wroclaw-git-backups/nix
1
0
Fork 0
mirror of https://github.com/NixOS/nix synced 2025-07-07 14:21:48 +02:00
Code Activity

Revert flake-schemas for now

Browse source
This commit is contained in:
Eelco Dolstra 2024-10-29 15:18:48 +01:00
parent 6ff6baaa35
commit ffcc42faf4
25 changed files with 827 additions and 730 deletions
Download patch file Download diff file Expand all files Collapse all files

1
doc/manual/src/SUMMARY.md.in
View file

@ -114,7 +114,6 @@
- [Store Path Specification](protocols/store-path.md)
- [Nix Archive (NAR) Format](protocols/nix-archive.md)
- [Derivation "ATerm" file format](protocols/derivation-aterm.md)
- [Flake Schemas](protocols/flake-schemas.md)
- [C API](c-api.md)
- [Glossary](glossary.md)
- [Development](development/index.md)

64
doc/manual/src/protocols/flake-schemas.md
View file

@ -1,64 +0,0 @@
# Flake Schemas
Flake schemas are a mechanism to allow tools like `nix flake show` and `nix flake check` to enumerate and check the contents of a flake
in a generic way, without requiring built-in knowledge of specific flake output types like `packages` or `nixosConfigurations`.
A flake can define schemas for its outputs by defining a `schemas` output. `schemas` should be an attribute set with an attribute for
every output type that you want to be supported. If a flake does not have a `schemas` attribute, Nix uses a built-in set of schemas (namely https://github.com/DeterminateSystems/flake-schemas).
A schema is an attribute set with the following attributes:
| Attribute | Description | Default |
| :---------- | :---------------------------------------------------------------------------------------------- | :------ |
| `version` | Should be set to 1 | |
| `doc` | A string containing documentation about the flake output type in Markdown format. | |
| `allowIFD` | Whether the evaluation of the output attributes of this flake can read from derivation outputs. | `true` |
| `inventory` | A function that returns the contents of the flake output (described [below](#inventory)). | |
# Inventory
The `inventory` function returns a _node_ describing the contents of the flake output. A node is either a _leaf node_ or a _non-leaf node_. This allows nested flake output attributes to be described (e.g. `x86_64-linux.hello` inside a `packages` output).
Non-leaf nodes must have the following attribute:
| Attribute | Description |
| :--------- | :------------------------------------------------------------------------------------- |
| `children` | An attribute set of nodes. If this attribute is missing, the attribute is a leaf node. |
Leaf nodes can have the following attributes:
| Attribute | Description |
| :----------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `derivation` | The main derivation of this node, if any. It must evaluate for `nix flake check` and `nix flake show` to succeed. |
| `evalChecks` | An attribute set of Boolean values, used by `nix flake check`. Each attribute must evaluate to `true`. |
| `isFlakeCheck` | Whether `nix flake check` should build the `derivation` attribute of this node. |
| `shortDescription` | A one-sentence description of the node (such as the `meta.description` attribute in Nixpkgs). |
| `what` | A brief human-readable string describing the type of the node, e.g. `"package"` or `"development environment"`. This is used by tools like `nix flake show` to describe the contents of a flake. |
Both leaf and non-leaf nodes can have the following attributes:
| Attribute | Description |
| :----------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `forSystems` | A list of Nix system types (e.g. `["x86_64-linux"]`) supported by this node. This is used by tools to skip nodes that cannot be built on the user's system. Setting this on a non-leaf node allows all the children to be skipped, regardless of the `forSystems` attributes of the children. If this attribute is not set, the node is never skipped. |
# Example
Here is a schema that checks that every element of the `nixosConfigurations` flake output evaluates and builds correctly (meaning that it has a `config.system.build.toplevel` attribute that yields a buildable derivation).
```nix
outputs = {
schemas.nixosConfigurations = {
version = 1;
doc = ''
The `nixosConfigurations` flake output defines NixOS system configurations.
'';
inventory = output: {
children = builtins.mapAttrs (configName: machine:
{
what = "NixOS configuration";
derivation = machine.config.system.build.toplevel;
}) output;
};
};
};
```