[category]);
+
+        Table2 table;
+        for (auto & [name, command] : commands) {
+            auto descr = command->description();
+            if (!descr.empty())
+                table.push_back(std::make_pair(name, descr));
+        }
+        printTable(out, table);
     }
-    printTable(out, table);
 }
 
 bool MultiCommand::processFlag(Strings::iterator & pos, Strings::iterator end)
 {
     if (Args::processFlag(pos, end)) return true;
-    if (command && command->processFlag(pos, end)) return true;
+    if (command && command->second->processFlag(pos, end)) return true;
     return false;
 }
 
 bool MultiCommand::processArgs(const Strings & args, bool finish)
 {
     if (command)
-        return command->processArgs(args, finish);
+        return command->second->processArgs(args, finish);
     else
         return Args::processArgs(args, finish);
 }

src/libutil/args.hh

@@ -32,13 +32,59 @@ protected:
     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;
-        Strings labels;
-        size_t arity = 0;
-        std::function<void(std::vector<std::string>)> handler;
         std::string category;
+        Strings labels;
+        Handler handler;
+
+        static Flag mkHashTypeFlag(std::string && longName, HashType * ht);
     };
 
     std::map<std::string, Flag::ptr> longFlags;
@@ -65,49 +111,7 @@ protected:
 
 public:
 
-    class FlagMaker
-    {
-        Args & args;
-        Flag::ptr flag;
-        friend class Args;
-        FlagMaker(Args & args) : args(args), flag(std::make_shared<Flag>()) { }
-    public:
-        ~FlagMaker();
-        FlagMaker & longName(const std::string & s) { flag->longName = s; return *this; }
-        FlagMaker & shortName(char s) { flag->shortName = s; return *this; }
-        FlagMaker & description(const std::string & s) { flag->description = s; return *this; }
-        FlagMaker & label(const std::string & l) { flag->arity = 1; flag->labels = {l}; return *this; }
-        FlagMaker & labels(const Strings & ls) { flag->arity = ls.size(); flag->labels = ls; return *this; }
-        FlagMaker & arity(size_t arity) { flag->arity = arity; return *this; }
-        FlagMaker & handler(std::function<void(std::vector<std::string>)> handler) { flag->handler = handler; return *this; }
-        FlagMaker & handler(std::function<void()> handler) { flag->handler = [handler](std::vector<std::string>) { handler(); }; return *this; }
-        FlagMaker & handler(std::function<void(std::string)> handler) {
-            flag->arity = 1;
-            flag->handler = [handler](std::vector<std::string> ss) { handler(std::move(ss[0])); };
-            return *this;
-        }
-        FlagMaker & category(const std::string & s) { flag->category = s; return *this; }
-
-        template<class T>
-        FlagMaker & dest(T * dest)
-        {
-            flag->arity = 1;
-            flag->handler = [=](std::vector<std::string> ss) { *dest = ss[0]; };
-            return *this;
-        }
-
-        template<class T>
-        FlagMaker & set(T * dest, const T & val)
-        {
-            flag->arity = 0;
-            flag->handler = [=](std::vector<std::string> ss) { *dest = val; };
-            return *this;
-        }
-
-        FlagMaker & mkHashTypeFlag(HashType * ht);
-    };
-
-    FlagMaker mkFlag();
+    void addFlag(Flag && flag);
 
     /* Helper functions for constructing flags / positional
        arguments. */
@@ -116,13 +120,13 @@ public:
         const std::string & label, const std::string & description,
         std::function<void(std::string)> fun)
     {
-        mkFlag()
-            .shortName(shortName)
-            .longName(longName)
-            .labels({label})
-            .description(description)
-            .arity(1)
-            .handler([=](std::vector<std::string> ss) { fun(ss[0]); });
+        addFlag({
+            .longName = longName,
+            .shortName = shortName,
+            .description = description,
+            .labels = {label},
+            .handler = {[=](std::string s) { fun(s); }}
+        });
     }
 
     void mkFlag(char shortName, const std::string & name,
@@ -135,11 +139,12 @@ public:
     void mkFlag(char shortName, const std::string & longName, const std::string & description,
         T * dest, const T & value)
     {
-        mkFlag()
-            .shortName(shortName)
-            .longName(longName)
-            .description(description)
-            .handler([=](std::vector<std::string> ss) { *dest = value; });
+        addFlag({
+            .longName = longName,
+            .shortName = shortName,
+            .description = description,
+            .handler = {[=]() { *dest = value; }}
+        });
     }
 
     template<class I>
@@ -155,18 +160,18 @@ public:
     void mkFlag(char shortName, const std::string & longName,
         const std::string & description, std::function<void(I)> fun)
     {
-        mkFlag()
-            .shortName(shortName)
-            .longName(longName)
-            .labels({"N"})
-            .description(description)
-            .arity(1)
-            .handler([=](std::vector<std::string> ss) {
+        addFlag({
+            .longName = longName,
+            .shortName = shortName,
+            .description = description,
+            .labels = {"N"},
+            .handler = {[=](std::string s) {
                 I n;
-                if (!string2Int(ss[0], n))
+                if (!string2Int(s, n))
                     throw UsageError("flag '--%s' requires a integer argument", longName);
                 fun(n);
-            });
+            }}
+        });
     }
 
     /* Expect a string argument. */
@@ -192,17 +197,10 @@ public:
    run() method. */
 struct Command : virtual Args
 {
-private:
-    std::string _name;
-
     friend class MultiCommand;
 
-public:
-
     virtual ~Command() { }
 
-    std::string name() { return _name; }
-
     virtual void prepare() { };
     virtual void run() = 0;
 
@@ -216,6 +214,12 @@ public:
 
     virtual Examples examples() { return Examples(); }
 
+    typedef int Category;
+
+    static constexpr Category catDefault = 0;
+
+    virtual Category category() { return catDefault; }
+
     void printHelp(const string & programName, std::ostream & out) override;
 };
 
@@ -228,7 +232,10 @@ class MultiCommand : virtual Args
 public:
     Commands commands;
 
-    std::shared_ptr<Command> command;
+    std::map<Command::Category, std::string> categories;
+
+    // Selected command, if any.
+    std::optional<std::pair<std::string, ref<Command>>> command;
 
     MultiCommand(const Commands & commands);
 

src/libutil/config.cc

@@ -177,12 +177,13 @@ void BaseSetting<T>::toJSON(JSONPlaceholder & out)
 template<typename T>
 void BaseSetting<T>::convertToArg(Args & args, const std::string & category)
 {
-    args.mkFlag()
-        .longName(name)
-        .description(description)
-        .arity(1)
-        .handler([=](std::vector<std::string> ss) { overriden = true; set(ss[0]); })
-        .category(category);
+    args.addFlag({
+        .longName = name,
+        .description = description,
+        .category = category,
+        .labels = {"value"},
+        .handler = {[=](std::string s) { overriden = true; set(s); }},
+    });
 }
 
 template<> void BaseSetting<std::string>::set(const std::string & str)
@@ -227,16 +228,18 @@ template<> std::string BaseSetting<bool>::to_string() const
 
 template<> void BaseSetting<bool>::convertToArg(Args & args, const std::string & category)
 {
-    args.mkFlag()
-        .longName(name)
-        .description(description)
-        .handler([=](std::vector<std::string> ss) { override(true); })
-        .category(category);
-    args.mkFlag()
-        .longName("no-" + name)
-        .description(description)
-        .handler([=](std::vector<std::string> ss) { override(false); })
-        .category(category);
+    args.addFlag({
+        .longName = name,
+        .description = description,
+        .category = category,
+        .handler = {[=]() { override(true); }}
+    });
+    args.addFlag({
+        .longName = "no-" + name,
+        .description = description,
+        .category = category,
+        .handler = {[=]() { override(false); }}
+    });
 }
 
 template<> void BaseSetting<Strings>::set(const std::string & str)

src/libutil/error.cc

@@ -0,0 +1,146 @@
+#include "error.hh"
+
+#include <iostream>
+#include <optional>
+
+namespace nix
+{
+
+std::optional<string> ErrorInfo::programName = std::nullopt;
+
+std::ostream& operator<<(std::ostream &os, const hintformat &hf)
+{
+    return os << hf.str();
+}
+
+string showErrPos(const ErrPos &errPos)
+{
+    if (errPos.column > 0) {
+        return fmt("(%1%:%2%)", errPos.lineNumber, errPos.column);
+    } else {
+        return fmt("(%1%)", errPos.lineNumber);
+    };
+}
+
+void printCodeLines(const string &prefix, const NixCode &nixCode)
+{
+    // previous line of code.
+    if (nixCode.prevLineOfCode.has_value()) {
+        std::cout << fmt("%1% %|2$5d|| %3%",
+                         prefix,
+                         (nixCode.errPos.lineNumber - 1),
+                         *nixCode.prevLineOfCode)
+                  << std::endl;
+    }
+
+    // line of code containing the error.%2$+5d%
+    std::cout << fmt("%1% %|2$5d|| %3%",
+                     prefix,
+                     (nixCode.errPos.lineNumber),
+                     nixCode.errLineOfCode)
+              << std::endl;
+
+    // error arrows for the column range.
+    if (nixCode.errPos.column > 0) {
+        int start = nixCode.errPos.column;
+        std::string spaces;
+        for (int i = 0; i < start; ++i) {
+            spaces.append(" ");
+        }
+
+        std::string arrows("^");
+
+        std::cout << fmt("%1%      |%2%" ANSI_RED "%3%" ANSI_NORMAL,
+                         prefix,
+                         spaces,
+                         arrows) << std::endl;
+    }
+
+    // next line of code.
+    if (nixCode.nextLineOfCode.has_value()) {
+        std::cout << fmt("%1% %|2$5d|| %3%",
+                         prefix,
+                         (nixCode.errPos.lineNumber + 1),
+                         *nixCode.nextLineOfCode)
+                  << std::endl;
+    }
+}
+
+void printErrorInfo(const ErrorInfo &einfo)
+{
+    int errwidth = 80;
+    string prefix = "    ";
+
+    string levelString;
+    switch (einfo.level) {
+    case ErrLevel::elError: {
+        levelString = ANSI_RED;
+        levelString += "error:";
+        levelString += ANSI_NORMAL;
+        break;
+    }
+    case ErrLevel::elWarning: {
+        levelString = ANSI_YELLOW;
+        levelString += "warning:";
+        levelString += ANSI_NORMAL;
+        break;
+    }
+    default: {
+        levelString = fmt("invalid error level: %1%", einfo.level);
+        break;
+    }
+    }
+
+    int ndl = prefix.length() + levelString.length() + 3 + einfo.name.length() + einfo.programName.value_or("").length();
+    int dashwidth = ndl > (errwidth - 3) ? 3 : errwidth - ndl;
+
+    string dashes;
+    for (int i = 0; i < dashwidth; ++i)
+        dashes.append("-");
+
+    // divider.
+    std::cout << fmt("%1%%2%" ANSI_BLUE " %3% %4% %5% %6%" ANSI_NORMAL,
+                     prefix,
+                     levelString,
+                     "---",
+                     einfo.name,
+                     dashes,
+                     einfo.programName.value_or(""))
+              << std::endl;
+
+    // filename.
+    if (einfo.nixCode.has_value()) {
+        if (einfo.nixCode->errPos.nixFile != "") {
+            string eline = einfo.nixCode->errLineOfCode != ""
+                           ? string(" ") + showErrPos(einfo.nixCode->errPos)
+                           : "";
+
+            std::cout << fmt("%1%in file: " ANSI_BLUE "%2%%3%" ANSI_NORMAL,
+                             prefix, 
+                             einfo.nixCode->errPos.nixFile,
+                             eline) << std::endl;
+            std::cout << prefix << std::endl;
+        } else {
+            std::cout << fmt("%1%from command line argument", prefix) << std::endl;
+            std::cout << prefix << std::endl;
+        }
+    }
+
+    // description
+    std::cout << prefix << einfo.description << std::endl;
+    std::cout << prefix << std::endl;
+
+    // lines of code.
+    if (einfo.nixCode->errLineOfCode != "") {
+        printCodeLines(prefix, *einfo.nixCode);
+        std::cout << prefix << std::endl;
+    }
+
+    // hint
+    if (einfo.hint.has_value()) {
+        std::cout << prefix << *einfo.hint << std::endl;
+        std::cout << prefix << std::endl;
+    }
+}
+
+}

src/libutil/error.hh

@@ -0,0 +1,121 @@
+#ifndef error_hh
+#define error_hh
+
+#include "ansicolor.hh"
+#include <string>
+#include <optional>
+#include <iostream>
+#include "types.hh"
+
+namespace nix
+{
+
+typedef enum {
+    elWarning,
+    elError
+} ErrLevel;
+
+struct ErrPos
+{
+    int lineNumber;
+    int column;
+    string nixFile;
+
+    template <class P>
+    ErrPos& operator=(const P &pos)
+    {
+        lineNumber = pos.line;
+        column = pos.column;
+        nixFile = pos.file;
+        return *this;
+    }
+
+    template <class P>
+    ErrPos(const P &p)
+    {
+      *this = p;
+    }
+};
+
+struct NixCode
+{
+    ErrPos errPos;
+    std::optional<string> prevLineOfCode;
+    string errLineOfCode;
+    std::optional<string> nextLineOfCode;
+};
+
+// ----------------------------------------------------------------
+// format function for hints.  same as fmt, except templated values
+// are always in yellow.
+
+template <class T>
+struct yellowify
+{
+    yellowify(T &s) : value(s) {}
+    T &value;
+};
+
+template <class T>
+std::ostream& operator<<(std::ostream &out, const yellowify<T> &y)
+{
+    return out << ANSI_YELLOW << y.value << ANSI_NORMAL;
+}
+
+class hintformat
+{
+public:
+    hintformat(string format) :fmt(format)
+    {
+        fmt.exceptions(boost::io::all_error_bits ^ boost::io::too_many_args_bit);
+    }
+    template<class T>
+    hintformat& operator%(const T &value)
+    {
+        fmt % yellowify(value);
+        return *this;
+    }
+
+    std::string str() const
+    {
+        return fmt.str();
+    }
+
+    template <typename U>
+    friend class AddHint;
+private:
+    format fmt;
+};
+
+std::ostream& operator<<(std::ostream &os, const hintformat &hf);
+
+template<typename... Args>
+inline hintformat hintfmt(const std::string & fs, const Args & ... args)
+{
+    hintformat f(fs);
+    formatHelper(f, args...);
+    return f;
+}
+
+// -------------------------------------------------
+// ErrorInfo.
+struct ErrorInfo
+{
+    ErrLevel level;
+    string name;
+    string description;
+    std::optional<hintformat> hint;
+    std::optional<NixCode> nixCode;
+
+    static std::optional<string> programName;
+};
+
+// --------------------------------------------------------
+// error printing
+
+// just to cout for now.
+void printErrorInfo(const ErrorInfo &einfo);
+
+}
+
+#endif

src/libutil/logging.cc

@@ -3,6 +3,7 @@
 
 #include <atomic>
 #include <nlohmann/json.hpp>
+#include <iostream>
 
 namespace nix {
 
@@ -24,6 +25,11 @@ void Logger::warn(const std::string & msg)
     log(lvlWarn, ANSI_YELLOW "warning:" ANSI_NORMAL " " + msg);
 }
 
+void Logger::writeToStdout(std::string_view s)
+{
+    std::cout << s << "\n";
+}
+
 class SimpleLogger : public Logger
 {
 public:
@@ -198,7 +204,7 @@ bool handleJSONLogMessage(const std::string & msg,
 
         if (action == "start") {
             auto type = (ActivityType) json["type"];
-            if (trusted || type == actDownload)
+            if (trusted || type == actFileTransfer)
                 activities.emplace(std::piecewise_construct,
                     std::forward_as_tuple(json["id"]),
                     std::forward_as_tuple(*logger, (Verbosity) json["level"], type,

src/libutil/logging.hh

@@ -17,7 +17,7 @@ typedef enum {
 typedef enum {
     actUnknown = 0,
     actCopyPath = 100,
-    actDownload = 101,
+    actFileTransfer = 101,
     actRealise = 102,
     actCopyPaths = 103,
     actBuilds = 104,
@@ -78,6 +78,16 @@ public:
     virtual void stopActivity(ActivityId act) { };
 
     virtual void result(ActivityId act, ResultType type, const Fields & fields) { };
+
+    virtual void writeToStdout(std::string_view s);
+
+    template<typename... Args>
+    inline void stdout(const std::string & fs, const Args & ... args)
+    {
+        boost::format f(fs);
+        formatHelper(f, args...);
+        writeToStdout(f.str());
+    }
 };
 
 ActivityId getCurActivity();

src/libutil/serialise.hh

@@ -148,6 +148,9 @@ struct StringSink : Sink
 {
     ref<std::string> s;
     StringSink() : s(make_ref<std::string>()) { };
+    explicit StringSink(const size_t reservedSize) : s(make_ref<std::string>()) {
+      s->reserve(reservedSize);
+    };
     StringSink(ref<std::string> s) : s(s) { };
     void operator () (const unsigned char * data, size_t len) override;
 };

src/libutil/tests/local.mk

@@ -0,0 +1,15 @@
+check: libutil-tests_RUN
+
+programs += libutil-tests
+
+libutil-tests_DIR := $(d)
+
+libutil-tests_INSTALL_DIR :=
+
+libutil-tests_SOURCES := $(wildcard $(d)/*.cc)
+
+libutil-tests_CXXFLAGS += -I src/libutil
+
+libutil-tests_LIBS = libutil
+
+libutil-tests_LDFLAGS := $(GTEST_LIBS)

src/libutil/tests/tests.cc

@@ -0,0 +1,589 @@
+#include "util.hh"
+#include "types.hh"
+
+#include <gtest/gtest.h>
+
+namespace nix {
+
+/* ----------- tests for util.hh ------------------------------------------------*/
+
+    /* ----------------------------------------------------------------------------
+     * absPath
+     * --------------------------------------------------------------------------*/
+
+    TEST(absPath, doesntChangeRoot) {
+        auto p = absPath("/");
+
+        ASSERT_EQ(p, "/");
+    }
+
+
+
+
+    TEST(absPath, turnsEmptyPathIntoCWD) {
+        char cwd[PATH_MAX+1];
+        auto p = absPath("");
+
+        ASSERT_EQ(p, getcwd((char*)&cwd, PATH_MAX));
+    }
+
+    TEST(absPath, usesOptionalBasePathWhenGiven) {
+        char _cwd[PATH_MAX+1];
+        char* cwd = getcwd((char*)&_cwd, PATH_MAX);
+
+        auto p = absPath("", cwd);
+
+        ASSERT_EQ(p, cwd);
+    }
+
+    TEST(absPath, isIdempotent) {
+        char _cwd[PATH_MAX+1];
+        char* cwd = getcwd((char*)&_cwd, PATH_MAX);
+        auto p1 = absPath(cwd);
+        auto p2 = absPath(p1);
+
+        ASSERT_EQ(p1, p2);
+    }
+
+
+    TEST(absPath, pathIsCanonicalised) {
+        auto path = "/some/path/with/trailing/dot/.";
+        auto p1 = absPath(path);
+        auto p2 = absPath(p1);
+
+        ASSERT_EQ(p1, "/some/path/with/trailing/dot");
+        ASSERT_EQ(p1, p2);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * canonPath
+     * --------------------------------------------------------------------------*/
+
+    TEST(canonPath, removesTrailingSlashes) {
+        auto path = "/this/is/a/path//";
+        auto p = canonPath(path);
+
+        ASSERT_EQ(p, "/this/is/a/path");
+    }
+
+    TEST(canonPath, removesDots) {
+        auto path = "/this/./is/a/path/./";
+        auto p = canonPath(path);
+
+        ASSERT_EQ(p, "/this/is/a/path");
+    }
+
+    TEST(canonPath, removesDots2) {
+        auto path = "/this/a/../is/a////path/foo/..";
+        auto p = canonPath(path);
+
+        ASSERT_EQ(p, "/this/is/a/path");
+    }
+
+    TEST(canonPath, requiresAbsolutePath) {
+        ASSERT_ANY_THROW(canonPath("."));
+        ASSERT_ANY_THROW(canonPath(".."));
+        ASSERT_ANY_THROW(canonPath("../"));
+        ASSERT_DEATH({ canonPath(""); }, "path != \"\"");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * dirOf
+     * --------------------------------------------------------------------------*/
+
+    TEST(dirOf, returnsEmptyStringForRoot) {
+        auto p = dirOf("/");
+
+        ASSERT_EQ(p, "/");
+    }
+
+    TEST(dirOf, returnsFirstPathComponent) {
+        auto p1 = dirOf("/dir/");
+        ASSERT_EQ(p1, "/dir");
+        auto p2 = dirOf("/dir");
+        ASSERT_EQ(p2, "/");
+        auto p3 = dirOf("/dir/..");
+        ASSERT_EQ(p3, "/dir");
+        auto p4 = dirOf("/dir/../");
+        ASSERT_EQ(p4, "/dir/..");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * baseNameOf
+     * --------------------------------------------------------------------------*/
+
+    TEST(baseNameOf, emptyPath) {
+        auto p1 = baseNameOf("");
+        ASSERT_EQ(p1, "");
+    }
+
+    TEST(baseNameOf, pathOnRoot) {
+        auto p1 = baseNameOf("/dir");
+        ASSERT_EQ(p1, "dir");
+    }
+
+    TEST(baseNameOf, relativePath) {
+        auto p1 = baseNameOf("dir/foo");
+        ASSERT_EQ(p1, "foo");
+    }
+
+    TEST(baseNameOf, pathWithTrailingSlashRoot) {
+        auto p1 = baseNameOf("/");
+        ASSERT_EQ(p1, "");
+    }
+
+    TEST(baseNameOf, trailingSlash) {
+        auto p1 = baseNameOf("/dir/");
+        ASSERT_EQ(p1, "dir");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * isInDir
+     * --------------------------------------------------------------------------*/
+
+    TEST(isInDir, trivialCase) {
+        auto p1 = isInDir("/foo/bar", "/foo");
+        ASSERT_EQ(p1, true);
+    }
+
+    TEST(isInDir, notInDir) {
+        auto p1 = isInDir("/zes/foo/bar", "/foo");
+        ASSERT_EQ(p1, false);
+    }
+
+    // XXX: hm, bug or feature? :) Looking at the implementation
+    // this might be problematic.
+    TEST(isInDir, emptyDir) {
+        auto p1 = isInDir("/zes/foo/bar", "");
+        ASSERT_EQ(p1, true);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * isDirOrInDir
+     * --------------------------------------------------------------------------*/
+
+    TEST(isDirOrInDir, trueForSameDirectory) {
+        ASSERT_EQ(isDirOrInDir("/nix", "/nix"), true);
+        ASSERT_EQ(isDirOrInDir("/", "/"), true);
+    }
+
+    TEST(isDirOrInDir, trueForEmptyPaths) {
+        ASSERT_EQ(isDirOrInDir("", ""), true);
+    }
+
+    TEST(isDirOrInDir, falseForDisjunctPaths) {
+        ASSERT_EQ(isDirOrInDir("/foo", "/bar"), false);
+    }
+
+    TEST(isDirOrInDir, relativePaths) {
+        ASSERT_EQ(isDirOrInDir("/foo/..", "/foo"), true);
+    }
+
+    // XXX: while it is possible to use "." or ".." in the
+    // first argument this doesn't seem to work in the second.
+    TEST(isDirOrInDir, DISABLED_shouldWork) {
+        ASSERT_EQ(isDirOrInDir("/foo/..", "/foo/."), true);
+
+    }
+
+    /* ----------------------------------------------------------------------------
+     * pathExists
+     * --------------------------------------------------------------------------*/
+
+    TEST(pathExists, rootExists) {
+        ASSERT_TRUE(pathExists("/"));
+    }
+
+    TEST(pathExists, cwdExists) {
+        ASSERT_TRUE(pathExists("."));
+    }
+
+    TEST(pathExists, bogusPathDoesNotExist) {
+        ASSERT_FALSE(pathExists("/home/schnitzel/darmstadt/pommes"));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * concatStringsSep
+     * --------------------------------------------------------------------------*/
+
+    TEST(concatStringsSep, buildCommaSeparatedString) {
+        Strings strings;
+        strings.push_back("this");
+        strings.push_back("is");
+        strings.push_back("great");
+
+        ASSERT_EQ(concatStringsSep(",", strings), "this,is,great");
+    }
+
+    TEST(concatStringsSep, buildStringWithEmptySeparator) {
+        Strings strings;
+        strings.push_back("this");
+        strings.push_back("is");
+        strings.push_back("great");
+
+        ASSERT_EQ(concatStringsSep("", strings), "thisisgreat");
+    }
+
+    TEST(concatStringsSep, buildSingleString) {
+        Strings strings;
+        strings.push_back("this");
+
+        ASSERT_EQ(concatStringsSep(",", strings), "this");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * hasPrefix
+     * --------------------------------------------------------------------------*/
+
+    TEST(hasPrefix, emptyStringHasNoPrefix) {
+        ASSERT_FALSE(hasPrefix("", "foo"));
+    }
+
+    TEST(hasPrefix, emptyStringIsAlwaysPrefix) {
+        ASSERT_TRUE(hasPrefix("foo", ""));
+        ASSERT_TRUE(hasPrefix("jshjkfhsadf", ""));
+    }
+
+    TEST(hasPrefix, trivialCase) {
+        ASSERT_TRUE(hasPrefix("foobar", "foo"));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * hasSuffix
+     * --------------------------------------------------------------------------*/
+
+    TEST(hasSuffix, emptyStringHasNoSuffix) {
+        ASSERT_FALSE(hasSuffix("", "foo"));
+    }
+
+    TEST(hasSuffix, trivialCase) {
+        ASSERT_TRUE(hasSuffix("foo", "foo"));
+        ASSERT_TRUE(hasSuffix("foobar", "bar"));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * base64Encode
+     * --------------------------------------------------------------------------*/
+
+    TEST(base64Encode, emptyString) {
+        ASSERT_EQ(base64Encode(""), "");
+    }
+
+    TEST(base64Encode, encodesAString) {
+        ASSERT_EQ(base64Encode("quod erat demonstrandum"), "cXVvZCBlcmF0IGRlbW9uc3RyYW5kdW0=");
+    }
+
+    TEST(base64Encode, encodeAndDecode) {
+        auto s = "quod erat demonstrandum";
+        auto encoded = base64Encode(s);
+        auto decoded = base64Decode(encoded);
+
+        ASSERT_EQ(decoded, s);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * base64Decode
+     * --------------------------------------------------------------------------*/
+
+    TEST(base64Decode, emptyString) {
+        ASSERT_EQ(base64Decode(""), "");
+    }
+
+    TEST(base64Decode, decodeAString) {
+        ASSERT_EQ(base64Decode("cXVvZCBlcmF0IGRlbW9uc3RyYW5kdW0="), "quod erat demonstrandum");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * toLower
+     * --------------------------------------------------------------------------*/
+
+    TEST(toLower, emptyString) {
+        ASSERT_EQ(toLower(""), "");
+    }
+
+    TEST(toLower, nonLetters) {
+        auto s = "!@(*$#)(@#=\\234_";
+        ASSERT_EQ(toLower(s), s);
+    }
+
+    // std::tolower() doesn't handle unicode characters. In the context of
+    // store paths this isn't relevant but doesn't hurt to record this behavior
+    // here.
+    TEST(toLower, umlauts) {
+        auto s = "ÄÖÜ";
+        ASSERT_EQ(toLower(s), "ÄÖÜ");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * string2Float
+     * --------------------------------------------------------------------------*/
+
+    TEST(string2Float, emptyString) {
+        double n;
+        ASSERT_EQ(string2Float("", n), false);
+    }
+
+    TEST(string2Float, trivialConversions) {
+        double n;
+        ASSERT_EQ(string2Float("1.0", n), true);
+        ASSERT_EQ(n, 1.0);
+
+        ASSERT_EQ(string2Float("0.0", n), true);
+        ASSERT_EQ(n, 0.0);
+
+        ASSERT_EQ(string2Float("-100.25", n), true);
+        ASSERT_EQ(n, (-100.25));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * string2Int
+     * --------------------------------------------------------------------------*/
+
+    TEST(string2Int, emptyString) {
+        double n;
+        ASSERT_EQ(string2Int("", n), false);
+    }
+
+    TEST(string2Int, trivialConversions) {
+        double n;
+        ASSERT_EQ(string2Int("1", n), true);
+        ASSERT_EQ(n, 1);
+
+        ASSERT_EQ(string2Int("0", n), true);
+        ASSERT_EQ(n, 0);
+
+        ASSERT_EQ(string2Int("-100", n), true);
+        ASSERT_EQ(n, (-100));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * statusOk
+     * --------------------------------------------------------------------------*/
+
+    TEST(statusOk, zeroIsOk) {
+        ASSERT_EQ(statusOk(0), true);
+        ASSERT_EQ(statusOk(1), false);
+    }
+
+
+    /* ----------------------------------------------------------------------------
+     * rewriteStrings
+     * --------------------------------------------------------------------------*/
+
+    TEST(rewriteStrings, emptyString) {
+        StringMap rewrites;
+        rewrites["this"] = "that";
+
+        ASSERT_EQ(rewriteStrings("", rewrites), "");
+    }
+
+    TEST(rewriteStrings, emptyRewrites) {
+        StringMap rewrites;
+
+        ASSERT_EQ(rewriteStrings("this and that", rewrites), "this and that");
+    }
+
+    TEST(rewriteStrings, successfulRewrite) {
+        StringMap rewrites;
+        rewrites["this"] = "that";
+
+        ASSERT_EQ(rewriteStrings("this and that", rewrites), "that and that");
+    }
+
+    TEST(rewriteStrings, doesntOccur) {
+        StringMap rewrites;
+        rewrites["foo"] = "bar";
+
+        ASSERT_EQ(rewriteStrings("this and that", rewrites), "this and that");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * replaceStrings
+     * --------------------------------------------------------------------------*/
+
+    TEST(replaceStrings, emptyString) {
+        ASSERT_EQ(replaceStrings("", "this", "that"), "");
+        ASSERT_EQ(replaceStrings("this and that", "", ""), "this and that");
+    }
+
+    TEST(replaceStrings, successfulReplace) {
+        ASSERT_EQ(replaceStrings("this and that", "this", "that"), "that and that");
+    }
+
+    TEST(replaceStrings, doesntOccur) {
+        ASSERT_EQ(replaceStrings("this and that", "foo", "bar"), "this and that");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * trim
+     * --------------------------------------------------------------------------*/
+
+    TEST(trim, emptyString) {
+        ASSERT_EQ(trim(""), "");
+    }
+
+    TEST(trim, removesWhitespace) {
+        ASSERT_EQ(trim("foo"), "foo");
+        ASSERT_EQ(trim("     foo "), "foo");
+        ASSERT_EQ(trim("     foo bar baz"), "foo bar baz");
+        ASSERT_EQ(trim("     \t foo bar baz\n"), "foo bar baz");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * chomp
+     * --------------------------------------------------------------------------*/
+
+    TEST(chomp, emptyString) {
+        ASSERT_EQ(chomp(""), "");
+    }
+
+    TEST(chomp, removesWhitespace) {
+        ASSERT_EQ(chomp("foo"), "foo");
+        ASSERT_EQ(chomp("foo "), "foo");
+        ASSERT_EQ(chomp(" foo "), " foo");
+        ASSERT_EQ(chomp(" foo bar baz  "), " foo bar baz");
+        ASSERT_EQ(chomp("\t foo bar baz\n"), "\t foo bar baz");
+    }
+
+    /* ----------------------------------------------------------------------------
+     * quoteStrings
+     * --------------------------------------------------------------------------*/
+
+    TEST(quoteStrings, empty) {
+        Strings s = { };
+        Strings expected = { };
+
+        ASSERT_EQ(quoteStrings(s), expected);
+    }
+
+    TEST(quoteStrings, emptyStrings) {
+        Strings s = { "", "", "" };
+        Strings expected = { "''", "''", "''" };
+        ASSERT_EQ(quoteStrings(s), expected);
+
+    }
+
+    TEST(quoteStrings, trivialQuote) {
+        Strings s = { "foo", "bar", "baz" };
+        Strings expected = { "'foo'", "'bar'", "'baz'" };
+
+        ASSERT_EQ(quoteStrings(s), expected);
+    }
+
+    TEST(quoteStrings, quotedStrings) {
+        Strings s = { "'foo'", "'bar'", "'baz'" };
+        Strings expected = { "''foo''", "''bar''", "''baz''" };
+
+        ASSERT_EQ(quoteStrings(s), expected);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * tokenizeString
+     * --------------------------------------------------------------------------*/
+
+    TEST(tokenizeString, empty) {
+        Strings expected = { };
+
+        ASSERT_EQ(tokenizeString<Strings>(""), expected);
+    }
+
+    TEST(tokenizeString, tokenizeSpacesWithDefaults) {
+        auto s = "foo bar baz";
+        Strings expected = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s), expected);
+    }
+
+    TEST(tokenizeString, tokenizeTabsWithDefaults) {
+        auto s = "foo\tbar\tbaz";
+        Strings expected = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s), expected);
+    }
+
+    TEST(tokenizeString, tokenizeTabsSpacesWithDefaults) {
+        auto s = "foo\t bar\t baz";
+        Strings expected = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s), expected);
+    }
+
+    TEST(tokenizeString, tokenizeTabsSpacesNewlineWithDefaults) {
+        auto s = "foo\t\n bar\t\n baz";
+        Strings expected = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s), expected);
+    }
+
+    TEST(tokenizeString, tokenizeTabsSpacesNewlineRetWithDefaults) {
+        auto s = "foo\t\n\r bar\t\n\r baz";
+        Strings expected = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s), expected);
+
+        auto s2 = "foo \t\n\r bar \t\n\r baz";
+        Strings expected2 = { "foo", "bar", "baz" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s2), expected2);
+    }
+
+    TEST(tokenizeString, tokenizeWithCustomSep) {
+        auto s = "foo\n,bar\n,baz\n";
+        Strings expected = { "foo\n", "bar\n", "baz\n" };
+
+        ASSERT_EQ(tokenizeString<Strings>(s, ","), expected);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * get
+     * --------------------------------------------------------------------------*/
+
+    TEST(get, emptyContainer) {
+        StringMap s = { };
+        auto expected = std::nullopt;
+
+        ASSERT_EQ(get(s, "one"), expected);
+    }
+
+    TEST(get, getFromContainer) {
+        StringMap s;
+        s["one"] = "yi";
+        s["two"] = "er";
+        auto expected = "yi";
+
+        ASSERT_EQ(get(s, "one"), expected);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * filterANSIEscapes
+     * --------------------------------------------------------------------------*/
+
+    TEST(filterANSIEscapes, emptyString) {
+        auto s = "";
+        auto expected = "";
+
+        ASSERT_EQ(filterANSIEscapes(s), expected);
+    }
+
+    TEST(filterANSIEscapes, doesntChangePrintableChars) {
+        auto s = "09 2q304ruyhr slk2-19024 kjsadh sar f";
+
+        ASSERT_EQ(filterANSIEscapes(s), s);
+    }
+
+    TEST(filterANSIEscapes, filtersColorCodes) {
+        auto s = "\u001b[30m A \u001b[31m B \u001b[32m C \u001b[33m D \u001b[0m";
+
+        ASSERT_EQ(filterANSIEscapes(s, true, 2), " A" );
+        ASSERT_EQ(filterANSIEscapes(s, true, 3), " A " );
+        ASSERT_EQ(filterANSIEscapes(s, true, 4), " A  " );
+        ASSERT_EQ(filterANSIEscapes(s, true, 5), " A  B" );
+        ASSERT_EQ(filterANSIEscapes(s, true, 8), " A  B  C" );
+    }
+
+    TEST(filterANSIEscapes, expandsTabs) {
+        auto s = "foo\tbar\tbaz";
+
+        ASSERT_EQ(filterANSIEscapes(s, true), "foo     bar     baz" );
+    }
+}

src/libutil/tests/url.cc

@@ -0,0 +1,266 @@
+#include "url.hh"
+#include <gtest/gtest.h>
+
+namespace nix {
+
+/* ----------- tests for url.hh --------------------------------------------------*/
+
+    string print_map(std::map<string, string> m) {
+        std::map<string, string>::iterator it;
+        string s = "{ ";
+        for (it = m.begin(); it != m.end(); ++it) {
+            s += "{ ";
+            s += it->first;
+            s += " = ";
+            s += it->second;
+            s += " } ";
+        }
+        s += "}";
+        return s;
+    }
+
+
+    std::ostream& operator<<(std::ostream& os, const ParsedURL& p) {
+        return os << "\n"
+                  << "url:       " << p.url << "\n"
+                  << "base:      " << p.base << "\n"
+                  << "scheme:    " << p.scheme << "\n"
+                  << "authority: " << p.authority.value() << "\n"
+                  << "path:      " << p.path << "\n"
+                  << "query:     " << print_map(p.query) << "\n"
+                  << "fragment:  " << p.fragment << "\n";
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrl) {
+        auto s = "http://www.example.org/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://www.example.org/file.tar.gz",
+            .base = "http://www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpsUrl) {
+        auto s = "https://www.example.org/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "https://www.example.org/file.tar.gz",
+            .base = "https://www.example.org/file.tar.gz",
+            .scheme = "https",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrlWithQueryAndFragment) {
+        auto s = "https://www.example.org/file.tar.gz?download=fast&when=now#hello";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "https://www.example.org/file.tar.gz",
+            .base = "https://www.example.org/file.tar.gz",
+            .scheme = "https",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "download", "fast" }, { "when", "now" } },
+            .fragment = "hello",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesSimpleHttpUrlWithComplexFragment) {
+        auto s = "http://www.example.org/file.tar.gz?field=value#?foo=bar%23";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://www.example.org/file.tar.gz",
+            .base = "http://www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "www.example.org",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "field", "value" } },
+            .fragment = "?foo=bar#",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+
+    TEST(parseURL, parseIPv4Address) {
+        auto s = "http://127.0.0.1:8080/file.tar.gz?download=fast&when=now#hello";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://127.0.0.1:8080/file.tar.gz",
+            .base = "https://127.0.0.1:8080/file.tar.gz",
+            .scheme = "http",
+            .authority = "127.0.0.1:8080",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { { "download", "fast" }, { "when", "now" } },
+            .fragment = "hello",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parseIPv6Address) {
+        auto s = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .base = "http://[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .scheme = "http",
+            .authority = "[2a02:8071:8192:c100:311d:192d:81ac:11ea]:8080",
+            .path = "",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+
+    }
+
+    TEST(parseURL, parseEmptyQueryParams) {
+        auto s = "http://127.0.0.1:8080/file.tar.gz?&&&&&";
+        auto parsed = parseURL(s);
+        ASSERT_EQ(parsed.query, (StringMap) { });
+    }
+
+    TEST(parseURL, parseUserPassword) {
+        auto s = "http://user:pass@www.example.org:8080/file.tar.gz";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "http://user:pass@www.example.org/file.tar.gz",
+            .base = "http://user:pass@www.example.org/file.tar.gz",
+            .scheme = "http",
+            .authority = "user:pass@www.example.org:8080",
+            .path = "/file.tar.gz",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parseFileURLWithQueryAndFragment) {
+        auto s = "file:///none/of/your/business";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "",
+            .base = "",
+            .scheme = "file",
+            .authority = "",
+            .path = "/none/of/your/business",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+
+    }
+
+    TEST(parseURL, parsedUrlsIsEqualToItself) {
+        auto s = "http://www.example.org/file.tar.gz";
+        auto url = parseURL(s);
+
+        ASSERT_TRUE(url == url);
+    }
+
+    TEST(parseURL, parseFTPUrl) {
+        auto s = "ftp://ftp.nixos.org/downloads/nixos.iso";
+        auto parsed = parseURL(s);
+
+        ParsedURL expected {
+            .url = "ftp://ftp.nixos.org/downloads/nixos.iso",
+            .base = "ftp://ftp.nixos.org/downloads/nixos.iso",
+            .scheme = "ftp",
+            .authority = "ftp.nixos.org",
+            .path = "/downloads/nixos.iso",
+            .query = (StringMap) { },
+            .fragment = "",
+        };
+
+        ASSERT_EQ(parsed, expected);
+    }
+
+    TEST(parseURL, parsesAnythingInUriFormat) {
+        auto s = "whatever://github.com/NixOS/nixpkgs.git";
+        auto parsed = parseURL(s);
+    }
+
+    TEST(parseURL, parsesAnythingInUriFormatWithoutDoubleSlash) {
+        auto s = "whatever:github.com/NixOS/nixpkgs.git";
+        auto parsed = parseURL(s);
+    }
+
+    TEST(parseURL, emptyStringIsInvalidURL) {
+        ASSERT_THROW(parseURL(""), Error);
+    }
+
+    /* ----------------------------------------------------------------------------
+     * decodeQuery
+     * --------------------------------------------------------------------------*/
+
+    TEST(decodeQuery, emptyStringYieldsEmptyMap) {
+        auto d = decodeQuery("");
+        ASSERT_EQ(d, (StringMap) { });
+    }
+
+    TEST(decodeQuery, simpleDecode) {
+        auto d = decodeQuery("yi=one&er=two");
+        ASSERT_EQ(d, ((StringMap) { { "yi", "one" }, { "er", "two" } }));
+    }
+
+    TEST(decodeQuery, decodeUrlEncodedArgs) {
+        auto d = decodeQuery("arg=%3D%3D%40%3D%3D");
+        ASSERT_EQ(d, ((StringMap) { { "arg", "==@==" } }));
+    }
+
+    TEST(decodeQuery, decodeArgWithEmptyValue) {
+        auto d = decodeQuery("arg=");
+        ASSERT_EQ(d, ((StringMap) { { "arg", ""} }));
+    }
+
+    /* ----------------------------------------------------------------------------
+     * percentDecode
+     * --------------------------------------------------------------------------*/
+
+    TEST(percentDecode, decodesUrlEncodedString) {
+        string s = "==@==";
+        string d = percentDecode("%3D%3D%40%3D%3D");
+        ASSERT_EQ(d, s);
+    }
+
+    TEST(percentDecode, multipleDecodesAreIdempotent) {
+        string once = percentDecode("%3D%3D%40%3D%3D");
+        string twice = percentDecode(once);
+
+        ASSERT_EQ(once, twice);
+    }
+
+    TEST(percentDecode, trailingPercent) {
+        string s = "==@==%";
+        string d = percentDecode("%3D%3D%40%3D%3D%25");
+
+        ASSERT_EQ(d, s);
+    }
+
+}

src/libutil/types.hh

@@ -41,7 +41,8 @@ struct FormatOrString
 {
     string s;
     FormatOrString(const string & s) : s(s) { };
-    FormatOrString(const format & f) : s(f.str()) { };
+    template<class F>
+    FormatOrString(const F & f) : s(f.str()) { };
     FormatOrString(const char * s) : s(s) { };
 };
 
@@ -51,12 +52,13 @@ struct FormatOrString
    ... a_n’. However, ‘fmt(s)’ is equivalent to ‘s’ (so no %-expansion
    takes place). */
 
-inline void formatHelper(boost::format & f)
+template<class F>
+inline void formatHelper(F & f)
 {
 }
 
-template<typename T, typename... Args>
-inline void formatHelper(boost::format & f, const T & x, const Args & ... args)
+template<class F, typename T, typename... Args>
+inline void formatHelper(F & f, const T & x, const Args & ... args)
 {
     formatHelper(f % x, args...);
 }
@@ -157,4 +159,12 @@ typedef list<Path> Paths;
 typedef set<Path> PathSet;
 
 
+/* Helper class to run code at startup. */
+template<typename T>
+struct OnStartup
+{
+    OnStartup(T && t) { t(); }
+};
+
+
 }

src/libutil/url.cc

@@ -0,0 +1,137 @@
+#include "url.hh"
+#include "util.hh"
+
+namespace nix {
+
+std::regex refRegex(refRegexS, std::regex::ECMAScript);
+std::regex revRegex(revRegexS, std::regex::ECMAScript);
+std::regex flakeIdRegex(flakeIdRegexS, std::regex::ECMAScript);
+
+ParsedURL parseURL(const std::string & url)
+{
+    static std::regex uriRegex(
+        "((" + schemeRegex + "):"
+        + "(?:(?://(" + authorityRegex + ")(" + absPathRegex + "))|(/?" + pathRegex + ")))"
+        + "(?:\\?(" + queryRegex + "))?"
+        + "(?:#(" + queryRegex + "))?",
+        std::regex::ECMAScript);
+
+    std::smatch match;
+
+    if (std::regex_match(url, match, uriRegex)) {
+        auto & base = match[1];
+        std::string scheme = match[2];
+        auto authority = match[3].matched
+            ? std::optional<std::string>(match[3]) : std::nullopt;
+        std::string path = match[4].matched ? match[4] : match[5];
+        auto & query = match[6];
+        auto & fragment = match[7];
+
+        auto isFile = scheme.find("file") != std::string::npos;
+
+        if (authority && *authority != "" && isFile)
+            throw Error("file:// URL '%s' has unexpected authority '%s'",
+                url, *authority);
+
+        if (isFile && path.empty())
+            path = "/";
+
+        return ParsedURL{
+            .url = url,
+            .base = base,
+            .scheme = scheme,
+            .authority = authority,
+            .path = path,
+            .query = decodeQuery(query),
+            .fragment = percentDecode(std::string(fragment))
+        };
+    }
+
+    else
+        throw BadURL("'%s' is not a valid URL", url);
+}
+
+std::string percentDecode(std::string_view in)
+{
+    std::string decoded;
+    for (size_t i = 0; i < in.size(); ) {
+        if (in[i] == '%') {
+            if (i + 2 >= in.size())
+                throw BadURL("invalid URI parameter '%s'", in);
+            try {
+                decoded += std::stoul(std::string(in, i + 1, 2), 0, 16);
+                i += 3;
+            } catch (...) {
+                throw BadURL("invalid URI parameter '%s'", in);
+            }
+        } else
+            decoded += in[i++];
+    }
+    return decoded;
+}
+
+std::map<std::string, std::string> decodeQuery(const std::string & query)
+{
+    std::map<std::string, std::string> result;
+
+    for (auto s : tokenizeString<Strings>(query, "&")) {
+        auto e = s.find('=');
+        if (e != std::string::npos)
+            result.emplace(
+                s.substr(0, e),
+                percentDecode(std::string_view(s).substr(e + 1)));
+    }
+
+    return result;
+}
+
+std::string percentEncode(std::string_view s)
+{
+    std::string res;
+    for (auto & c : s)
+        if ((c >= 'a' && c <= 'z')
+            || (c >= 'A' && c <= 'Z')
+            || (c >= '0' && c <= '9')
+            || strchr("-._~!$&'()*+,;=:@", c))
+            res += c;
+        else
+            res += fmt("%%%02x", (unsigned int) c);
+    return res;
+}
+
+std::string encodeQuery(const std::map<std::string, std::string> & ss)
+{
+    std::string res;
+    bool first = true;
+    for (auto & [name, value] : ss) {
+        if (!first) res += '&';
+        first = false;
+        res += percentEncode(name);
+        res += '=';
+        res += percentEncode(value);
+    }
+    return res;
+}
+
+std::string ParsedURL::to_string() const
+{
+    return
+        scheme
+        + ":"
+        + (authority ? "//" + *authority : "")
+        + path
+        + (query.empty() ? "" : "?" + encodeQuery(query))
+        + (fragment.empty() ? "" : "#" + percentEncode(fragment));
+}
+
+bool ParsedURL::operator ==(const ParsedURL & other) const
+{
+    return
+        scheme == other.scheme
+        && authority == other.authority
+        && path == other.path
+        && query == other.query
+        && fragment == other.fragment;
+}
+
+}

src/libutil/url.hh

@@ -0,0 +1,62 @@
+#pragma once
+
+#include "types.hh"
+
+#include <regex>
+
+namespace nix {
+
+struct ParsedURL
+{
+    std::string url;
+    std::string base; // URL without query/fragment
+    std::string scheme;
+    std::optional<std::string> authority;
+    std::string path;
+    std::map<std::string, std::string> query;
+    std::string fragment;
+
+    std::string to_string() const;
+
+    bool operator ==(const ParsedURL & other) const;
+};
+
+MakeError(BadURL, Error);
+
+std::string percentDecode(std::string_view in);
+
+std::map<std::string, std::string> decodeQuery(const std::string & query);
+
+ParsedURL parseURL(const std::string & url);
+
+// URI stuff.
+const static std::string pctEncoded = "(?:%[0-9a-fA-F][0-9a-fA-F])";
+const static std::string schemeRegex = "(?:[a-z+]+)";
+const static std::string ipv6AddressRegex = "(?:\\[[0-9a-fA-F:]+\\])";
+const static std::string unreservedRegex = "(?:[a-zA-Z0-9-._~])";
+const static std::string subdelimsRegex = "(?:[!$&'\"()*+,;=])";
+const static std::string hostnameRegex = "(?:(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + ")*)";
+const static std::string hostRegex = "(?:" + ipv6AddressRegex + "|" + hostnameRegex + ")";
+const static std::string userRegex = "(?:(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + "|:)*)";
+const static std::string authorityRegex = "(?:" + userRegex + "@)?" + hostRegex + "(?::[0-9]+)?";
+const static std::string pcharRegex = "(?:" + unreservedRegex + "|" + pctEncoded + "|" + subdelimsRegex + "|[:@])";
+const static std::string queryRegex = "(?:" + pcharRegex + "|[/? \"])*";
+const static std::string segmentRegex = "(?:" + pcharRegex + "+)";
+const static std::string absPathRegex = "(?:(?:/" + segmentRegex + ")*/?)";
+const static std::string pathRegex = "(?:" + segmentRegex + "(?:/" + segmentRegex + ")*/?)";
+
+// A Git ref (i.e. branch or tag name).
+const static std::string refRegexS = "[a-zA-Z0-9][a-zA-Z0-9_.-]*"; // FIXME: check
+extern std::regex refRegex;
+
+// A Git revision (a SHA-1 commit hash).
+const static std::string revRegexS = "[0-9a-fA-F]{40}";
+extern std::regex revRegex;
+
+// A ref or revision, or a ref followed by a revision.
+const static std::string refAndOrRevRegex = "(?:(" + revRegexS + ")|(?:(" + refRegexS + ")(?:/(" + revRegexS + "))?))";
+
+const static std::string flakeIdRegexS = "[a-zA-Z][a-zA-Z0-9_-]*";
+extern std::regex flakeIdRegex;
+
+}

src/libutil/util.cc

@@ -268,16 +268,13 @@ bool isLink(const Path & path)
 }
 
 
-DirEntries readDirectory(const Path & path)
+DirEntries readDirectory(DIR *dir, const Path & path)
 {
     DirEntries entries;
     entries.reserve(64);
 
-    AutoCloseDir dir(opendir(path.c_str()));
-    if (!dir) throw SysError(format("opening directory '%1%'") % path);
-
     struct dirent * dirent;
-    while (errno = 0, dirent = readdir(dir.get())) { /* sic */
+    while (errno = 0, dirent = readdir(dir)) { /* sic */
         checkInterrupt();
         string name = dirent->d_name;
         if (name == "." || name == "..") continue;
@@ -294,6 +291,14 @@ DirEntries readDirectory(const Path & path)
     return entries;
 }
 
+DirEntries readDirectory(const Path & path)
+{
+    AutoCloseDir dir(opendir(path.c_str()));
+    if (!dir) throw SysError(format("opening directory '%1%'") % path);
+
+    return readDirectory(dir.get(), path);
+}
+
 
 unsigned char getFileType(const Path & path)
 {
@@ -311,19 +316,16 @@ string readFile(int fd)
     if (fstat(fd, &st) == -1)
         throw SysError("statting file");
 
-    std::vector<unsigned char> buf(st.st_size);
-    readFull(fd, buf.data(), st.st_size);
-
-    return string((char *) buf.data(), st.st_size);
+    return drainFD(fd, true, st.st_size);
 }
 
 
-string readFile(const Path & path, bool drain)
+string readFile(const Path & path)
 {
     AutoCloseFD fd = open(path.c_str(), O_RDONLY | O_CLOEXEC);
     if (!fd)
         throw SysError(format("opening file '%1%'") % path);
-    return drain ? drainFD(fd.get()) : readFile(fd.get());
+    return readFile(fd.get());
 }
 
 
@@ -389,12 +391,14 @@ void writeLine(int fd, string s)
 }
 
 
-static void _deletePath(const Path & path, unsigned long long & bytesFreed)
+static void _deletePath(int parentfd, const Path & path, unsigned long long & bytesFreed)
 {
     checkInterrupt();
 
+    string name(baseNameOf(path));
+
     struct stat st;
-    if (lstat(path.c_str(), &st) == -1) {
+    if (fstatat(parentfd, name.c_str(), &st, AT_SYMLINK_NOFOLLOW) == -1) {
         if (errno == ENOENT) return;
         throw SysError(format("getting status of '%1%'") % path);
     }
@@ -406,20 +410,45 @@ static void _deletePath(const Path & path, unsigned long long & bytesFreed)
         /* Make the directory accessible. */
         const auto PERM_MASK = S_IRUSR | S_IWUSR | S_IXUSR;
         if ((st.st_mode & PERM_MASK) != PERM_MASK) {
-            if (chmod(path.c_str(), st.st_mode | PERM_MASK) == -1)
+            if (fchmodat(parentfd, name.c_str(), st.st_mode | PERM_MASK, 0) == -1)
                 throw SysError(format("chmod '%1%'") % path);
         }
 
-        for (auto & i : readDirectory(path))
-            _deletePath(path + "/" + i.name, bytesFreed);
+        int fd = openat(parentfd, path.c_str(), O_RDONLY);
+        if (!fd)
+            throw SysError(format("opening directory '%1%'") % path);
+        AutoCloseDir dir(fdopendir(fd));
+        if (!dir)
+            throw SysError(format("opening directory '%1%'") % path);
+        for (auto & i : readDirectory(dir.get(), path))
+            _deletePath(dirfd(dir.get()), path + "/" + i.name, bytesFreed);
     }
 
-    if (remove(path.c_str()) == -1) {
+    int flags = S_ISDIR(st.st_mode) ? AT_REMOVEDIR : 0;
+    if (unlinkat(parentfd, name.c_str(), flags) == -1) {
         if (errno == ENOENT) return;
         throw SysError(format("cannot unlink '%1%'") % path);
     }
 }
 
+static void _deletePath(const Path & path, unsigned long long & bytesFreed)
+{
+    Path dir = dirOf(path);
+    if (dir == "")
+        dir = "/";
+
+    AutoCloseFD dirfd(open(dir.c_str(), O_RDONLY));
+    if (!dirfd) {
+        // This really shouldn't fail silently, but it's left this way
+        // for backwards compatibility.
+        if (errno == ENOENT) return;
+
+        throw SysError(format("opening directory '%1%'") % path);
+    }
+
+    _deletePath(dirfd.get(), path, bytesFreed);
+}
+
 
 void deletePath(const Path & path)
 {
@@ -478,6 +507,17 @@ Path createTempDir(const Path & tmpRoot, const Path & prefix,
 }
 
 
+std::pair<AutoCloseFD, Path> createTempFile(const Path & prefix)
+{
+    Path tmpl(getEnv("TMPDIR").value_or("/tmp") + "/" + prefix + ".XXXXXX");
+    // Strictly speaking, this is UB, but who cares...
+    AutoCloseFD fd(mkstemp((char *) tmpl.c_str()));
+    if (!fd)
+        throw SysError("creating temporary file '%s'", tmpl);
+    return {std::move(fd), tmpl};
+}
+
+
 std::string getUserName()
 {
     auto pw = getpwuid(geteuid());
@@ -622,9 +662,9 @@ void writeFull(int fd, const string & s, bool allowInterrupts)
 }
 
 
-string drainFD(int fd, bool block)
+string drainFD(int fd, bool block, const size_t reserveSize)
 {
-    StringSink sink;
+    StringSink sink(reserveSize);
     drainFD(fd, sink, block);
     return std::move(*sink.s);
 }

src/libutil/util.hh

@@ -2,6 +2,7 @@
 
 #include "types.hh"
 #include "logging.hh"
+#include "ansicolor.hh"
 
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -16,6 +17,7 @@
 #include <sstream>
 #include <optional>
 #include <future>
+#include <iterator>
 
 #ifndef HAVE_STRUCT_DIRENT_D_TYPE
 #define DT_UNKNOWN 0
@@ -56,12 +58,12 @@ Path canonPath(const Path & path, bool resolveSymlinks = false);
 
 /* Return the directory part of the given canonical path, i.e.,
    everything before the final `/'.  If the path is the root or an
-   immediate child thereof (e.g., `/foo'), this means an empty string
-   is returned. */
+   immediate child thereof (e.g., `/foo'), this means `/'
+   is returned.*/
 Path dirOf(const Path & path);
 
 /* Return the base name of the given canonical path, i.e., everything
-   following the final `/'. */
+   following the final `/' (trailing slashes are removed). */
 std::string_view baseNameOf(std::string_view path);
 
 /* Check whether 'path' is a descendant of 'dir'. */
@@ -101,7 +103,7 @@ unsigned char getFileType(const Path & path);
 
 /* Read the contents of a file into a string. */
 string readFile(int fd);
-string readFile(const Path & path, bool drain = false);
+string readFile(const Path & path);
 void readFile(const Path & path, Sink & sink);
 
 /* Write a string to a file. */
@@ -122,10 +124,6 @@ void deletePath(const Path & path);
 
 void deletePath(const Path & path, unsigned long long & bytesFreed);
 
-/* Create a temporary directory. */
-Path createTempDir(const Path & tmpRoot = "", const Path & prefix = "nix",
-    bool includePid = true, bool useGlobalCounter = true, mode_t mode = 0755);
-
 std::string getUserName();
 
 /* Return $HOME or the user's home directory from /etc/passwd. */
@@ -164,7 +162,7 @@ MakeError(EndOfFile, Error);
 
 
 /* Read a file descriptor until EOF occurs. */
-string drainFD(int fd, bool block = true);
+string drainFD(int fd, bool block = true, const size_t reserveSize=0);
 
 void drainFD(int fd, Sink & sink, bool block = true);
 
@@ -205,6 +203,14 @@ public:
 };
 
 
+/* Create a temporary directory. */
+Path createTempDir(const Path & tmpRoot = "", const Path & prefix = "nix",
+    bool includePid = true, bool useGlobalCounter = true, mode_t mode = 0755);
+
+/* Create a temporary file, returning a file handle and its path. */
+std::pair<AutoCloseFD, Path> createTempFile(const Path & prefix = "nix");
+
+
 class Pipe
 {
 public:
@@ -383,17 +389,6 @@ string replaceStrings(const std::string & s,
 std::string rewriteStrings(const std::string & s, const StringMap & rewrites);
 
 
-/* If a set contains 'from', remove it and insert 'to'. */
-template<typename T>
-void replaceInSet(std::set<T> & set, const T & from, const T & to)
-{
-    auto i = set.find(from);
-    if (i == set.end()) return;
-    set.erase(i);
-    set.insert(to);
-}
-
-
 /* Convert the exit status of a child as returned by wait() into an
    error string. */
 string statusToString(int status);
@@ -441,15 +436,6 @@ std::string shellEscape(const std::string & s);
 void ignoreException();
 
 
-/* Some ANSI escape sequences. */
-#define ANSI_NORMAL "\e[0m"
-#define ANSI_BOLD "\e[1m"
-#define ANSI_FAINT "\e[2m"
-#define ANSI_RED "\e[31;1m"
-#define ANSI_GREEN "\e[32;1m"
-#define ANSI_YELLOW "\e[33;1m"
-#define ANSI_BLUE "\e[34;1m"
-
 
 /* Tree formatting. */
 constexpr char treeConn[] = "├───";

src/nix-build/nix-build.cc

@@ -16,7 +16,7 @@
 #include "get-drvs.hh"
 #include "common-eval-args.hh"
 #include "attr-path.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 using namespace nix;
 using namespace std::string_literals;

src/nix-channel/nix-channel.cc

@@ -1,8 +1,9 @@
 #include "shared.hh"
 #include "globals.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "store-api.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
+#include "fetchers.hh"
 
 #include <fcntl.h>
 #include <regex>
@@ -86,12 +87,9 @@ static void update(const StringSet & channelNames)
         // We want to download the url to a file to see if it's a tarball while also checking if we
         // got redirected in the process, so that we can grab the various parts of a nix channel
         // definition from a consistent location if the redirect changes mid-download.
-        CachedDownloadRequest request(url);
-        request.ttl = 0;
-        auto dl = getDownloader();
-        auto result = dl->downloadCached(store, request);
-        auto filename = result.path;
-        url = chomp(result.effectiveUri);
+        auto result = fetchers::downloadFile(store, url, std::string(baseNameOf(url)), false);
+        auto filename = store->toRealPath(result.storePath);
+        url = result.effectiveUrl;
 
         // If the URL contains a version number, append it to the name
         // attribute (so that "nix-env -q" on the channels profile
@@ -114,11 +112,10 @@ static void update(const StringSet & channelNames)
         if (!unpacked) {
             // Download the channel tarball.
             try {
-                filename = dl->downloadCached(store, CachedDownloadRequest(url + "/nixexprs.tar.xz")).path;
-            } catch (DownloadError & e) {
-                filename = dl->downloadCached(store, CachedDownloadRequest(url + "/nixexprs.tar.bz2")).path;
+                filename = store->toRealPath(fetchers::downloadFile(store, url + "/nixexprs.tar.xz", "nixexprs.tar.xz", false).storePath);
+            } catch (FileTransferError & e) {
+                filename = store->toRealPath(fetchers::downloadFile(store, url + "/nixexprs.tar.bz2", "nixexprs.tar.bz2", false).storePath);
             }
-            chomp(filename);
         }
 
         // Regardless of where it came from, add the expression representing this channel to accumulated expression
@@ -185,6 +182,8 @@ static int _main(int argc, char ** argv)
             } else if (*arg == "--rollback") {
                 cmd = cRollback;
             } else {
+                if (hasPrefix(*arg, "-"))
+                    throw UsageError("unsupported argument '%s'", *arg);
                 args.push_back(std::move(*arg));
             }
             return true;

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

@@ -2,7 +2,7 @@
 #include "profiles.hh"
 #include "shared.hh"
 #include "globals.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 #include <iostream>
 #include <cerrno>

src/nix-copy-closure/nix-copy-closure.cc

@@ -1,6 +1,6 @@
 #include "shared.hh"
 #include "store-api.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 using namespace nix;
 

src/nix-daemon/nix-daemon.cc

@@ -6,7 +6,7 @@
 #include "globals.hh"
 #include "derivations.hh"
 #include "finally.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 #include "daemon.hh"
 
 #include <algorithm>

src/nix-env/nix-env.cc

@@ -13,7 +13,7 @@
 #include "json.hh"
 #include "value-to-json.hh"
 #include "xml-writer.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 #include <cerrno>
 #include <ctime>
@@ -718,28 +718,39 @@ static void uninstallDerivations(Globals & globals, Strings & selectors,
     while (true) {
         string lockToken = optimisticLockProfile(profile);
 
-        DrvInfos installedElems = queryInstalled(*globals.state, profile);
-        DrvInfos newElems;
+        DrvInfos workingElems = queryInstalled(*globals.state, profile);
 
-        for (auto & i : installedElems) {
-            DrvName drvName(i.queryName());
-            bool found = false;
-            for (auto & j : selectors)
-                /* !!! the repeated calls to followLinksToStorePath()
-                   are expensive, should pre-compute them. */
-                if ((isPath(j) && globals.state->store->parseStorePath(i.queryOutPath()) == globals.state->store->followLinksToStorePath(j))
-                    || DrvName(j).matches(drvName))
-                {
-                    printInfo("uninstalling '%s'", i.queryName());
-                    found = true;
-                    break;
-                }
-            if (!found) newElems.push_back(i);
+        for (auto & selector : selectors) {
+            DrvInfos::iterator split = workingElems.begin();
+            if (isPath(selector)) {
+                StorePath selectorStorePath = globals.state->store->followLinksToStorePath(selector);
+                split = std::partition(
+                    workingElems.begin(), workingElems.end(),
+                    [&selectorStorePath, globals](auto &elem) {
+                        return selectorStorePath != globals.state->store->parseStorePath(elem.queryOutPath());
+                    }
+                );
+            } else {
+                DrvName selectorName(selector);
+                split = std::partition(
+                    workingElems.begin(), workingElems.end(),
+                    [&selectorName](auto &elem){
+                        DrvName elemName(elem.queryName());
+                        return !selectorName.matches(elemName);
+                    }
+                );
+            }
+            if (split == workingElems.end())
+                warn("selector '%s' matched no installed derivations", selector);
+            for (auto removedElem = split; removedElem != workingElems.end(); removedElem++) {
+                printInfo("uninstalling '%s'", removedElem->queryName());
+            }
+            workingElems.erase(split, workingElems.end());
         }
 
         if (globals.dryRun) return;
 
-        if (createUserEnv(*globals.state, newElems,
+        if (createUserEnv(*globals.state, workingElems,
                 profile, settings.envKeepDerivations, lockToken)) break;
     }
 }

src/nix-env/user-env.cc

@@ -15,6 +15,8 @@ namespace nix {
 DrvInfos queryInstalled(EvalState & state, const Path & userEnv)
 {
     DrvInfos elems;
+    if (pathExists(userEnv + "/manifest.json"))
+        throw Error("profile '%s' is incompatible with 'nix-env'; please use 'nix profile' instead", userEnv);
     Path manifestFile = userEnv + "/manifest.nix";
     if (pathExists(manifestFile)) {
         Value v;

src/nix-instantiate/nix-instantiate.cc

@@ -9,7 +9,7 @@
 #include "util.hh"
 #include "store-api.hh"
 #include "common-eval-args.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 #include <map>
 #include <iostream>

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

@@ -1,14 +1,14 @@
 #include "hash.hh"
 #include "shared.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "store-api.hh"
 #include "eval.hh"
 #include "eval-inline.hh"
 #include "common-eval-args.hh"
 #include "attr-path.hh"
-#include "legacy.hh"
 #include "finally.hh"
-#include "progress-bar.hh"
+#include "../nix/legacy.hh"
+#include "../nix/progress-bar.hh"
 #include "tarfile.hh"
 
 #include <iostream>
@@ -181,9 +181,9 @@ static int _main(int argc, char * * argv)
 
                 FdSink sink(fd.get());
 
-                DownloadRequest req(actualUri);
+                FileTransferRequest req(actualUri);
                 req.decompress = false;
-                getDownloader()->download(std::move(req), sink);
+                getFileTransfer()->download(std::move(req), sink);
             }
 
             /* Optionally unpack the file. */

src/nix-store/nix-store.cc

@@ -9,7 +9,7 @@
 #include "util.hh"
 #include "worker-protocol.hh"
 #include "graphml.hh"
-#include "legacy.hh"
+#include "../nix/legacy.hh"
 
 #include <iostream>
 #include <algorithm>
@@ -174,7 +174,7 @@ static void opAdd(Strings opFlags, Strings opArgs)
    store. */
 static void opAddFixed(Strings opFlags, Strings opArgs)
 {
-    FileIngestionMethod recursive = FileIngestionMethod::Flat;
+    auto recursive = FileIngestionMethod::Flat;
 
     for (auto & i : opFlags)
         if (i == "--recursive") recursive = FileIngestionMethod::Recursive;
@@ -194,7 +194,7 @@ static void opAddFixed(Strings opFlags, Strings opArgs)
 /* Hack to support caching in `nix-prefetch-url'. */
 static void opPrintFixedPath(Strings opFlags, Strings opArgs)
 {
-    FileIngestionMethod recursive = FileIngestionMethod::Flat;
+    auto recursive = FileIngestionMethod::Flat;
 
     for (auto i : opFlags)
         if (i == "--recursive") recursive = FileIngestionMethod::Recursive;

src/nix/add-to-store.cc

@@ -14,12 +14,13 @@ struct CmdAddToStore : MixDryRun, StoreCommand
     {
         expectArg("path", &path);
 
-        mkFlag()
-            .longName("name")
-            .shortName('n')
-            .description("name component of the store path")
-            .labels({"name"})
-            .dest(&namePart);
+        addFlag({
+            .longName = "name",
+            .shortName = 'n',
+            .description = "name component of the store path",
+            .labels = {"name"},
+            .handler = {&namePart},
+        });
     }
 
     std::string description() override
@@ -33,6 +34,8 @@ struct CmdAddToStore : MixDryRun, StoreCommand
         };
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store) override
     {
         if (!namePart) namePart = baseNameOf(path);
@@ -50,7 +53,7 @@ struct CmdAddToStore : MixDryRun, StoreCommand
         if (!dryRun)
             store->addToStore(info, sink.s);
 
-        std::cout << fmt("%s\n", store->printStorePath(info.path));
+        logger->stdout("%s", store->printStorePath(info.path));
     }
 };
 

src/nix/build.cc

@@ -5,23 +5,25 @@
 
 using namespace nix;
 
-struct CmdBuild : MixDryRun, InstallablesCommand
+struct CmdBuild : InstallablesCommand, MixDryRun, MixProfile
 {
     Path outLink = "result";
 
     CmdBuild()
     {
-        mkFlag()
-            .longName("out-link")
-            .shortName('o')
-            .description("path of the symlink to the build result")
-            .labels({"path"})
-            .dest(&outLink);
+        addFlag({
+            .longName = "out-link",
+            .shortName = 'o',
+            .description = "path of the symlink to the build result",
+            .labels = {"path"},
+            .handler = {&outLink},
+        });
 
-        mkFlag()
-            .longName("no-link")
-            .description("do not create a symlink to the build result")
-            .set(&outLink, Path(""));
+        addFlag({
+            .longName = "no-link",
+            .description = "do not create a symlink to the build result",
+            .handler = {&outLink, Path("")},
+        });
     }
 
     std::string description() override
@@ -40,6 +42,10 @@ struct CmdBuild : MixDryRun, InstallablesCommand
                 "To build the build.x86_64-linux attribute from release.nix:",
                 "nix build -f release.nix build.x86_64-linux"
             },
+            Example{
+                "To make a profile point at GNU Hello:",
+                "nix build --profile /tmp/profile nixpkgs.hello"
+            },
         };
     }
 
@@ -49,18 +55,19 @@ struct CmdBuild : MixDryRun, InstallablesCommand
 
         if (dryRun) return;
 
-        for (size_t i = 0; i < buildables.size(); ++i) {
-            auto & b(buildables[i]);
-
-            if (outLink != "")
-                for (auto & output : b.outputs)
+        if (outLink != "") {
+            for (size_t i = 0; i < buildables.size(); ++i) {
+                for (auto & output : buildables[i].outputs)
                     if (auto store2 = store.dynamic_pointer_cast<LocalFSStore>()) {
                         std::string symlink = outLink;
                         if (i) symlink += fmt("-%d", i);
                         if (output.first != "out") symlink += fmt("-%s", output.first);
                         store2->addPermRoot(output.second, absPath(symlink), true);
                     }
+            }
         }
+
+        updateProfile(buildables);
     }
 };
 

src/nix/cat.cc

@@ -30,9 +30,11 @@ struct CmdCatStore : StoreCommand, MixCat
 
     std::string description() override
     {
-        return "print the contents of a store file on stdout";
+        return "print the contents of a file in the Nix store on stdout";
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store) override
     {
         cat(store->getFSAccessor());
@@ -51,9 +53,11 @@ struct CmdCatNar : StoreCommand, MixCat
 
     std::string description() override
     {
-        return "print the contents of a file inside a NAR file";
+        return "print the contents of a file inside a NAR file on stdout";
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store) override
     {
         cat(makeNarAccessor(make_ref<std::string>(readFile(narPath))));

src/nix/command.cc

@@ -2,6 +2,9 @@
 #include "store-api.hh"
 #include "derivations.hh"
 #include "nixexpr.hh"
+#include "profiles.hh"
+
+extern char * * environ;
 
 namespace nix {
 
@@ -32,16 +35,18 @@ StorePathsCommand::StorePathsCommand(FileIngestionMethod recursive)
     : recursive(recursive)
 {
     if (recursive == FileIngestionMethod::Recursive)
-        mkFlag()
-            .longName("no-recursive")
-            .description("apply operation to specified paths only")
-            .set(&this->recursive, FileIngestionMethod::Flat);
+        addFlag({
+            .longName = "no-recursive",
+            .description = "apply operation to specified paths only",
+            .handler = {&this->recursive, FileIngestionMethod::Flat},
+        });
     else
-        mkFlag()
-            .longName("recursive")
-            .shortName('r')
-            .description("apply operation to closure of the specified paths")
-            .set(&this->recursive, FileIngestionMethod::Recursive);
+        addFlag({
+            .longName = "recursive",
+            .shortName = 'r',
+            .description = "apply operation to closure of the specified paths",
+            .handler = {&this->recursive, FileIngestionMethod::Recursive},
+        });
 
     mkFlag(0, "all", "apply operation to the entire store", &all);
 }
@@ -96,4 +101,98 @@ Strings editorFor(const Pos & pos)
     return args;
 }
 
+MixProfile::MixProfile()
+{
+    addFlag({
+        .longName = "profile",
+        .description = "profile to update",
+        .labels = {"path"},
+        .handler = {&profile},
+    });
+}
+
+void MixProfile::updateProfile(const StorePath & storePath)
+{
+    if (!profile) return;
+    auto store = getStore().dynamic_pointer_cast<LocalFSStore>();
+    if (!store) throw Error("'--profile' is not supported for this Nix store");
+    auto profile2 = absPath(*profile);
+    switchLink(profile2,
+        createGeneration(
+            ref<LocalFSStore>(store),
+            profile2, store->printStorePath(storePath)));
+}
+
+void MixProfile::updateProfile(const Buildables & buildables)
+{
+    if (!profile) return;
+
+    std::optional<StorePath> result;
+
+    for (auto & buildable : buildables) {
+        for (auto & output : buildable.outputs) {
+            if (result)
+                throw Error("'--profile' requires that the arguments produce a single store path, but there are multiple");
+            result = output.second.clone();
+        }
+    }
+
+    if (!result)
+        throw Error("'--profile' requires that the arguments produce a single store path, but there are none");
+
+    updateProfile(*result);
+}
+
+MixDefaultProfile::MixDefaultProfile()
+{
+    profile = getDefaultProfile();
+}
+
+MixEnvironment::MixEnvironment() : ignoreEnvironment(false)
+{
+    addFlag({
+        .longName = "ignore-environment",
+        .shortName = 'i',
+        .description = "clear the entire environment (except those specified with --keep)",
+        .handler = {&ignoreEnvironment, true},
+    });
+
+    addFlag({
+        .longName = "keep",
+        .shortName = 'k',
+        .description = "keep specified environment variable",
+        .labels = {"name"},
+        .handler = {[&](std::string s) { keep.insert(s); }},
+    });
+
+    addFlag({
+        .longName = "unset",
+        .shortName = 'u',
+        .description = "unset specified environment variable",
+        .labels = {"name"},
+        .handler = {[&](std::string s) { unset.insert(s); }},
+    });
+}
+
+void MixEnvironment::setEnviron() {
+    if (ignoreEnvironment) {
+        if (!unset.empty())
+            throw UsageError("--unset does not make sense with --ignore-environment");
+
+        for (const auto & var : keep) {
+            auto val = getenv(var.c_str());
+            if (val) stringsEnv.emplace_back(fmt("%s=%s", var.c_str(), val));
+        }
+
+        vectorEnv = stringsToCharPtrs(stringsEnv);
+        environ = vectorEnv.data();
+    } else {
+        if (!keep.empty())
+            throw UsageError("--keep does not make sense without --ignore-environment");
+
+        for (const auto & var : unset)
+            unsetenv(var.c_str());
+    }
+}
+
 }

src/nix/command.hh

@@ -1,5 +1,6 @@
 #pragma once
 
+#include "installables.hh"
 #include "args.hh"
 #include "common-eval-args.hh"
 #include "path.hh"
@@ -9,6 +10,10 @@ namespace nix {
 
 extern std::string programPath;
 
+static constexpr Command::Category catSecondary = 100;
+static constexpr Command::Category catUtility = 101;
+static constexpr Command::Category catNixInstallation = 102;
+
 /* A command that requires a Nix store. */
 struct StoreCommand : virtual Command
 {
@@ -22,34 +27,7 @@ private:
     std::shared_ptr<Store> _store;
 };
 
-struct Buildable
-{
-    std::optional<StorePath> drvPath;
-    std::map<std::string, StorePath> outputs;
-};
-
-typedef std::vector<Buildable> Buildables;
-
-struct Installable
-{
-    virtual ~Installable() { }
-
-    virtual std::string what() = 0;
-
-    virtual Buildables toBuildables()
-    {
-        throw Error("argument '%s' cannot be built", what());
-    }
-
-    Buildable toBuildable();
-
-    virtual std::pair<Value *, Pos> toValue(EvalState & state)
-    {
-        throw Error("argument '%s' cannot be evaluated", what());
-    }
-};
-
-struct SourceExprCommand : virtual Args, StoreCommand, MixEvalArgs
+struct SourceExprCommand : virtual StoreCommand, MixEvalArgs
 {
     Path file;
 
@@ -67,7 +45,7 @@ private:
 
     std::shared_ptr<EvalState> evalState;
 
-    Value * vSourceExpr = 0;
+    RootValue vSourceExpr;
 };
 
 enum RealiseMode { Build, NoBuild, DryRun };
@@ -184,4 +162,36 @@ std::set<StorePath> toDerivations(ref<Store> store,
    filename:lineno. */
 Strings editorFor(const Pos & pos);
 
+struct MixProfile : virtual StoreCommand
+{
+    std::optional<Path> profile;
+
+    MixProfile();
+
+    /* If 'profile' is set, make it point at 'storePath'. */
+    void updateProfile(const StorePath & storePath);
+
+    /* If 'profile' is set, make it point at the store path produced
+       by 'buildables'. */
+    void updateProfile(const Buildables & buildables);
+};
+
+struct MixDefaultProfile : MixProfile
+{
+    MixDefaultProfile();
+};
+
+struct MixEnvironment : virtual Args {
+
+    StringSet keep, unset;
+    Strings stringsEnv;
+    std::vector<char*> vectorEnv;
+    bool ignoreEnvironment;
+
+    MixEnvironment();
+
+    /* Modify global environ based on ignoreEnvironment, keep, and unset. It's expected that exec will be called before this class goes out of scope, otherwise environ will become invalid. */
+    void setEnviron();
+};
+
 }

src/nix/copy.cc

@@ -19,27 +19,32 @@ struct CmdCopy : StorePathsCommand
     CmdCopy()
         : StorePathsCommand(FileIngestionMethod::Recursive)
     {
-        mkFlag()
-            .longName("from")
-            .labels({"store-uri"})
-            .description("URI of the source Nix store")
-            .dest(&srcUri);
-        mkFlag()
-            .longName("to")
-            .labels({"store-uri"})
-            .description("URI of the destination Nix store")
-            .dest(&dstUri);
+        addFlag({
+            .longName = "from",
+            .description = "URI of the source Nix store",
+            .labels = {"store-uri"},
+            .handler = {&srcUri},
+        });
 
-        mkFlag()
-            .longName("no-check-sigs")
-            .description("do not require that paths are signed by trusted keys")
-            .set(&checkSigs, NoCheckSigs);
+        addFlag({
+            .longName = "to",
+            .description = "URI of the destination Nix store",
+            .labels = {"store-uri"},
+            .handler = {&dstUri},
+        });
 
-        mkFlag()
-            .longName("substitute-on-destination")
-            .shortName('s')
-            .description("whether to try substitutes on the destination store (only supported by SSH)")
-            .set(&substitute, Substitute);
+        addFlag({
+            .longName = "no-check-sigs",
+            .description = "do not require that paths are signed by trusted keys",
+            .handler = {&checkSigs, NoCheckSigs},
+        });
+
+        addFlag({
+            .longName = "substitute-on-destination",
+            .shortName = 's',
+            .description = "whether to try substitutes on the destination store (only supported by SSH)",
+            .handler = {&substitute, Substitute},
+        });
     }
 
     std::string description() override
@@ -75,6 +80,8 @@ struct CmdCopy : StorePathsCommand
         };
     }
 
+    Category category() override { return catSecondary; }
+
     ref<Store> createStore() override
     {
         return srcUri.empty() ? StoreCommand::createStore() : openStore(srcUri);

src/nix/dev-shell.cc

@@ -0,0 +1,347 @@
+#include "eval.hh"
+#include "command.hh"
+#include "common-args.hh"
+#include "shared.hh"
+#include "store-api.hh"
+#include "derivations.hh"
+#include "affinity.hh"
+#include "progress-bar.hh"
+
+#include <regex>
+
+using namespace nix;
+
+struct Var
+{
+    bool exported = true;
+    bool associative = false;
+    std::string value; // quoted string or array
+};
+
+struct BuildEnvironment
+{
+    std::map<std::string, Var> env;
+    std::string bashFunctions;
+};
+
+BuildEnvironment readEnvironment(const Path & path)
+{
+    BuildEnvironment res;
+
+    std::set<std::string> exported;
+
+    debug("reading environment file '%s'", path);
+
+    auto file = readFile(path);
+
+    auto pos = file.cbegin();
+
+    static std::string varNameRegex =
+        R"re((?:[a-zA-Z_][a-zA-Z0-9_]*))re";
+
+    static std::regex declareRegex(
+        "^declare -x (" + varNameRegex + ")" +
+        R"re((?:="((?:[^"\\]|\\.)*)")?\n)re");
+
+    static std::string simpleStringRegex =
+        R"re((?:[a-zA-Z0-9_/:\.\-\+=]*))re";
+
+    static std::string quotedStringRegex =
+        R"re((?:\$?'(?:[^'\\]|\\[abeEfnrtv\\'"?])*'))re";
+
+    static std::string indexedArrayRegex =
+        R"re((?:\(( *\[[0-9]+]="(?:[^"\\]|\\.)*")**\)))re";
+
+    static std::regex varRegex(
+        "^(" + varNameRegex + ")=(" + simpleStringRegex + "|" + quotedStringRegex + "|" + indexedArrayRegex + ")\n");
+
+    /* Note: we distinguish between an indexed and associative array
+       using the space before the closing parenthesis. Will
+       undoubtedly regret this some day. */
+    static std::regex assocArrayRegex(
+        "^(" + varNameRegex + ")=" + R"re((?:\(( *\[[^\]]+\]="(?:[^"\\]|\\.)*")* *\)))re" + "\n");
+
+    static std::regex functionRegex(
+        "^" + varNameRegex + " \\(\\) *\n");
+
+    while (pos != file.end()) {
+
+        std::smatch match;
+
+        if (std::regex_search(pos, file.cend(), match, declareRegex)) {
+            pos = match[0].second;
+            exported.insert(match[1]);
+        }
+
+        else if (std::regex_search(pos, file.cend(), match, varRegex)) {
+            pos = match[0].second;
+            res.env.insert({match[1], Var { .exported = exported.count(match[1]) > 0, .value = match[2] }});
+        }
+
+        else if (std::regex_search(pos, file.cend(), match, assocArrayRegex)) {
+            pos = match[0].second;
+            res.env.insert({match[1], Var { .associative = true, .value = match[2] }});
+        }
+
+        else if (std::regex_search(pos, file.cend(), match, functionRegex)) {
+            res.bashFunctions = std::string(pos, file.cend());
+            break;
+        }
+
+        else throw Error("shell environment '%s' has unexpected line '%s'",
+            path, file.substr(pos - file.cbegin(), 60));
+    }
+
+    return res;
+}
+
+const static std::string getEnvSh =
+    #include "get-env.sh.gen.hh"
+    ;
+
+/* Given an existing derivation, return the shell environment as
+   initialised by stdenv's setup script. We do this by building a
+   modified derivation with the same dependencies and nearly the same
+   initial environment variables, that just writes the resulting
+   environment to a file and exits. */
+StorePath getDerivationEnvironment(ref<Store> store, const StorePath & drvPath)
+{
+    auto drv = store->derivationFromPath(drvPath);
+
+    auto builder = baseNameOf(drv.builder);
+    if (builder != "bash")
+        throw Error("'nix dev-shell' only works on derivations that use 'bash' as their builder");
+
+    auto getEnvShPath = store->addTextToStore("get-env.sh", getEnvSh, {});
+
+    drv.args = {store->printStorePath(getEnvShPath)};
+
+    /* Remove derivation checks. */
+    drv.env.erase("allowedReferences");
+    drv.env.erase("allowedRequisites");
+    drv.env.erase("disallowedReferences");
+    drv.env.erase("disallowedRequisites");
+
+    /* Rehash and write the derivation. FIXME: would be nice to use
+       'buildDerivation', but that's privileged. */
+    auto drvName = std::string(drvPath.name());
+    assert(hasSuffix(drvName, ".drv"));
+    drvName.resize(drvName.size() - 4);
+    drvName += "-env";
+    for (auto & output : drv.outputs)
+        drv.env.erase(output.first);
+    drv.env["out"] = "";
+    drv.env["outputs"] = "out";
+    drv.inputSrcs.insert(std::move(getEnvShPath));
+    Hash h = hashDerivationModulo(*store, drv, true);
+    auto shellOutPath = store->makeOutputPath("out", h, drvName);
+    drv.outputs.insert_or_assign("out", DerivationOutput(shellOutPath.clone(), FileSystemHash {
+        .method = FileIngestionMethod::Flat,
+        .hash = Hash { }
+    }));
+    drv.env["out"] = store->printStorePath(shellOutPath);
+    auto shellDrvPath2 = writeDerivation(store, drv, drvName);
+
+    /* Build the derivation. */
+    store->buildPaths({shellDrvPath2});
+
+    assert(store->isValidPath(shellOutPath));
+
+    return shellOutPath;
+}
+
+struct Common : InstallableCommand, MixProfile
+{
+    std::set<string> ignoreVars{
+        "BASHOPTS",
+        "EUID",
+        "HOME", // FIXME: don't ignore in pure mode?
+        "NIX_BUILD_TOP",
+        "NIX_ENFORCE_PURITY",
+        "NIX_LOG_FD",
+        "PPID",
+        "PWD",
+        "SHELLOPTS",
+        "SHLVL",
+        "SSL_CERT_FILE", // FIXME: only want to ignore /no-cert-file.crt
+        "TEMP",
+        "TEMPDIR",
+        "TERM",
+        "TMP",
+        "TMPDIR",
+        "TZ",
+        "UID",
+    };
+
+    void makeRcScript(const BuildEnvironment & buildEnvironment, std::ostream & out)
+    {
+        out << "unset shellHook\n";
+
+        out << "nix_saved_PATH=\"$PATH\"\n";
+
+        for (auto & i : buildEnvironment.env) {
+            if (!ignoreVars.count(i.first) && !hasPrefix(i.first, "BASH_")) {
+                if (i.second.associative)
+                    out << fmt("declare -A %s=(%s)\n", i.first, i.second.value);
+                else {
+                    out << fmt("%s=%s\n", i.first, i.second.value);
+                    if (i.second.exported)
+                        out << fmt("export %s\n", i.first);
+                }
+            }
+        }
+
+        out << "PATH=\"$PATH:$nix_saved_PATH\"\n";
+
+        out << buildEnvironment.bashFunctions << "\n";
+
+        // FIXME: set outputs
+
+        out << "export NIX_BUILD_TOP=\"$(mktemp -d --tmpdir nix-shell.XXXXXX)\"\n";
+        for (auto & i : {"TMP", "TMPDIR", "TEMP", "TEMPDIR"})
+            out << fmt("export %s=\"$NIX_BUILD_TOP\"\n", i);
+
+        out << "eval \"$shellHook\"\n";
+    }
+
+    StorePath getShellOutPath(ref<Store> store)
+    {
+        auto path = installable->getStorePath();
+        if (path && hasSuffix(path->to_string(), "-env"))
+            return path->clone();
+        else {
+            auto drvs = toDerivations(store, {installable});
+
+            if (drvs.size() != 1)
+                throw Error("'%s' needs to evaluate to a single derivation, but it evaluated to %d derivations",
+                    installable->what(), drvs.size());
+
+            auto & drvPath = *drvs.begin();
+
+            return getDerivationEnvironment(store, drvPath);
+        }
+    }
+
+    std::pair<BuildEnvironment, std::string> getBuildEnvironment(ref<Store> store)
+    {
+        auto shellOutPath = getShellOutPath(store);
+
+        auto strPath = store->printStorePath(shellOutPath);
+
+        updateProfile(shellOutPath);
+
+        return {readEnvironment(strPath), strPath};
+    }
+};
+
+struct CmdDevShell : Common, MixEnvironment
+{
+    std::vector<std::string> command;
+
+    CmdDevShell()
+    {
+        addFlag({
+            .longName = "command",
+            .shortName = 'c',
+            .description = "command and arguments to be executed insted of an interactive shell",
+            .labels = {"command", "args"},
+            .handler = {[&](std::vector<std::string> ss) {
+                if (ss.empty()) throw UsageError("--command requires at least one argument");
+                command = ss;
+            }}
+        });
+    }
+
+    std::string description() override
+    {
+        return "run a bash shell that provides the build environment of a derivation";
+    }
+
+    Examples examples() override
+    {
+        return {
+            Example{
+                "To get the build environment of GNU hello:",
+                "nix dev-shell nixpkgs.hello"
+            },
+            Example{
+                "To store the build environment in a profile:",
+                "nix dev-shell --profile /tmp/my-shell nixpkgs.hello"
+            },
+            Example{
+                "To use a build environment previously recorded in a profile:",
+                "nix dev-shell /tmp/my-shell"
+            },
+        };
+    }
+
+    void run(ref<Store> store) override
+    {
+        auto [buildEnvironment, gcroot] = getBuildEnvironment(store);
+
+        auto [rcFileFd, rcFilePath] = createTempFile("nix-shell");
+
+        std::ostringstream ss;
+        makeRcScript(buildEnvironment, ss);
+
+        ss << fmt("rm -f '%s'\n", rcFilePath);
+
+        if (!command.empty()) {
+            std::vector<std::string> args;
+            for (auto s : command)
+                args.push_back(shellEscape(s));
+            ss << fmt("exec %s\n", concatStringsSep(" ", args));
+        }
+
+        writeFull(rcFileFd.get(), ss.str());
+
+        stopProgressBar();
+
+        auto shell = getEnv("SHELL").value_or("bash");
+
+        setEnviron();
+        // prevent garbage collection until shell exits
+        setenv("NIX_GCROOT", gcroot.data(), 1);
+
+        auto args = Strings{std::string(baseNameOf(shell)), "--rcfile", rcFilePath};
+
+        restoreAffinity();
+        restoreSignals();
+
+        execvp(shell.c_str(), stringsToCharPtrs(args).data());
+
+        throw SysError("executing shell '%s'", shell);
+    }
+};
+
+struct CmdPrintDevEnv : Common
+{
+    std::string description() override
+    {
+        return "print shell code that can be sourced by bash to reproduce the build environment of a derivation";
+    }
+
+    Examples examples() override
+    {
+        return {
+            Example{
+                "To apply the build environment of GNU hello to the current shell:",
+                ". <(nix print-dev-env nixpkgs.hello)"
+            },
+        };
+    }
+
+    Category category() override { return catUtility; }
+
+    void run(ref<Store> store) override
+    {
+        auto buildEnvironment = getBuildEnvironment(store).first;
+
+        stopProgressBar();
+
+        makeRcScript(buildEnvironment, std::cout);
+    }
+};
+
+static auto r1 = registerCommand<CmdPrintDevEnv>("print-dev-env");
+static auto r2 = registerCommand<CmdDevShell>("dev-shell");

src/nix/doctor.cc

@@ -40,9 +40,11 @@ struct CmdDoctor : StoreCommand
 
     std::string description() override
     {
-        return "check your system for potential problems and print a PASS or FAIL for each check.";
+        return "check your system for potential problems and print a PASS or FAIL for each check";
     }
 
+    Category category() override { return catNixInstallation; }
+
     void run(ref<Store> store) override
     {
         logger->log("Running checks against store uri: " + store->getUri());

src/nix/dump-path.cc

@@ -20,6 +20,8 @@ struct CmdDumpPath : StorePathCommand
         };
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store, const StorePath & storePath) override
     {
         FdSink sink(STDOUT_FILENO);

src/nix/edit.cc

@@ -25,6 +25,8 @@ struct CmdEdit : InstallableCommand
         };
     }
 
+    Category category() override { return catSecondary; }
+
     void run(ref<Store> store) override
     {
         auto state = getEvalState();

src/nix/eval.cc

@@ -45,6 +45,8 @@ struct CmdEval : MixJSON, InstallableCommand
         };
     }
 
+    Category category() override { return catSecondary; }
+
     void run(ref<Store> store) override
     {
         if (raw && json)
@@ -55,16 +57,15 @@ struct CmdEval : MixJSON, InstallableCommand
         auto v = installable->toValue(*state).first;
         PathSet context;
 
-        stopProgressBar();
-
         if (raw) {
+            stopProgressBar();
             std::cout << state->coerceToString(noPos, *v, context);
         } else if (json) {
             JSONPlaceholder jsonOut(std::cout);
             printValueAsJSON(*state, true, *v, jsonOut, context);
         } else {
             state->forceValueDeep(*v);
-            std::cout << *v << "\n";
+            logger->stdout("%s", *v);
         }
     }
 };

src/nix/get-env.sh

@@ -0,0 +1,9 @@
+set -e
+if [ -e .attrs.sh ]; then source .attrs.sh; fi
+export IN_NIX_SHELL=impure
+export dontAddDisableDepTrack=1
+if [[ -n $stdenv ]]; then
+    source $stdenv/setup
+fi
+export > $out
+set >> $out

src/nix/hash.cc

@@ -23,9 +23,7 @@ struct CmdHash : Command
         mkFlag(0, "base64", "print hash in base-64", &base, Base64);
         mkFlag(0, "base32", "print hash in base-32 (Nix-specific)", &base, Base32);
         mkFlag(0, "base16", "print hash in base-16", &base, Base16);
-        mkFlag()
-            .longName("type")
-            .mkHashTypeFlag(&ht);
+        addFlag(Flag::mkHashTypeFlag("type", &ht));
         #if 0
         mkFlag()
             .longName("modulo")
@@ -43,6 +41,8 @@ struct CmdHash : Command
             : "print cryptographic hash of the NAR serialisation of a path";
     }
 
+    Category category() override { return catUtility; }
+
     void run() override
     {
         for (auto path : paths) {
@@ -60,8 +60,7 @@ struct CmdHash : Command
 
             Hash h = hashSink->finish().first;
             if (truncate && h.hashSize > 20) h = compressHash(h, 20);
-            std::cout << format("%1%\n") %
-                h.to_string(base, base == SRI);
+            logger->stdout(h.to_string(base, base == SRI));
         }
     }
 };
@@ -77,9 +76,7 @@ struct CmdToBase : Command
 
     CmdToBase(Base base) : base(base)
     {
-        mkFlag()
-            .longName("type")
-            .mkHashTypeFlag(&ht);
+        addFlag(Flag::mkHashTypeFlag("type", &ht));
         expectArgs("strings", &args);
     }
 
@@ -92,10 +89,12 @@ struct CmdToBase : Command
             "SRI");
     }
 
+    Category category() override { return catUtility; }
+
     void run() override
     {
         for (auto s : args)
-            std::cout << fmt("%s\n", Hash(s, ht).to_string(base, base == SRI));
+            logger->stdout(Hash(s, ht).to_string(base, base == SRI));
     }
 };
 

src/nix/installables.cc

@@ -12,26 +12,28 @@
 
 namespace nix {
 
+
 SourceExprCommand::SourceExprCommand()
 {
-    mkFlag()
-        .shortName('f')
-        .longName("file")
-        .label("file")
-        .description("evaluate FILE rather than the default")
-        .dest(&file);
+    addFlag({
+        .longName = "file",
+        .shortName = 'f',
+        .description = "evaluate FILE rather than the default",
+        .labels = {"file"},
+        .handler = {&file}
+    });
 }
 
 Value * SourceExprCommand::getSourceExpr(EvalState & state)
 {
-    if (vSourceExpr) return vSourceExpr;
+    if (vSourceExpr) return *vSourceExpr;
 
     auto sToplevel = state.symbols.create("_toplevel");
 
-    vSourceExpr = state.allocValue();
+    vSourceExpr = allocRootValue(state.allocValue());
 
     if (file != "")
-        state.evalFile(lookupFileArg(state, file), *vSourceExpr);
+        state.evalFile(lookupFileArg(state, file), **vSourceExpr);
 
     else {
 
@@ -39,9 +41,9 @@ Value * SourceExprCommand::getSourceExpr(EvalState & state)
 
         auto searchPath = state.getSearchPath();
 
-        state.mkAttrs(*vSourceExpr, 1024);
+        state.mkAttrs(**vSourceExpr, 1024);
 
-        mkBool(*state.allocAttr(*vSourceExpr, sToplevel), true);
+        mkBool(*state.allocAttr(**vSourceExpr, sToplevel), true);
 
         std::unordered_set<std::string> seen;
 
@@ -52,7 +54,7 @@ Value * SourceExprCommand::getSourceExpr(EvalState & state)
             mkPrimOpApp(*v1, state.getBuiltin("findFile"), state.getBuiltin("nixPath"));
             Value * v2 = state.allocValue();
             mkApp(*v2, *v1, mkString(*state.allocValue(), name));
-            mkApp(*state.allocAttr(*vSourceExpr, state.symbols.create(name)),
+            mkApp(*state.allocAttr(**vSourceExpr, state.symbols.create(name)),
                 state.getBuiltin("import"), *v2);
         };
 
@@ -66,10 +68,10 @@ Value * SourceExprCommand::getSourceExpr(EvalState & state)
             } else
                 addEntry(i.first);
 
-        vSourceExpr->attrs->sort();
+        (*vSourceExpr)->attrs->sort();
     }
 
-    return vSourceExpr;
+    return *vSourceExpr;
 }
 
 ref<EvalState> SourceExprCommand::getEvalState()
@@ -109,6 +111,11 @@ struct InstallableStorePath : Installable
         bs.push_back(std::move(b));
         return bs;
     }
+
+    std::optional<StorePath> getStorePath() override
+    {
+        return storePath.clone();
+    }
 };
 
 struct InstallableValue : Installable

src/nix/installables.hh

@@ -0,0 +1,45 @@
+#pragma once
+
+#include "util.hh"
+#include "path.hh"
+#include "eval.hh"
+
+#include <optional>
+
+namespace nix {
+
+struct Buildable
+{
+    std::optional<StorePath> drvPath;
+    std::map<std::string, StorePath> outputs;
+};
+
+typedef std::vector<Buildable> Buildables;
+
+struct Installable
+{
+    virtual ~Installable() { }
+
+    virtual std::string what() = 0;
+
+    virtual Buildables toBuildables()
+    {
+        throw Error("argument '%s' cannot be built", what());
+    }
+
+    Buildable toBuildable();
+
+    virtual std::pair<Value *, Pos> toValue(EvalState & state)
+    {
+        throw Error("argument '%s' cannot be evaluated", what());
+    }
+
+    /* Return a value only if this installable is a store path or a
+       symlink to it. */
+    virtual std::optional<StorePath> getStorePath()
+    {
+        return {};
+    }
+};
+
+}

src/nix/local.mk

@@ -15,7 +15,9 @@ nix_SOURCES := \
   $(wildcard src/nix-prefetch-url/*.cc) \
   $(wildcard src/nix-store/*.cc) \
 
-nix_LIBS = libexpr libmain libstore libutil libnixrust
+nix_CXXFLAGS += -I src/libutil -I src/libstore -I src/libfetchers -I src/libexpr -I src/libmain
+
+nix_LIBS = libexpr libmain libfetchers libstore libutil libnixrust
 
 nix_LDFLAGS = -pthread $(SODIUM_LIBS) $(EDITLINE_LIBS) $(BOOST_LDFLAGS) -lboost_context -lboost_thread -lboost_system
 
@@ -25,3 +27,5 @@ $(foreach name, \
 $(eval $(call install-symlink, $(bindir)/nix, $(libexecdir)/nix/build-remote))
 
 src/nix-env/user-env.cc: src/nix-env/buildenv.nix.gen.hh
+
+src/nix/dev-shell.cc: src/nix/get-env.sh.gen.hh

src/nix/log.cc

@@ -31,6 +31,8 @@ struct CmdLog : InstallableCommand
         };
     }
 
+    Category category() override { return catSecondary; }
+
     void run(ref<Store> store) override
     {
         settings.readOnlyMode = true;

src/nix/ls.cc

@@ -34,16 +34,14 @@ struct MixLs : virtual Args, MixJSON
                         (st.isExecutable ? "-r-xr-xr-x" : "-r--r--r--") :
                     st.type == FSAccessor::Type::tSymlink ? "lrwxrwxrwx" :
                     "dr-xr-xr-x";
-                std::cout <<
-                    (format("%s %20d %s") % tp % st.fileSize % relPath);
+                auto line = fmt("%s %20d %s", tp, st.fileSize, relPath);
                 if (st.type == FSAccessor::Type::tSymlink)
-                    std::cout << " -> " << accessor->readLink(curPath)
-                    ;
-                std::cout << "\n";
+                    line += " -> " + accessor->readLink(curPath);
+                logger->stdout(line);
                 if (recursive && st.type == FSAccessor::Type::tDirectory)
                     doPath(st, curPath, relPath, false);
             } else {
-                std::cout << relPath << "\n";
+                logger->stdout(relPath);
                 if (recursive) {
                     auto st = accessor->stat(curPath);
                     if (st.type == FSAccessor::Type::tDirectory)
@@ -102,9 +100,11 @@ struct CmdLsStore : StoreCommand, MixLs
 
     std::string description() override
     {
-        return "show information about a store path";
+        return "show information about a path in the Nix store";
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store) override
     {
         list(store->getFSAccessor());
@@ -133,12 +133,14 @@ struct CmdLsNar : Command, MixLs
 
     std::string description() override
     {
-        return "show information about the contents of a NAR file";
+        return "show information about a path inside a NAR file";
     }
 
+    Category category() override { return catUtility; }
+
     void run() override
     {
-        list(makeNarAccessor(make_ref<std::string>(readFile(narPath, true))));
+        list(makeNarAccessor(make_ref<std::string>(readFile(narPath))));
     }
 };
 

src/nix/main.cc

@@ -8,7 +8,7 @@
 #include "shared.hh"
 #include "store-api.hh"
 #include "progress-bar.hh"
-#include "download.hh"
+#include "filetransfer.hh"
 #include "finally.hh"
 
 #include <sys/types.h>
@@ -59,15 +59,22 @@ struct NixArgs : virtual MultiCommand, virtual MixCommonArgs
 
     NixArgs() : MultiCommand(*RegisterCommand::commands), MixCommonArgs("nix")
     {
-        mkFlag()
-            .longName("help")
-            .description("show usage information")
-            .handler([&]() { showHelpAndExit(); });
+        categories.clear();
+        categories[Command::catDefault] = "Main commands";
+        categories[catSecondary] = "Infrequently used commands";
+        categories[catUtility] = "Utility/scripting commands";
+        categories[catNixInstallation] = "Commands for upgrading or troubleshooting your Nix installation";
 
-        mkFlag()
-            .longName("help-config")
-            .description("show configuration options")
-            .handler([&]() {
+        addFlag({
+            .longName = "help",
+            .description = "show usage information",
+            .handler = {[&]() { showHelpAndExit(); }},
+        });
+
+        addFlag({
+            .longName = "help-config",
+            .description = "show configuration options",
+            .handler = {[&]() {
                 std::cout << "The following configuration options are available:\n\n";
                 Table2 tbl;
                 std::map<std::string, Config::SettingInfo> settings;
@@ -76,28 +83,33 @@ struct NixArgs : virtual MultiCommand, virtual MixCommonArgs
                     tbl.emplace_back(s.first, s.second.description);
                 printTable(std::cout, tbl);
                 throw Exit();
-            });
+            }},
+        });
 
-        mkFlag()
-            .longName("print-build-logs")
-            .shortName('L')
-            .description("print full build logs on stderr")
-            .set(&printBuildLogs, true);
+        addFlag({
+            .longName = "print-build-logs",
+            .shortName = 'L',
+            .description = "print full build logs on stderr",
+            .handler = {&printBuildLogs, true},
+        });
 
-        mkFlag()
-            .longName("version")
-            .description("show version information")
-            .handler([&]() { printVersion(programName); });
+        addFlag({
+            .longName = "version",
+            .description = "show version information",
+            .handler = {[&]() { printVersion(programName); }},
+        });
 
-        mkFlag()
-            .longName("no-net")
-            .description("disable substituters and consider all previously downloaded files up-to-date")
-            .handler([&]() { useNet = false; });
+        addFlag({
+            .longName = "no-net",
+            .description = "disable substituters and consider all previously downloaded files up-to-date",
+            .handler = {[&]() { useNet = false; }},
+        });
 
-        mkFlag()
-            .longName("refresh")
-            .description("consider all previously downloaded files out-of-date")
-            .handler([&]() { refresh = true; });
+        addFlag({
+            .longName = "refresh",
+            .description = "consider all previously downloaded files out-of-date",
+            .handler = {[&]() { refresh = true; }},
+        });
     }
 
     void printFlags(std::ostream & out) override
@@ -105,8 +117,8 @@ struct NixArgs : virtual MultiCommand, virtual MixCommonArgs
         Args::printFlags(out);
         std::cout <<
             "\n"
-            "In addition, most configuration settings can be overriden using '--<name> <value>'.\n"
-            "Boolean settings can be overriden using '--<name>' or '--no-<name>'. See 'nix\n"
+            "In addition, most configuration settings can be overriden using '--" ANSI_ITALIC "name value" ANSI_NORMAL "'.\n"
+            "Boolean settings can be overriden using '--" ANSI_ITALIC "name" ANSI_NORMAL "' or '--no-" ANSI_ITALIC "name" ANSI_NORMAL "'. See 'nix\n"
             "--help-config' for a list of configuration settings.\n";
     }
 
@@ -115,10 +127,10 @@ struct NixArgs : virtual MultiCommand, virtual MixCommonArgs
         MultiCommand::printHelp(programName, out);
 
 #if 0
-        out << "\nFor full documentation, run 'man " << programName << "' or 'man " << programName << "-<COMMAND>'.\n";
+        out << "\nFor full documentation, run 'man " << programName << "' or 'man " << programName << "-" ANSI_ITALIC "COMMAND" ANSI_NORMAL "'.\n";
 #endif
 
-        std::cout << "\nNote: this program is EXPERIMENTAL and subject to change.\n";
+        std::cout << "\nNote: this program is " ANSI_RED "EXPERIMENTAL" ANSI_NORMAL " and subject to change.\n";
     }
 
     void showHelpAndExit()
@@ -155,12 +167,15 @@ void mainWrapped(int argc, char * * argv)
 
     args.parseCmdline(argvToStrings(argc, argv));
 
-    settings.requireExperimentalFeature("nix-command");
-
     initPlugins();
 
     if (!args.command) args.showHelpAndExit();
 
+    if (args.command->first != "repl"
+        && args.command->first != "doctor"
+        && args.command->first != "upgrade-nix")
+        settings.requireExperimentalFeature("nix-command");
+
     Finally f([]() { stopProgressBar(); });
 
     startProgressBar(args.printBuildLogs);
@@ -176,17 +191,17 @@ void mainWrapped(int argc, char * * argv)
             settings.useSubstitutes = false;
         if (!settings.tarballTtl.overriden)
             settings.tarballTtl = std::numeric_limits<unsigned int>::max();
-        if (!downloadSettings.tries.overriden)
-            downloadSettings.tries = 0;
-        if (!downloadSettings.connectTimeout.overriden)
-            downloadSettings.connectTimeout = 1;
+        if (!fileTransferSettings.tries.overriden)
+            fileTransferSettings.tries = 0;
+        if (!fileTransferSettings.connectTimeout.overriden)
+            fileTransferSettings.connectTimeout = 1;
     }
 
     if (args.refresh)
         settings.tarballTtl = 0;
 
-    args.command->prepare();
-    args.command->run();
+    args.command->second->prepare();
+    args.command->second->run();
 }
 
 }

src/nix/make-content-addressable.cc

@@ -31,6 +31,9 @@ struct CmdMakeContentAddressable : StorePathsCommand, MixJSON
             },
         };
     }
+
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store, StorePaths storePaths) override
     {
         auto paths = store->topoSortPaths(storePathsToSet(storePaths));

src/nix/optimise-store.cc

@@ -23,6 +23,8 @@ struct CmdOptimiseStore : StoreCommand
         };
     }
 
+    Category category() override { return catUtility; }
+
     void run(ref<Store> store) override
     {
         store->optimiseStore();

src/nix/path-info.cc

@@ -29,6 +29,8 @@ struct CmdPathInfo : StorePathsCommand, MixJSON
         return "query information about store paths";
     }
 
+    Category category() override { return catSecondary; }
+
     Examples examples() override
     {
         return {