Initial commit

This commit is contained in:
Victor Fuentes 2024-05-01 18:14:04 -04:00
commit 74ee9adefc
Signed by: vlinkz
GPG key ID: 0A88B68D6A9ACAE0
222 changed files with 28547 additions and 0 deletions

1
.gitignore vendored Normal file
View file

@ -0,0 +1 @@
/result

21
LICENSE Normal file
View file

@ -0,0 +1,21 @@
Copyright (c) 2003-2024 Eelco Dolstra and the Nixpkgs/NixOS contributors
Copyright (c) 2024 the Aux contributors
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

27
flake.lock generated Normal file
View file

@ -0,0 +1,27 @@
{
"nodes": {
"nixpkgs": {
"locked": {
"lastModified": 1714253743,
"narHash": "sha256-mdTQw2XlariysyScCv2tTE45QSU9v/ezLcHJ22f0Nxc=",
"owner": "auxolotl",
"repo": "nixpkgs",
"rev": "58a1abdbae3217ca6b702f03d3b35125d88a2994",
"type": "github"
},
"original": {
"owner": "auxolotl",
"ref": "nixos-unstable",
"repo": "nixpkgs",
"type": "github"
}
},
"root": {
"inputs": {
"nixpkgs": "nixpkgs"
}
}
},
"root": "root",
"version": 7
}

28
flake.nix Normal file
View file

@ -0,0 +1,28 @@
{
inputs = {
nixpkgs.url = "github:auxolotl/nixpkgs/nixos-unstable";
};
outputs = { self, nixpkgs, ... }:
let
lib = import ./lib;
forAllSystems = lib.genAttrs lib.systems.flakeExposed;
in
{
nixPackages = forAllSystems (system:
(import nixpkgs { inherit system; })
);
# auxPackages = forAllSystems (system:
# (import ./. { inherit system; })
# );
# To test, run nix build .#tests.x86_64-linux.release
tests = forAllSystems (system:
({
systems = import ./lib/tests/systems.nix;
release = import ./lib/tests/release.nix { pkgs = self.nixPackages.${system}; };
})
);
};
}

1
lib/.version Normal file
View file

@ -0,0 +1 @@
24.05

159
lib/README.md Normal file
View file

@ -0,0 +1,159 @@
# Nixpkgs lib
This directory contains the implementation, documentation and tests for the Nixpkgs `lib` library.
## Overview
The evaluation entry point for `lib` is [`default.nix`](default.nix).
This file evaluates to an attribute set containing two separate kinds of attributes:
- Sub-libraries:
Attribute sets grouping together similar functionality.
Each sub-library is defined in a separate file usually matching its attribute name.
Example: `lib.lists` is a sub-library containing list-related functionality such as `lib.lists.take` and `lib.lists.imap0`.
These are defined in the file [`lists.nix`](lists.nix).
- Aliases:
Attributes that point to an attribute of the same name in some sub-library.
Example: `lib.take` is an alias for `lib.lists.take`.
Most files in this directory are definitions of sub-libraries, but there are a few others:
- [`minver.nix`](minver.nix): A string of the minimum version of Nix that is required to evaluate Nixpkgs.
- [`tests`](tests): Tests, see [Running tests](#running-tests)
- [`release.nix`](tests/release.nix): A derivation aggregating all tests
- [`misc.nix`](tests/misc.nix): Evaluation unit tests for most sub-libraries
- `*.sh`: Bash scripts that run tests for specific sub-libraries
- All other files in this directory exist to support the tests
- [`systems`](systems): The `lib.systems` sub-library, structured into a directory instead of a file due to its complexity
- [`path`](path): The `lib.path` sub-library, which includes tests as well as a document describing the design goals of `lib.path`
- All other files in this directory are sub-libraries
### Module system
The [module system](https://nixos.org/manual/nixpkgs/#module-system) spans multiple sub-libraries:
- [`modules.nix`](modules.nix): `lib.modules` for the core functions and anything not relating to option definitions
- [`options.nix`](options.nix): `lib.options` for anything relating to option definitions
- [`types.nix`](types.nix): `lib.types` for module system types
## PR Guidelines
Follow these guidelines for proposing a change to the interface of `lib`.
### Provide a Motivation
Clearly describe why the change is necessary and its use cases.
Make sure that the change benefits the user more than the added mental effort of looking it up and keeping track of its definition.
If the same can reasonably be done with the existing interface,
consider just updating the documentation with more examples and links.
This is also known as the [Fairbairn Threshold](https://wiki.haskell.org/Fairbairn_threshold).
Through this principle we avoid the human cost of duplicated functionality in an overly large library.
### Make one PR for each change
Don't have multiple changes in one PR, instead split it up into multiple ones.
This keeps the conversation focused and has a higher chance of getting merged.
### Name the interface appropriately
When introducing new names to the interface, such as new function, or new function attributes,
make sure to name it appropriately.
Names should be self-explanatory and consistent with the rest of `lib`.
If there's no obvious best name, include the alternatives you considered.
### Write documentation
Update the [reference documentation](#reference-documentation) to reflect the change.
Be generous with links to related functionality.
### Write tests
Add good test coverage for the change, including:
- Tests for edge cases, such as empty values or lists.
- Tests for tricky inputs, such as a string with string context or a path that doesn't exist.
- Test all code paths, such as `if-then-else` branches and returned attributes.
- If the tests for the sub-library are written in bash,
test messages of custom errors, such as `throw` or `abortMsg`,
At the time this is only not necessary for sub-libraries tested with [`tests/misc.nix`](./tests/misc.nix).
See [running tests](#running-tests) for more details on the test suites.
### Write tidy code
Name variables well, even if they're internal.
The code should be as self-explanatory as possible.
Be generous with code comments when appropriate.
As a baseline, follow the [Nixpkgs code conventions](https://github.com/NixOS/nixpkgs/blob/master/CONTRIBUTING.md#code-conventions).
### Write efficient code
Nix generally does not have free abstractions.
Be aware that seemingly straightforward changes can cause more allocations and a decrease in performance.
That said, don't optimise prematurely, especially in new code.
## Reference documentation
Reference documentation for library functions is written above each function as a multi-line comment.
These comments are processed using [nixdoc](https://github.com/nix-community/nixdoc) and [rendered in the Nixpkgs manual](https://nixos.org/manual/nixpkgs/stable/#chap-functions).
The nixdoc README describes the [comment format](https://github.com/nix-community/nixdoc#comment-format).
See [doc/README.md](../doc/README.md) for how to build the manual.
## Running tests
All library tests can be run by building the derivation in [`tests/release.nix`](tests/release.nix):
```bash
nix-build tests/release.nix
```
Some commands for quicker iteration over parts of the test suite are also available:
```bash
# Run all evaluation unit tests in tests/misc.nix
# if the resulting list is empty, all tests passed
nix-instantiate --eval --strict tests/misc.nix
# Run the module system tests
tests/modules.sh
# Run the lib.sources tests
tests/sources.sh
# Run the lib.filesystem tests
tests/filesystem.sh
# Run the lib.path property tests
path/tests/prop.sh
# Run the lib.fileset tests
fileset/tests.sh
```
## Commit conventions
- Make sure you read about the [commit conventions](../CONTRIBUTING.md#commit-conventions) common to Nixpkgs as a whole.
- Format the commit messages in the following way:
```
lib.(section): (init | add additional argument | refactor | etc)
(Motivation for change. Additional information.)
```
Examples:
* lib.getExe': check arguments
* lib.fileset: Add an additional argument in the design docs
Closes #264537

99
lib/ascii-table.nix Normal file
View file

@ -0,0 +1,99 @@
{ "\t" = 9;
"\n" = 10;
"\r" = 13;
" " = 32;
"!" = 33;
"\"" = 34;
"#" = 35;
"$" = 36;
"%" = 37;
"&" = 38;
"'" = 39;
"(" = 40;
")" = 41;
"*" = 42;
"+" = 43;
"," = 44;
"-" = 45;
"." = 46;
"/" = 47;
"0" = 48;
"1" = 49;
"2" = 50;
"3" = 51;
"4" = 52;
"5" = 53;
"6" = 54;
"7" = 55;
"8" = 56;
"9" = 57;
":" = 58;
";" = 59;
"<" = 60;
"=" = 61;
">" = 62;
"?" = 63;
"@" = 64;
"A" = 65;
"B" = 66;
"C" = 67;
"D" = 68;
"E" = 69;
"F" = 70;
"G" = 71;
"H" = 72;
"I" = 73;
"J" = 74;
"K" = 75;
"L" = 76;
"M" = 77;
"N" = 78;
"O" = 79;
"P" = 80;
"Q" = 81;
"R" = 82;
"S" = 83;
"T" = 84;
"U" = 85;
"V" = 86;
"W" = 87;
"X" = 88;
"Y" = 89;
"Z" = 90;
"[" = 91;
"\\" = 92;
"]" = 93;
"^" = 94;
"_" = 95;
"`" = 96;
"a" = 97;
"b" = 98;
"c" = 99;
"d" = 100;
"e" = 101;
"f" = 102;
"g" = 103;
"h" = 104;
"i" = 105;
"j" = 106;
"k" = 107;
"l" = 108;
"m" = 109;
"n" = 110;
"o" = 111;
"p" = 112;
"q" = 113;
"r" = 114;
"s" = 115;
"t" = 116;
"u" = 117;
"v" = 118;
"w" = 119;
"x" = 120;
"y" = 121;
"z" = 122;
"{" = 123;
"|" = 124;
"}" = 125;
"~" = 126;
}

144
lib/asserts.nix Normal file
View file

@ -0,0 +1,144 @@
{ lib }:
rec {
/**
Throw if pred is false, else return pred.
Intended to be used to augment asserts with helpful error messages.
# Inputs
`pred`
: Predicate that needs to succeed, otherwise `msg` is thrown
`msg`
: Message to throw in case `pred` fails
# Type
```
assertMsg :: Bool -> String -> Bool
```
# Examples
:::{.example}
## `lib.asserts.assertMsg` usage example
```nix
assertMsg false "nope"
stderr> error: nope
assert assertMsg ("foo" == "bar") "foo is not bar, silly"; ""
stderr> error: foo is not bar, silly
```
:::
*/
# TODO(Profpatsch): add tests that check stderr
assertMsg =
pred:
msg:
pred || builtins.throw msg;
/**
Specialized `assertMsg` for checking if `val` is one of the elements
of the list `xs`. Useful for checking enums.
# Inputs
`name`
: The name of the variable the user entered `val` into, for inclusion in the error message
`val`
: The value of what the user provided, to be compared against the values in `xs`
`xs`
: The list of valid values
# Type
```
assertOneOf :: String -> ComparableVal -> List ComparableVal -> Bool
```
# Examples
:::{.example}
## `lib.asserts.assertOneOf` usage example
```nix
let sslLibrary = "libressl";
in assertOneOf "sslLibrary" sslLibrary [ "openssl" "bearssl" ]
stderr> error: sslLibrary must be one of [
stderr> "openssl"
stderr> "bearssl"
stderr> ], but is: "libressl"
```
:::
*/
assertOneOf =
name:
val:
xs:
assertMsg
(lib.elem val xs)
"${name} must be one of ${
lib.generators.toPretty {} xs}, but is: ${
lib.generators.toPretty {} val}";
/**
Specialized `assertMsg` for checking if every one of `vals` is one of the elements
of the list `xs`. Useful for checking lists of supported attributes.
# Inputs
`name`
: The name of the variable the user entered `val` into, for inclusion in the error message
`vals`
: The list of values of what the user provided, to be compared against the values in `xs`
`xs`
: The list of valid values
# Type
```
assertEachOneOf :: String -> List ComparableVal -> List ComparableVal -> Bool
```
# Examples
:::{.example}
## `lib.asserts.assertEachOneOf` usage example
```nix
let sslLibraries = [ "libressl" "bearssl" ];
in assertEachOneOf "sslLibraries" sslLibraries [ "openssl" "bearssl" ]
stderr> error: each element in sslLibraries must be one of [
stderr> "openssl"
stderr> "bearssl"
stderr> ], but is: [
stderr> "libressl"
stderr> "bearssl"
stderr> ]
```
:::
*/
assertEachOneOf =
name:
vals:
xs:
assertMsg
(lib.all (val: lib.elem val xs) vals)
"each element in ${name} must be one of ${
lib.generators.toPretty {} xs}, but is: ${
lib.generators.toPretty {} vals}";
}

2040
lib/attrsets.nix Normal file

File diff suppressed because it is too large Load diff

104
lib/cli.nix Normal file
View file

@ -0,0 +1,104 @@
{ lib }:
rec {
/**
Automatically convert an attribute set to command-line options.
This helps protect against malformed command lines and also to reduce
boilerplate related to command-line construction for simple use cases.
`toGNUCommandLine` returns a list of nix strings.
`toGNUCommandLineShell` returns an escaped shell string.
# Inputs
`options`
: 1\. Function argument
`attrs`
: 2\. Function argument
# Examples
:::{.example}
## `lib.cli.toGNUCommandLineShell` usage example
```nix
cli.toGNUCommandLine {} {
data = builtins.toJSON { id = 0; };
X = "PUT";
retry = 3;
retry-delay = null;
url = [ "https://example.com/foo" "https://example.com/bar" ];
silent = false;
verbose = true;
}
=> [
"-X" "PUT"
"--data" "{\"id\":0}"
"--retry" "3"
"--url" "https://example.com/foo"
"--url" "https://example.com/bar"
"--verbose"
]
cli.toGNUCommandLineShell {} {
data = builtins.toJSON { id = 0; };
X = "PUT";
retry = 3;
retry-delay = null;
url = [ "https://example.com/foo" "https://example.com/bar" ];
silent = false;
verbose = true;
}
=> "'-X' 'PUT' '--data' '{\"id\":0}' '--retry' '3' '--url' 'https://example.com/foo' '--url' 'https://example.com/bar' '--verbose'";
```
:::
*/
toGNUCommandLineShell =
options: attrs: lib.escapeShellArgs (toGNUCommandLine options attrs);
toGNUCommandLine = {
# how to string-format the option name;
# by default one character is a short option (`-`),
# more than one characters a long option (`--`).
mkOptionName ?
k: if builtins.stringLength k == 1
then "-${k}"
else "--${k}",
# how to format a boolean value to a command list;
# by default its a flag option
# (only the option name if true, left out completely if false).
mkBool ? k: v: lib.optional v (mkOptionName k),
# how to format a list value to a command list;
# by default the option name is repeated for each value
# and `mkOption` is applied to the values themselves.
mkList ? k: v: lib.concatMap (mkOption k) v,
# how to format any remaining value to a command list;
# on the toplevel, booleans and lists are handled by `mkBool` and `mkList`,
# though they can still appear as values of a list.
# By default, everything is printed verbatim and complex types
# are forbidden (lists, attrsets, functions). `null` values are omitted.
mkOption ?
k: v: if v == null
then []
else [ (mkOptionName k) (lib.generators.mkValueStringDefault {} v) ]
}:
options:
let
render = k: v:
if builtins.isBool v then mkBool k v
else if builtins.isList v then mkList k v
else mkOption k v;
in
builtins.concatLists (lib.mapAttrsToList render options);
}

664
lib/customisation.nix Normal file
View file

@ -0,0 +1,664 @@
{ lib }:
let
inherit (builtins)
intersectAttrs;
inherit (lib)
functionArgs isFunction mirrorFunctionArgs isAttrs setFunctionArgs
optionalAttrs attrNames filter elemAt concatStringsSep sortOn take length
filterAttrs optionalString flip pathIsDirectory head pipe isDerivation listToAttrs
mapAttrs seq flatten deepSeq warnIf isInOldestRelease extends
;
inherit (lib.strings) levenshtein levenshteinAtMost;
in
rec {
/**
`overrideDerivation drv f` takes a derivation (i.e., the result
of a call to the builtin function `derivation`) and returns a new
derivation in which the attributes of the original are overridden
according to the function `f`. The function `f` is called with
the original derivation attributes.
`overrideDerivation` allows certain "ad-hoc" customisation
scenarios (e.g. in ~/.config/nixpkgs/config.nix). For instance,
if you want to "patch" the derivation returned by a package
function in Nixpkgs to build another version than what the
function itself provides.
For another application, see build-support/vm, where this
function is used to build arbitrary derivations inside a QEMU
virtual machine.
Note that in order to preserve evaluation errors, the new derivation's
outPath depends on the old one's, which means that this function cannot
be used in circular situations when the old derivation also depends on the
new one.
You should in general prefer `drv.overrideAttrs` over this function;
see the nixpkgs manual for more information on overriding.
# Inputs
`drv`
: 1\. Function argument
`f`
: 2\. Function argument
# Type
```
overrideDerivation :: Derivation -> ( Derivation -> AttrSet ) -> Derivation
```
# Examples
:::{.example}
## `lib.customisation.overrideDerivation` usage example
```nix
mySed = overrideDerivation pkgs.gnused (oldAttrs: {
name = "sed-4.2.2-pre";
src = fetchurl {
url = ftp://alpha.gnu.org/gnu/sed/sed-4.2.2-pre.tar.bz2;
hash = "sha256-MxBJRcM2rYzQYwJ5XKxhXTQByvSg5jZc5cSHEZoB2IY=";
};
patches = [];
});
```
:::
*/
overrideDerivation = drv: f:
let
newDrv = derivation (drv.drvAttrs // (f drv));
in flip (extendDerivation (seq drv.drvPath true)) newDrv (
{ meta = drv.meta or {};
passthru = if drv ? passthru then drv.passthru else {};
}
//
(drv.passthru or {})
//
optionalAttrs (drv ? __spliced) {
__spliced = {} // (mapAttrs (_: sDrv: overrideDerivation sDrv f) drv.__spliced);
});
/**
`makeOverridable` takes a function from attribute set to attribute set and
injects `override` attribute which can be used to override arguments of
the function.
Please refer to documentation on [`<pkg>.overrideDerivation`](#sec-pkg-overrideDerivation) to learn about `overrideDerivation` and caveats
related to its use.
# Inputs
`f`
: 1\. Function argument
# Type
```
makeOverridable :: (AttrSet -> a) -> AttrSet -> a
```
# Examples
:::{.example}
## `lib.customisation.makeOverridable` usage example
```nix
nix-repl> x = {a, b}: { result = a + b; }
nix-repl> y = lib.makeOverridable x { a = 1; b = 2; }
nix-repl> y
{ override = «lambda»; overrideDerivation = «lambda»; result = 3; }
nix-repl> y.override { a = 10; }
{ override = «lambda»; overrideDerivation = «lambda»; result = 12; }
```
:::
*/
makeOverridable = f:
let
# Creates a functor with the same arguments as f
mirrorArgs = mirrorFunctionArgs f;
in
mirrorArgs (origArgs:
let
result = f origArgs;
# Changes the original arguments with (potentially a function that returns) a set of new attributes
overrideWith = newArgs: origArgs // (if isFunction newArgs then newArgs origArgs else newArgs);
# Re-call the function but with different arguments
overrideArgs = mirrorArgs (newArgs: makeOverridable f (overrideWith newArgs));
# Change the result of the function call by applying g to it
overrideResult = g: makeOverridable (mirrorArgs (args: g (f args))) origArgs;
in
if isAttrs result then
result // {
override = overrideArgs;
overrideDerivation = fdrv: overrideResult (x: overrideDerivation x fdrv);
${if result ? overrideAttrs then "overrideAttrs" else null} = fdrv:
overrideResult (x: x.overrideAttrs fdrv);
}
else if isFunction result then
# Transform the result into a functor while propagating its arguments
setFunctionArgs result (functionArgs result) // {
override = overrideArgs;
}
else result);
/**
Call the package function in the file `fn` with the required
arguments automatically. The function is called with the
arguments `args`, but any missing arguments are obtained from
`autoArgs`. This function is intended to be partially
parameterised, e.g.,
```nix
callPackage = callPackageWith pkgs;
pkgs = {
libfoo = callPackage ./foo.nix { };
libbar = callPackage ./bar.nix { };
};
```
If the `libbar` function expects an argument named `libfoo`, it is
automatically passed as an argument. Overrides or missing
arguments can be supplied in `args`, e.g.
```nix
libbar = callPackage ./bar.nix {
libfoo = null;
enableX11 = true;
};
```
<!-- TODO: Apply "Example:" tag to the examples above -->
# Inputs
`autoArgs`
: 1\. Function argument
`fn`
: 2\. Function argument
`args`
: 3\. Function argument
# Type
```
callPackageWith :: AttrSet -> ((AttrSet -> a) | Path) -> AttrSet -> a
```
*/
callPackageWith = autoArgs: fn: args:
let
f = if isFunction fn then fn else import fn;
fargs = functionArgs f;
# All arguments that will be passed to the function
# This includes automatic ones and ones passed explicitly
allArgs = intersectAttrs fargs autoArgs // args;
# a list of argument names that the function requires, but
# wouldn't be passed to it
missingArgs =
# Filter out arguments that have a default value
(filterAttrs (name: value: ! value)
# Filter out arguments that would be passed
(removeAttrs fargs (attrNames allArgs)));
# Get a list of suggested argument names for a given missing one
getSuggestions = arg: pipe (autoArgs // args) [
attrNames
# Only use ones that are at most 2 edits away. While mork would work,
# levenshteinAtMost is only fast for 2 or less.
(filter (levenshteinAtMost 2 arg))
# Put strings with shorter distance first
(sortOn (levenshtein arg))
# Only take the first couple results
(take 3)
# Quote all entries
(map (x: "\"" + x + "\""))
];
prettySuggestions = suggestions:
if suggestions == [] then ""
else if length suggestions == 1 then ", did you mean ${elemAt suggestions 0}?"
else ", did you mean ${concatStringsSep ", " (lib.init suggestions)} or ${lib.last suggestions}?";
errorForArg = arg:
let
loc = builtins.unsafeGetAttrPos arg fargs;
# loc' can be removed once lib/minver.nix is >2.3.4, since that includes
# https://github.com/NixOS/nix/pull/3468 which makes loc be non-null
loc' = if loc != null then loc.file + ":" + toString loc.line
else if ! isFunction fn then
toString fn + optionalString (pathIsDirectory fn) "/default.nix"
else "<unknown location>";
in "Function called without required argument \"${arg}\" at "
+ "${loc'}${prettySuggestions (getSuggestions arg)}";
# Only show the error for the first missing argument
error = errorForArg (head (attrNames missingArgs));
in if missingArgs == {}
then makeOverridable f allArgs
# This needs to be an abort so it can't be caught with `builtins.tryEval`,
# which is used by nix-env and ofborg to filter out packages that don't evaluate.
# This way we're forced to fix such errors in Nixpkgs,
# which is especially relevant with allowAliases = false
else abort "lib.customisation.callPackageWith: ${error}";
/**
Like callPackage, but for a function that returns an attribute
set of derivations. The override function is added to the
individual attributes.
# Inputs
`autoArgs`
: 1\. Function argument
`fn`
: 2\. Function argument
`args`
: 3\. Function argument
# Type
```
callPackagesWith :: AttrSet -> ((AttrSet -> AttrSet) | Path) -> AttrSet -> AttrSet
```
*/
callPackagesWith = autoArgs: fn: args:
let
f = if isFunction fn then fn else import fn;
auto = intersectAttrs (functionArgs f) autoArgs;
mirrorArgs = mirrorFunctionArgs f;
origArgs = auto // args;
pkgs = f origArgs;
mkAttrOverridable = name: _: makeOverridable (mirrorArgs (newArgs: (f newArgs).${name})) origArgs;
in
if isDerivation pkgs then throw
("function `callPackages` was called on a *single* derivation "
+ ''"${pkgs.name or "<unknown-name>"}";''
+ " did you mean to use `callPackage` instead?")
else mapAttrs mkAttrOverridable pkgs;
/**
Add attributes to each output of a derivation without changing
the derivation itself and check a given condition when evaluating.
# Inputs
`condition`
: 1\. Function argument
`passthru`
: 2\. Function argument
`drv`
: 3\. Function argument
# Type
```
extendDerivation :: Bool -> Any -> Derivation -> Derivation
```
*/
extendDerivation = condition: passthru: drv:
let
outputs = drv.outputs or [ "out" ];
commonAttrs = drv // (listToAttrs outputsList) //
({ all = map (x: x.value) outputsList; }) // passthru;
outputToAttrListElement = outputName:
{ name = outputName;
value = commonAttrs // {
inherit (drv.${outputName}) type outputName;
outputSpecified = true;
drvPath = assert condition; drv.${outputName}.drvPath;
outPath = assert condition; drv.${outputName}.outPath;
} //
# TODO: give the derivation control over the outputs.
# `overrideAttrs` may not be the only attribute that needs
# updating when switching outputs.
optionalAttrs (passthru?overrideAttrs) {
# TODO: also add overrideAttrs when overrideAttrs is not custom, e.g. when not splicing.
overrideAttrs = f: (passthru.overrideAttrs f).${outputName};
};
};
outputsList = map outputToAttrListElement outputs;
in commonAttrs // {
drvPath = assert condition; drv.drvPath;
outPath = assert condition; drv.outPath;
};
/**
Strip a derivation of all non-essential attributes, returning
only those needed by hydra-eval-jobs. Also strictly evaluate the
result to ensure that there are no thunks kept alive to prevent
garbage collection.
# Inputs
`drv`
: 1\. Function argument
# Type
```
hydraJob :: (Derivation | Null) -> (Derivation | Null)
```
*/
hydraJob = drv:
let
outputs = drv.outputs or ["out"];
commonAttrs =
{ inherit (drv) name system meta; inherit outputs; }
// optionalAttrs (drv._hydraAggregate or false) {
_hydraAggregate = true;
constituents = map hydraJob (flatten drv.constituents);
}
// (listToAttrs outputsList);
makeOutput = outputName:
let output = drv.${outputName}; in
{ name = outputName;
value = commonAttrs // {
outPath = output.outPath;
drvPath = output.drvPath;
type = "derivation";
inherit outputName;
};
};
outputsList = map makeOutput outputs;
drv' = (head outputsList).value;
in if drv == null then null else
deepSeq drv' drv';
/**
Make an attribute set (a "scope") from functions that take arguments from that same attribute set.
See [](#ex-makeScope) for how to use it.
# Inputs
1. `newScope` (`AttrSet -> ((AttrSet -> a) | Path) -> AttrSet -> a`)
A function that takes an attribute set `attrs` and returns what ends up as `callPackage` in the output.
Typical values are `callPackageWith` or the output attribute `newScope`.
2. `f` (`AttrSet -> AttrSet`)
A function that takes an attribute set as returned by `makeScope newScope f` (a "scope") and returns any attribute set.
This function is used to compute the fixpoint of the resulting scope using `callPackage`.
Its argument is the lazily evaluated reference to the value of that fixpoint, and is typically called `self` or `final`.
See [](#ex-makeScope) for how to use it.
See [](#sec-functions-library-fixedPoints) for details on fixpoint computation.
# Output
`makeScope` returns an attribute set of a form called `scope`, which also contains the final attributes produced by `f`:
```
scope :: {
callPackage :: ((AttrSet -> a) | Path) -> AttrSet -> a
newScope = AttrSet -> scope
overrideScope = (scope -> scope -> AttrSet) -> scope
packages :: AttrSet -> AttrSet
}
```
- `callPackage` (`((AttrSet -> a) | Path) -> AttrSet -> a`)
A function that
1. Takes a function `p`, or a path to a Nix file that contains a function `p`, which takes an attribute set and returns value of arbitrary type `a`,
2. Takes an attribute set `args` with explicit attributes to pass to `p`,
3. Calls `f` with attributes from the original attribute set `attrs` passed to `newScope` updated with `args, i.e. `attrs // args`, if they match the attributes in the argument of `p`.
All such functions `p` will be called with the same value for `attrs`.
See [](#ex-makeScope-callPackage) for how to use it.
- `newScope` (`AttrSet -> scope`)
Takes an attribute set `attrs` and returns a scope that extends the original scope.
- `overrideScope` (`(scope -> scope -> AttrSet) -> scope`)
Takes a function `g` of the form `final: prev: { # attributes }` to act as an overlay on `f`, and returns a new scope with values determined by `extends g f`.
See [](https://nixos.org/manual/nixpkgs/unstable/#function-library-lib.fixedPoints.extends) for details.
This allows subsequent modification of the final attribute set in a consistent way, i.e. all functions `p` invoked with `callPackage` will be called with the modified values.
- `packages` (`AttrSet -> AttrSet`)
The value of the argument `f` to `makeScope`.
- final attributes
The final values returned by `f`.
# Examples
:::{#ex-makeScope .example}
# Create an interdependent package set on top of `pkgs`
The functions in `foo.nix` and `bar.nix` can depend on each other, in the sense that `foo.nix` can contain a function that expects `bar` as an attribute in its argument.
```nix
let
pkgs = import <nixpkgs> { };
in
pkgs.lib.makeScope pkgs.newScope (self: {
foo = self.callPackage ./foo.nix { };
bar = self.callPackage ./bar.nix { };
})
```
evaluates to
```nix
{
callPackage = «lambda»;
newScope = «lambda»;
overrideScope = «lambda»;
packages = «lambda»;
foo = «derivation»;
bar = «derivation»;
}
```
:::
:::{#ex-makeScope-callPackage .example}
# Using `callPackage` from a scope
```nix
let
pkgs = import <nixpkgs> { };
inherit (pkgs) lib;
scope = lib.makeScope lib.callPackageWith (self: { a = 1; b = 2; });
three = scope.callPackage ({ a, b }: a + b) { };
four = scope.callPackage ({ a, b }: a + b) { a = 2; };
in
[ three four ]
```
evaluates to
```nix
[ 3 4 ]
```
:::
# Type
```
makeScope :: (AttrSet -> ((AttrSet -> a) | Path) -> AttrSet -> a) -> (AttrSet -> AttrSet) -> scope
```
*/
makeScope = newScope: f:
let self = f self // {
newScope = scope: newScope (self // scope);
callPackage = self.newScope {};
overrideScope = g: makeScope newScope (extends g f);
# Remove after 24.11 is released.
overrideScope' = g: warnIf (isInOldestRelease 2311)
"`overrideScope'` (from `lib.makeScope`) has been renamed to `overrideScope`."
(makeScope newScope (extends g f));
packages = f;
};
in self;
/**
backward compatibility with old uncurried form; deprecated
# Inputs
`splicePackages`
: 1\. Function argument
`newScope`
: 2\. Function argument
`otherSplices`
: 3\. Function argument
`keep`
: 4\. Function argument
`extra`
: 5\. Function argument
`f`
: 6\. Function argument
*/
makeScopeWithSplicing =
splicePackages: newScope: otherSplices: keep: extra: f:
makeScopeWithSplicing'
{ inherit splicePackages newScope; }
{ inherit otherSplices keep extra f; };
/**
Like makeScope, but aims to support cross compilation. It's still ugly, but
hopefully it helps a little bit.
# Type
```
makeScopeWithSplicing' ::
{ splicePackages :: Splice -> AttrSet
, newScope :: AttrSet -> ((AttrSet -> a) | Path) -> AttrSet -> a
}
-> { otherSplices :: Splice, keep :: AttrSet -> AttrSet, extra :: AttrSet -> AttrSet }
-> AttrSet
Splice ::
{ pkgsBuildBuild :: AttrSet
, pkgsBuildHost :: AttrSet
, pkgsBuildTarget :: AttrSet
, pkgsHostHost :: AttrSet
, pkgsHostTarget :: AttrSet
, pkgsTargetTarget :: AttrSet
}
```
*/
makeScopeWithSplicing' =
{ splicePackages
, newScope
}:
{ otherSplices
# Attrs from `self` which won't be spliced.
# Avoid using keep, it's only used for a python hook workaround, added in PR #104201.
# ex: `keep = (self: { inherit (self) aAttr; })`
, keep ? (_self: {})
# Additional attrs to add to the sets `callPackage`.
# When the package is from a subset (but not a subset within a package IS #211340)
# within `spliced0` it will be spliced.
# When using an package outside the set but it's available from `pkgs`, use the package from `pkgs.__splicedPackages`.
# If the package is not available within the set or in `pkgs`, such as a package in a let binding, it will not be spliced
# ex:
# ```
# nix-repl> darwin.apple_sdk.frameworks.CoreFoundation
# «derivation ...CoreFoundation-11.0.0.drv»
# nix-repl> darwin.CoreFoundation
# error: attribute 'CoreFoundation' missing
# nix-repl> darwin.callPackage ({ CoreFoundation }: CoreFoundation) { }
# «derivation ...CoreFoundation-11.0.0.drv»
# ```
, extra ? (_spliced0: {})
, f
}:
let
spliced0 = splicePackages {
pkgsBuildBuild = otherSplices.selfBuildBuild;
pkgsBuildHost = otherSplices.selfBuildHost;
pkgsBuildTarget = otherSplices.selfBuildTarget;
pkgsHostHost = otherSplices.selfHostHost;
pkgsHostTarget = self; # Not `otherSplices.selfHostTarget`;
pkgsTargetTarget = otherSplices.selfTargetTarget;
};
spliced = extra spliced0 // spliced0 // keep self;
self = f self // {
newScope = scope: newScope (spliced // scope);
callPackage = newScope spliced; # == self.newScope {};
# N.B. the other stages of the package set spliced in are *not*
# overridden.
overrideScope = g: (makeScopeWithSplicing'
{ inherit splicePackages newScope; }
{ inherit otherSplices keep extra;
f = extends g f;
});
packages = f;
};
in self;
}

246
lib/debug.nix Normal file
View file

@ -0,0 +1,246 @@
/* Collection of functions useful for debugging
broken nix expressions.
* `trace`-like functions take two values, print
the first to stderr and return the second.
* `traceVal`-like functions take one argument
which both printed and returned.
* `traceSeq`-like functions fully evaluate their
traced value before printing (not just to weak
head normal form like trace does by default).
* Functions that end in `-Fn` take an additional
function as their first argument, which is applied
to the traced value before it is printed.
*/
{ lib }:
let
inherit (lib)
isList
isAttrs
substring
attrValues
concatLists
const
elem
generators
id
mapAttrs
trace;
in
rec {
# -- TRACING --
/* Conditionally trace the supplied message, based on a predicate.
Type: traceIf :: bool -> string -> a -> a
Example:
traceIf true "hello" 3
trace: hello
=> 3
*/
traceIf =
# Predicate to check
pred:
# Message that should be traced
msg:
# Value to return
x: if pred then trace msg x else x;
/* Trace the supplied value after applying a function to it, and
return the original value.
Type: traceValFn :: (a -> b) -> a -> a
Example:
traceValFn (v: "mystring ${v}") "foo"
trace: mystring foo
=> "foo"
*/
traceValFn =
# Function to apply
f:
# Value to trace and return
x: trace (f x) x;
/* Trace the supplied value and return it.
Type: traceVal :: a -> a
Example:
traceVal 42
# trace: 42
=> 42
*/
traceVal = traceValFn id;
/* `builtins.trace`, but the value is `builtins.deepSeq`ed first.
Type: traceSeq :: a -> b -> b
Example:
trace { a.b.c = 3; } null
trace: { a = <CODE>; }
=> null
traceSeq { a.b.c = 3; } null
trace: { a = { b = { c = 3; }; }; }
=> null
*/
traceSeq =
# The value to trace
x:
# The value to return
y: trace (builtins.deepSeq x x) y;
/* Like `traceSeq`, but only evaluate down to depth n.
This is very useful because lots of `traceSeq` usages
lead to an infinite recursion.
Example:
traceSeqN 2 { a.b.c = 3; } null
trace: { a = { b = {}; }; }
=> null
Type: traceSeqN :: Int -> a -> b -> b
*/
traceSeqN = depth: x: y:
let snip = v: if isList v then noQuotes "[]" v
else if isAttrs v then noQuotes "{}" v
else v;
noQuotes = str: v: { __pretty = const str; val = v; };
modify = n: fn: v: if (n == 0) then fn v
else if isList v then map (modify (n - 1) fn) v
else if isAttrs v then mapAttrs
(const (modify (n - 1) fn)) v
else v;
in trace (generators.toPretty { allowPrettyValues = true; }
(modify depth snip x)) y;
/* A combination of `traceVal` and `traceSeq` that applies a
provided function to the value to be traced after `deepSeq`ing
it.
*/
traceValSeqFn =
# Function to apply
f:
# Value to trace
v: traceValFn f (builtins.deepSeq v v);
/* A combination of `traceVal` and `traceSeq`. */
traceValSeq = traceValSeqFn id;
/* A combination of `traceVal` and `traceSeqN` that applies a
provided function to the value to be traced. */
traceValSeqNFn =
# Function to apply
f:
depth:
# Value to trace
v: traceSeqN depth (f v) v;
/* A combination of `traceVal` and `traceSeqN`. */
traceValSeqN = traceValSeqNFn id;
/* Trace the input and output of a function `f` named `name`,
both down to `depth`.
This is useful for adding around a function call,
to see the before/after of values as they are transformed.
Example:
traceFnSeqN 2 "id" (x: x) { a.b.c = 3; }
trace: { fn = "id"; from = { a.b = {}; }; to = { a.b = {}; }; }
=> { a.b.c = 3; }
*/
traceFnSeqN = depth: name: f: v:
let res = f v;
in lib.traceSeqN
(depth + 1)
{
fn = name;
from = v;
to = res;
}
res;
# -- TESTING --
/* Evaluates a set of tests.
A test is an attribute set `{expr, expected}`,
denoting an expression and its expected result.
The result is a `list` of __failed tests__, each represented as
`{name, expected, result}`,
- expected
- What was passed as `expected`
- result
- The actual `result` of the test
Used for regression testing of the functions in lib; see
tests.nix for more examples.
Important: Only attributes that start with `test` are executed.
- If you want to run only a subset of the tests add the attribute `tests = ["testName"];`
Example:
runTests {
testAndOk = {
expr = lib.and true false;
expected = false;
};
testAndFail = {
expr = lib.and true false;
expected = true;
};
}
->
[
{
name = "testAndFail";
expected = true;
result = false;
}
]
Type:
runTests :: {
tests = [ String ];
${testName} :: {
expr :: a;
expected :: a;
};
}
->
[
{
name :: String;
expected :: a;
result :: a;
}
]
*/
runTests =
# Tests to run
tests: concatLists (attrValues (mapAttrs (name: test:
let testsToRun = if tests ? tests then tests.tests else [];
in if (substring 0 4 name == "test" || elem name testsToRun)
&& ((testsToRun == []) || elem name tests.tests)
&& (test.expr != test.expected)
then [ { inherit name; expected = test.expected; result = test.expr; } ]
else [] ) tests));
/* Create a test assuming that list elements are `true`.
Example:
{ testX = allTrue [ true ]; }
*/
testAllTrue = expr: { inherit expr; expected = map (x: true) expr; };
}

171
lib/default.nix Normal file
View file

@ -0,0 +1,171 @@
/* Library of low-level helper functions for nix expressions.
*
* Please implement (mostly) exhaustive unit tests
* for new functions in `./tests.nix`.
*/
let
inherit (import ./fixed-points.nix { inherit lib; }) makeExtensible;
lib = makeExtensible (self: let
callLibs = file: import file { lib = self; };
in {
# often used, or depending on very little
trivial = callLibs ./trivial.nix;
fixedPoints = callLibs ./fixed-points.nix;
# datatypes
attrsets = callLibs ./attrsets.nix;
lists = callLibs ./lists.nix;
strings = callLibs ./strings.nix;
stringsWithDeps = callLibs ./strings-with-deps.nix;
# packaging
customisation = callLibs ./customisation.nix;
derivations = callLibs ./derivations.nix;
maintainers = import ../maintainers/maintainer-list.nix;
teams = callLibs ../maintainers/team-list.nix;
meta = callLibs ./meta.nix;
versions = callLibs ./versions.nix;
# module system
modules = callLibs ./modules.nix;
options = callLibs ./options.nix;
types = callLibs ./types.nix;
# constants
licenses = callLibs ./licenses.nix;
sourceTypes = callLibs ./source-types.nix;
systems = callLibs ./systems;
# serialization
cli = callLibs ./cli.nix;
gvariant = callLibs ./gvariant.nix;
generators = callLibs ./generators.nix;
# misc
asserts = callLibs ./asserts.nix;
debug = callLibs ./debug.nix;
misc = callLibs ./deprecated/misc.nix;
# domain-specific
fetchers = callLibs ./fetchers.nix;
# Eval-time filesystem handling
path = callLibs ./path;
filesystem = callLibs ./filesystem.nix;
fileset = callLibs ./fileset;
sources = callLibs ./sources.nix;
# back-compat aliases
platforms = self.systems.doubles;
# linux kernel configuration
kernel = callLibs ./kernel.nix;
inherit (builtins) add addErrorContext attrNames concatLists
deepSeq elem elemAt filter genericClosure genList getAttr
hasAttr head isAttrs isBool isInt isList isPath isString length
lessThan listToAttrs pathExists readFile replaceStrings seq
stringLength sub substring tail trace;
inherit (self.trivial) id const pipe concat or and xor bitAnd bitOr bitXor
bitNot boolToString mergeAttrs flip mapNullable inNixShell isFloat min max
importJSON importTOML warn warnIf warnIfNot throwIf throwIfNot checkListOfEnum
info showWarnings nixpkgsVersion version isInOldestRelease
mod compare splitByAndCompare
functionArgs setFunctionArgs isFunction toFunction mirrorFunctionArgs
toHexString toBaseDigits inPureEvalMode;
inherit (self.fixedPoints) fix fix' converge extends composeExtensions
composeManyExtensions makeExtensible makeExtensibleWithCustomName;
inherit (self.attrsets) attrByPath hasAttrByPath setAttrByPath
getAttrFromPath attrVals attrValues getAttrs catAttrs filterAttrs
filterAttrsRecursive foldlAttrs foldAttrs collect nameValuePair mapAttrs
mapAttrs' mapAttrsToList attrsToList concatMapAttrs mapAttrsRecursive
mapAttrsRecursiveCond genAttrs isDerivation toDerivation optionalAttrs
zipAttrsWithNames zipAttrsWith zipAttrs recursiveUpdateUntil
recursiveUpdate matchAttrs mergeAttrsList overrideExisting showAttrPath getOutput
getBin getLib getDev getMan chooseDevOutputs zipWithNames zip
recurseIntoAttrs dontRecurseIntoAttrs cartesianProduct cartesianProductOfSets
mapCartesianProduct updateManyAttrsByPath;
inherit (self.lists) singleton forEach foldr fold foldl foldl' imap0 imap1
ifilter0 concatMap flatten remove findSingle findFirst any all count
optional optionals toList range replicate partition zipListsWith zipLists
reverseList listDfs toposort sort sortOn naturalSort compareLists take
drop sublist last init crossLists unique allUnique intersectLists
subtractLists mutuallyExclusive groupBy groupBy';
inherit (self.strings) concatStrings concatMapStrings concatImapStrings
intersperse concatStringsSep concatMapStringsSep
concatImapStringsSep concatLines makeSearchPath makeSearchPathOutput
makeLibraryPath makeIncludePath makeBinPath optionalString
hasInfix hasPrefix hasSuffix stringToCharacters stringAsChars escape
escapeShellArg escapeShellArgs
isStorePath isStringLike
isValidPosixName toShellVar toShellVars
escapeRegex escapeURL escapeXML replaceChars lowerChars
upperChars toLower toUpper addContextFrom splitString
removePrefix removeSuffix versionOlder versionAtLeast
getName getVersion
cmakeOptionType cmakeBool cmakeFeature
mesonOption mesonBool mesonEnable
nameFromURL enableFeature enableFeatureAs withFeature
withFeatureAs fixedWidthString fixedWidthNumber
toInt toIntBase10 readPathsFromFile fileContents;
inherit (self.stringsWithDeps) textClosureList textClosureMap
noDepEntry fullDepEntry packEntry stringAfter;
inherit (self.customisation) overrideDerivation makeOverridable
callPackageWith callPackagesWith extendDerivation hydraJob
makeScope makeScopeWithSplicing makeScopeWithSplicing';
inherit (self.derivations) lazyDerivation optionalDrvAttr;
inherit (self.meta) addMetaAttrs dontDistribute setName updateName
appendToName mapDerivationAttrset setPrio lowPrio lowPrioSet hiPrio
hiPrioSet getLicenseFromSpdxId getExe getExe';
inherit (self.filesystem) pathType pathIsDirectory pathIsRegularFile
packagesFromDirectoryRecursive;
inherit (self.sources) cleanSourceFilter
cleanSource sourceByRegex sourceFilesBySuffices
commitIdFromGitRepo cleanSourceWith pathHasContext
canCleanSource pathIsGitRepo;
inherit (self.modules) evalModules setDefaultModuleLocation
unifyModuleSyntax applyModuleArgsIfFunction mergeModules
mergeModules' mergeOptionDecls mergeDefinitions
pushDownProperties dischargeProperties filterOverrides
sortProperties fixupOptionType mkIf mkAssert mkMerge mkOverride
mkOptionDefault mkDefault mkImageMediaOverride mkForce mkVMOverride
mkFixStrictness mkOrder mkBefore mkAfter mkAliasDefinitions
mkAliasAndWrapDefinitions fixMergeModules mkRemovedOptionModule
mkRenamedOptionModule mkRenamedOptionModuleWith
mkMergedOptionModule mkChangedOptionModule
mkAliasOptionModule mkDerivedConfig doRename
mkAliasOptionModuleMD;
evalOptionValue = lib.warn "External use of `lib.evalOptionValue` is deprecated. If your use case isn't covered by non-deprecated functions, we'd like to know more and perhaps support your use case well, instead of providing access to these low level functions. In this case please open an issue in https://github.com/nixos/nixpkgs/issues/." self.modules.evalOptionValue;
inherit (self.options) isOption mkEnableOption mkSinkUndeclaredOptions
mergeDefaultOption mergeOneOption mergeEqualOption mergeUniqueOption
getValues getFiles
optionAttrSetToDocList optionAttrSetToDocList'
scrubOptionValue literalExpression literalExample
showOption showOptionWithDefLocs showFiles
unknownModule mkOption mkPackageOption mkPackageOptionMD
mdDoc literalMD;
inherit (self.types) isType setType defaultTypeMerge defaultFunctor
isOptionType mkOptionType;
inherit (self.asserts)
assertMsg assertOneOf;
inherit (self.debug) traceIf traceVal traceValFn
traceSeq traceSeqN traceValSeq
traceValSeqFn traceValSeqN traceValSeqNFn traceFnSeqN
runTests testAllTrue;
inherit (self.misc) maybeEnv defaultMergeArg defaultMerge foldArgs
maybeAttrNullable maybeAttr ifEnable checkFlag getValue
checkReqs uniqList uniqListExt condConcat lazyGenericClosure
innerModifySumArgs modifySumArgs innerClosePropagation
closePropagation mapAttrsFlatten nvs setAttr setAttrMerge
mergeAttrsWithFunc mergeAttrsConcatenateValues
mergeAttrsNoOverride mergeAttrByFunc mergeAttrsByFuncDefaults
mergeAttrsByFuncDefaultsClean mergeAttrBy
fakeHash fakeSha256 fakeSha512
nixType imap;
inherit (self.versions)
splitVersion;
});
in lib

11
lib/deprecated/README.md Normal file
View file

@ -0,0 +1,11 @@
# lib/deprecated
Do not add any new functions to this directory.
This directory contains the `lib.misc` sublibrary, which - as a location - is deprecated.
Furthermore, some of the functions inside are of *dubious* utility, and should perhaps be avoided,
while some functions *may still be needed*.
This directory does not play a role in the deprecation process for library functions.
They should be deprecated in place, by putting a `lib.warn` or `lib.warnIf` call around the function.

374
lib/deprecated/misc.nix Normal file
View file

@ -0,0 +1,374 @@
{ lib }:
let
inherit (lib)
and
any
attrByPath
attrNames
compare
concat
concatMap
elem
filter
foldl
foldr
genericClosure
head
imap1
init
isAttrs
isFunction
isInt
isList
lists
listToAttrs
mapAttrs
mergeAttrs
meta
nameValuePair
tail
toList
;
inherit (lib.attrsets) removeAttrs;
# returns default if env var is not set
maybeEnv = name: default:
let value = builtins.getEnv name; in
if value == "" then default else value;
defaultMergeArg = x : y: if builtins.isAttrs y then
y
else
(y x);
defaultMerge = x: y: x // (defaultMergeArg x y);
foldArgs = merger: f: init: x:
let arg = (merger init (defaultMergeArg init x));
# now add the function with composed args already applied to the final attrs
base = (setAttrMerge "passthru" {} (f arg)
( z: z // {
function = foldArgs merger f arg;
args = (attrByPath ["passthru" "args"] {} z) // x;
} ));
withStdOverrides = base // {
override = base.passthru.function;
};
in
withStdOverrides;
# shortcut for attrByPath ["name"] default attrs
maybeAttrNullable = maybeAttr;
# shortcut for attrByPath ["name"] default attrs
maybeAttr = name: default: attrs: attrs.${name} or default;
# Return the second argument if the first one is true or the empty version
# of the second argument.
ifEnable = cond: val:
if cond then val
else if builtins.isList val then []
else if builtins.isAttrs val then {}
# else if builtins.isString val then ""
else if val == true || val == false then false
else null;
# Return true only if there is an attribute and it is true.
checkFlag = attrSet: name:
if name == "true" then true else
if name == "false" then false else
if (elem name (attrByPath ["flags"] [] attrSet)) then true else
attrByPath [name] false attrSet ;
# Input : attrSet, [ [name default] ... ], name
# Output : its value or default.
getValue = attrSet: argList: name:
( attrByPath [name] (if checkFlag attrSet name then true else
if argList == [] then null else
let x = builtins.head argList; in
if (head x) == name then
(head (tail x))
else (getValue attrSet
(tail argList) name)) attrSet );
# Input : attrSet, [[name default] ...], [ [flagname reqs..] ... ]
# Output : are reqs satisfied? It's asserted.
checkReqs = attrSet: argList: condList:
(
foldr and true
(map (x: let name = (head x); in
((checkFlag attrSet name) ->
(foldr and true
(map (y: let val=(getValue attrSet argList y); in
(val!=null) && (val!=false))
(tail x))))) condList));
# This function has O(n^2) performance.
uniqList = { inputList, acc ? [] }:
let go = xs: acc:
if xs == []
then []
else let x = head xs;
y = if elem x acc then [] else [x];
in y ++ go (tail xs) (y ++ acc);
in go inputList acc;
uniqListExt = { inputList,
outputList ? [],
getter ? (x: x),
compare ? (x: y: x==y) }:
if inputList == [] then outputList else
let x = head inputList;
isX = y: (compare (getter y) (getter x));
newOutputList = outputList ++
(if any isX outputList then [] else [x]);
in uniqListExt { outputList = newOutputList;
inputList = (tail inputList);
inherit getter compare;
};
condConcat = name: list: checker:
if list == [] then name else
if checker (head list) then
condConcat
(name + (head (tail list)))
(tail (tail list))
checker
else condConcat
name (tail (tail list)) checker;
lazyGenericClosure = {startSet, operator}:
let
work = list: doneKeys: result:
if list == [] then
result
else
let x = head list; key = x.key; in
if elem key doneKeys then
work (tail list) doneKeys result
else
work (tail list ++ operator x) ([key] ++ doneKeys) ([x] ++ result);
in
work startSet [] [];
innerModifySumArgs = f: x: a: b: if b == null then (f a b) // x else
innerModifySumArgs f x (a // b);
modifySumArgs = f: x: innerModifySumArgs f x {};
innerClosePropagation = acc: xs:
if xs == []
then acc
else let y = head xs;
ys = tail xs;
in if ! isAttrs y
then innerClosePropagation acc ys
else let acc' = [y] ++ acc;
in innerClosePropagation
acc'
(uniqList { inputList = (maybeAttrNullable "propagatedBuildInputs" [] y)
++ (maybeAttrNullable "propagatedNativeBuildInputs" [] y)
++ ys;
acc = acc';
}
);
closePropagationSlow = list: (uniqList {inputList = (innerClosePropagation [] list);});
# This is an optimisation of closePropagation which avoids the O(n^2) behavior
# Using a list of derivations, it generates the full closure of the propagatedXXXBuildInputs
# The ordering / sorting / comparison is done based on the `outPath`
# attribute of each derivation.
# On some benchmarks, it performs up to 15 times faster than closePropagation.
# See https://github.com/NixOS/nixpkgs/pull/194391 for details.
closePropagationFast = list:
builtins.map (x: x.val) (builtins.genericClosure {
startSet = builtins.map (x: {
key = x.outPath;
val = x;
}) (builtins.filter (x: x != null) list);
operator = item:
if !builtins.isAttrs item.val then
[ ]
else
builtins.concatMap (x:
if x != null then [{
key = x.outPath;
val = x;
}] else
[ ]) ((item.val.propagatedBuildInputs or [ ])
++ (item.val.propagatedNativeBuildInputs or [ ]));
});
closePropagation = if builtins ? genericClosure
then closePropagationFast
else closePropagationSlow;
# calls a function (f attr value ) for each record item. returns a list
mapAttrsFlatten = f: r: map (attr: f attr r.${attr}) (attrNames r);
# attribute set containing one attribute
nvs = name: value: listToAttrs [ (nameValuePair name value) ];
# adds / replaces an attribute of an attribute set
setAttr = set: name: v: set // (nvs name v);
# setAttrMerge (similar to mergeAttrsWithFunc but only merges the values of a particular name)
# setAttrMerge "a" [] { a = [2];} (x: x ++ [3]) -> { a = [2 3]; }
# setAttrMerge "a" [] { } (x: x ++ [3]) -> { a = [ 3]; }
setAttrMerge = name: default: attrs: f:
setAttr attrs name (f (maybeAttr name default attrs));
# Using f = a: b = b the result is similar to //
# merge attributes with custom function handling the case that the attribute
# exists in both sets
mergeAttrsWithFunc = f: set1: set2:
foldr (n: set: if set ? ${n}
then setAttr set n (f set.${n} set2.${n})
else set )
(set2 // set1) (attrNames set2);
# merging two attribute set concatenating the values of same attribute names
# eg { a = 7; } { a = [ 2 3 ]; } becomes { a = [ 7 2 3 ]; }
mergeAttrsConcatenateValues = mergeAttrsWithFunc ( a: b: (toList a) ++ (toList b) );
# merges attributes using //, if a name exists in both attributes
# an error will be triggered unless its listed in mergeLists
# so you can mergeAttrsNoOverride { buildInputs = [a]; } { buildInputs = [a]; } {} to get
# { buildInputs = [a b]; }
# merging buildPhase doesn't really make sense. The cases will be rare where appending /prefixing will fit your needs?
# in these cases the first buildPhase will override the second one
# ! deprecated, use mergeAttrByFunc instead
mergeAttrsNoOverride = { mergeLists ? ["buildInputs" "propagatedBuildInputs"],
overrideSnd ? [ "buildPhase" ]
}: attrs1: attrs2:
foldr (n: set:
setAttr set n ( if set ? ${n}
then # merge
if elem n mergeLists # attribute contains list, merge them by concatenating
then attrs2.${n} ++ attrs1.${n}
else if elem n overrideSnd
then attrs1.${n}
else throw "error mergeAttrsNoOverride, attribute ${n} given in both attributes - no merge func defined"
else attrs2.${n} # add attribute not existing in attr1
)) attrs1 (attrNames attrs2);
# example usage:
# mergeAttrByFunc {
# inherit mergeAttrBy; # defined below
# buildInputs = [ a b ];
# } {
# buildInputs = [ c d ];
# };
# will result in
# { mergeAttrsBy = [...]; buildInputs = [ a b c d ]; }
# is used by defaultOverridableDelayableArgs and can be used when composing using
# foldArgs, composedArgsAndFun or applyAndFun. Example: composableDerivation in all-packages.nix
mergeAttrByFunc = x: y:
let
mergeAttrBy2 = { mergeAttrBy = mergeAttrs; }
// (maybeAttr "mergeAttrBy" {} x)
// (maybeAttr "mergeAttrBy" {} y); in
foldr mergeAttrs {} [
x y
(mapAttrs ( a: v: # merge special names using given functions
if x ? ${a}
then if y ? ${a}
then v x.${a} y.${a} # both have attr, use merge func
else x.${a} # only x has attr
else y.${a} # only y has attr)
) (removeAttrs mergeAttrBy2
# don't merge attrs which are neither in x nor y
(filter (a: ! x ? ${a} && ! y ? ${a})
(attrNames mergeAttrBy2))
)
)
];
mergeAttrsByFuncDefaults = foldl mergeAttrByFunc { inherit mergeAttrBy; };
mergeAttrsByFuncDefaultsClean = list: removeAttrs (mergeAttrsByFuncDefaults list) ["mergeAttrBy"];
# sane defaults (same name as attr name so that inherit can be used)
mergeAttrBy = # { buildInputs = concatList; [...]; passthru = mergeAttr; [..]; }
listToAttrs (map (n: nameValuePair n concat)
[ "nativeBuildInputs" "buildInputs" "propagatedBuildInputs" "configureFlags" "prePhases" "postAll" "patches" ])
// listToAttrs (map (n: nameValuePair n mergeAttrs) [ "passthru" "meta" "cfg" "flags" ])
// listToAttrs (map (n: nameValuePair n (a: b: "${a}\n${b}") ) [ "preConfigure" "postInstall" ])
;
nixType = x:
if isAttrs x then
if x ? outPath then "derivation"
else "attrs"
else if isFunction x then "function"
else if isList x then "list"
else if x == true then "bool"
else if x == false then "bool"
else if x == null then "null"
else if isInt x then "int"
else "string";
/**
# Deprecated
For historical reasons, imap has an index starting at 1.
But for consistency with the rest of the library we want an index
starting at zero.
*/
imap = imap1;
# Fake hashes. Can be used as hash placeholders, when computing hash ahead isn't trivial
fakeHash = "sha256-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=";
fakeSha256 = "0000000000000000000000000000000000000000000000000000000000000000";
fakeSha512 = "00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000";
in
# Everything in this attrset is the public interface of the file.
{
inherit
checkFlag
checkReqs
closePropagation
closePropagationFast
closePropagationSlow
condConcat
defaultMerge
defaultMergeArg
fakeHash
fakeSha256
fakeSha512
foldArgs
getValue
ifEnable
imap
innerClosePropagation
innerModifySumArgs
lazyGenericClosure
mapAttrsFlatten
maybeAttr
maybeAttrNullable
maybeEnv
mergeAttrBy
mergeAttrByFunc
mergeAttrsByFuncDefaults
mergeAttrsByFuncDefaultsClean
mergeAttrsConcatenateValues
mergeAttrsNoOverride
mergeAttrsWithFunc
modifySumArgs
nixType
nvs
setAttr
setAttrMerge
uniqList
uniqListExt
;
}

177
lib/derivations.nix Normal file
View file

@ -0,0 +1,177 @@
{ lib }:
let
inherit (lib)
genAttrs
isString
throwIfNot
;
showMaybeAttrPosPre = prefix: attrName: v:
let pos = builtins.unsafeGetAttrPos attrName v;
in if pos == null then "" else "${prefix}${pos.file}:${toString pos.line}:${toString pos.column}";
showMaybePackagePosPre = prefix: pkg:
if pkg?meta.position && isString pkg.meta.position
then "${prefix}${pkg.meta.position}"
else "";
in
{
/*
Restrict a derivation to a predictable set of attribute names, so
that the returned attrset is not strict in the actual derivation,
saving a lot of computation when the derivation is non-trivial.
This is useful in situations where a derivation might only be used for its
passthru attributes, improving evaluation performance.
The returned attribute set is lazy in `derivation`. Specifically, this
means that the derivation will not be evaluated in at least the
situations below.
For illustration and/or testing, we define derivation such that its
evaluation is very noticeable.
let derivation = throw "This won't be evaluated.";
In the following expressions, `derivation` will _not_ be evaluated:
(lazyDerivation { inherit derivation; }).type
attrNames (lazyDerivation { inherit derivation; })
(lazyDerivation { inherit derivation; } // { foo = true; }).foo
(lazyDerivation { inherit derivation; meta.foo = true; }).meta
In these expressions, `derivation` _will_ be evaluated:
"${lazyDerivation { inherit derivation }}"
(lazyDerivation { inherit derivation }).outPath
(lazyDerivation { inherit derivation }).meta
And the following expressions are not valid, because the refer to
implementation details and/or attributes that may not be present on
some derivations:
(lazyDerivation { inherit derivation }).buildInputs
(lazyDerivation { inherit derivation }).passthru
(lazyDerivation { inherit derivation }).pythonPath
*/
lazyDerivation =
args@{
# The derivation to be wrapped.
derivation
, # Optional meta attribute.
#
# While this function is primarily about derivations, it can improve
# the `meta` package attribute, which is usually specified through
# `mkDerivation`.
meta ? null
, # Optional extra values to add to the returned attrset.
#
# This can be used for adding package attributes, such as `tests`.
passthru ? { }
, # Optional list of assumed outputs. Default: ["out"]
#
# This must match the set of outputs that the returned derivation has.
# You must use this when the derivation has multiple outputs.
outputs ? [ "out" ]
}:
let
# These checks are strict in `drv` and some `drv` attributes, but the
# attrset spine returned by lazyDerivation does not depend on it.
# Instead, the individual derivation attributes do depend on it.
checked =
throwIfNot (derivation.type or null == "derivation")
"lazyDerivation: input must be a derivation."
throwIfNot
# NOTE: Technically we could require our outputs to be a subset of the
# actual ones, or even leave them unchecked and fail on a lazy basis.
# However, consider the case where an output is added in the underlying
# derivation, such as dev. lazyDerivation would remove it and cause it
# to fail as a buildInputs item, without any indication as to what
# happened. Hence the more stringent condition. We could consider
# adding a flag to control this behavior if there's a valid case for it,
# but the documentation must have a note like this.
(derivation.outputs == outputs)
''
lib.lazyDerivation: The derivation ${derivation.name or "<unknown>"} has outputs that don't match the assumed outputs.
Assumed outputs passed to lazyDerivation${showMaybeAttrPosPre ",\n at " "outputs" args}:
${lib.generators.toPretty { multiline = false; } outputs};
Actual outputs of the derivation${showMaybePackagePosPre ",\n defined at " derivation}:
${lib.generators.toPretty { multiline = false; } derivation.outputs}
If the outputs are known ahead of evaluating the derivation,
then update the lazyDerivation call to match the actual outputs, in the same order.
If lazyDerivation is passed a literal value, just change it to the actual outputs.
As a result it will work as before / as intended.
Otherwise, when the outputs are dynamic and can't be known ahead of time, it won't
be possible to add laziness, but lib.lazyDerivation may still be useful for trimming
the attributes.
If you want to keep trimming the attributes, make sure that the package is in a
variable (don't evaluate it twice!) and pass the variable and its outputs attribute
to lib.lazyDerivation. This largely defeats laziness, but keeps the trimming.
If none of the above works for you, replace the lib.lazyDerivation call by the
expression in the derivation argument.
''
derivation;
in
{
# Hardcoded `type`
#
# `lazyDerivation` requires its `derivation` argument to be a derivation,
# so if it is not, that is a programming error by the caller and not
# something that `lazyDerivation` consumers should be able to correct
# for after the fact.
# So, to improve laziness, we assume correctness here and check it only
# when actual derivation values are accessed later.
type = "derivation";
# A fixed set of derivation values, so that `lazyDerivation` can return
# its attrset before evaluating `derivation`.
# This must only list attributes that are available on _all_ derivations.
inherit (checked) outPath outputName drvPath name system;
inherit outputs;
# The meta attribute can either be taken from the derivation, or if the
# `lazyDerivation` caller knew a shortcut, be taken from there.
meta = args.meta or checked.meta;
}
// genAttrs outputs (outputName: checked.${outputName})
// passthru;
/* Conditionally set a derivation attribute.
Because `mkDerivation` sets `__ignoreNulls = true`, a derivation
attribute set to `null` will not impact the derivation output hash.
Thus, this function passes through its `value` argument if the `cond`
is `true`, but returns `null` if not.
Type: optionalDrvAttr :: Bool -> a -> a | Null
Example:
(stdenv.mkDerivation {
name = "foo";
x = optionalDrvAttr true 1;
y = optionalDrvAttr false 1;
}).drvPath == (stdenv.mkDerivation {
name = "foo";
x = 1;
}).drvPath
=> true
*/
optionalDrvAttr =
# Condition
cond:
# Attribute value
value: if cond then value else null;
}

13
lib/fetchers.nix Normal file
View file

@ -0,0 +1,13 @@
# snippets that can be shared by multiple fetchers (pkgs/build-support)
{ lib }:
{
proxyImpureEnvVars = [
# We borrow these environment variables from the caller to allow
# easy proxy configuration. This is impure, but a fixed-output
# derivation like fetchurl is allowed to do so since its result is
# by definition pure.
"http_proxy" "https_proxy" "ftp_proxy" "all_proxy" "no_proxy"
];
}

267
lib/fileset/README.md Normal file
View file

@ -0,0 +1,267 @@
# File set library
This is the internal contributor documentation.
The user documentation is [in the Nixpkgs manual](https://nixos.org/manual/nixpkgs/unstable/#sec-fileset).
## Goals
The main goal of the file set library is to be able to select local files that should be added to the Nix store.
It should have the following properties:
- Easy:
The functions should have obvious semantics, be low in number and be composable.
- Safe:
Throw early and helpful errors when mistakes are detected.
- Lazy:
Only compute values when necessary.
Non-goals are:
- Efficient:
If the abstraction proves itself worthwhile but too slow, it can be still be optimized further.
## Tests
Tests are declared in [`tests.sh`](./tests.sh) and can be run using
```
./tests.sh
```
## Benchmark
A simple benchmark against the HEAD commit can be run using
```
./benchmark.sh HEAD
```
This is intended to be run manually and is not checked by CI.
## Internal representation
The internal representation is versioned in order to allow file sets from different Nixpkgs versions to be composed with each other, see [`internal.nix`](./internal.nix) for the versions and conversions between them.
This section describes only the current representation, but past versions will have to be supported by the code.
### `fileset`
An attribute set with these values:
- `_type` (constant string `"fileset"`):
Tag to indicate this value is a file set.
- `_internalVersion` (constant `3`, the current version):
Version of the representation.
- `_internalIsEmptyWithoutBase` (bool):
Whether this file set is the empty file set without a base path.
If `true`, `_internalBase*` and `_internalTree` are not set.
This is the only way to represent an empty file set without needing a base path.
Such a value can be used as the identity element for `union` and the return value of `unions []` and co.
- `_internalBase` (path):
Any files outside of this path cannot influence the set of files.
This is always a directory and should be as long as possible.
This is used by `lib.fileset.toSource` to check that all files are under the `root` argument
- `_internalBaseRoot` (path):
The filesystem root of `_internalBase`, same as `(lib.path.splitRoot _internalBase).root`.
This is here because this needs to be computed anyway, and this computation shouldn't be duplicated.
- `_internalBaseComponents` (list of strings):
The path components of `_internalBase`, same as `lib.path.subpath.components (lib.path.splitRoot _internalBase).subpath`.
This is here because this needs to be computed anyway, and this computation shouldn't be duplicated.
- `_internalTree` ([filesetTree](#filesettree)):
A tree representation of all included files under `_internalBase`.
- `__noEval` (error):
An error indicating that directly evaluating file sets is not supported.
## `filesetTree`
One of the following:
- `{ <name> = filesetTree; }`:
A directory with a nested `filesetTree` value for directory entries.
Entries not included may either be omitted or set to `null`, as necessary to improve efficiency or laziness.
- `"directory"`:
A directory with all its files included recursively, allowing early cutoff for some operations.
This specific string is chosen to be compatible with `builtins.readDir` for a simpler implementation.
- `"regular"`, `"symlink"`, `"unknown"` or any other non-`"directory"` string:
A nested file with its file type.
These specific strings are chosen to be compatible with `builtins.readDir` for a simpler implementation.
Distinguishing between different file types is not strictly necessary for the functionality this library,
but it does allow nicer printing of file sets.
- `null`:
A file or directory that is excluded from the tree.
It may still exist on the file system.
## API design decisions
This section justifies API design decisions.
### Internal structure
The representation of the file set data type is internal and can be changed over time.
Arguments:
- (+) The point of this library is to provide high-level functions, users don't need to be concerned with how it's implemented
- (+) It allows adjustments to the representation, which is especially useful in the early days of the library.
- (+) It still allows the representation to be stabilized later if necessary and if it has proven itself
### Influence tracking
File set operations internally track the top-most directory that could influence the exact contents of a file set.
Specifically, `toSource` requires that the given `fileset` is completely determined by files within the directory specified by the `root` argument.
For example, even with `dir/file.txt` being the only file in `./.`, `toSource { root = ./dir; fileset = ./.; }` gives an error.
This is because `fileset` may as well be the result of filtering `./.` in a way that excludes `dir`.
Arguments:
- (+) This gives us the guarantee that adding new files to a project never breaks a file set expression.
This is also true in a lesser form for removed files:
only removing files explicitly referenced by paths can break a file set expression.
- (+) This can be removed later, if we discover it's too restrictive
- (-) It leads to errors when a sensible result could sometimes be returned, such as in the above example.
### Empty file set without a base
There is a special representation for an empty file set without a base path.
This is used for return values that should be empty but when there's no base path that would makes sense.
Arguments:
- Alternative: This could also be represented using `_internalBase = /.` and `_internalTree = null`.
- (+) Removes the need for a special representation.
- (-) Due to [influence tracking](#influence-tracking),
`union empty ./.` would have `/.` as the base path,
which would then prevent `toSource { root = ./.; fileset = union empty ./.; }` from working,
which is not as one would expect.
- (-) With the assumption that there can be multiple filesystem roots (as established with the [path library](../path/README.md)),
this would have to cause an error with `union empty pathWithAnotherFilesystemRoot`,
which is not as one would expect.
- Alternative: Do not have such a value and error when it would be needed as a return value
- (+) Removes the need for a special representation.
- (-) Leaves us with no identity element for `union` and no reasonable return value for `unions []`.
From a set theory perspective, which has a well-known notion of empty sets, this is unintuitive.
### No intersection for lists
While there is `intersection a b`, there is no function `intersections [ a b c ]`.
Arguments:
- (+) There is no known use case for such a function, it can be added later if a use case arises
- (+) There is no suitable return value for `intersections [ ]`, see also "Nullary intersections" [here](https://en.wikipedia.org/w/index.php?title=List_of_set_identities_and_relations&oldid=1177174035#Definitions)
- (-) Could throw an error for that case
- (-) Create a special value to represent "all the files" and return that
- (+) Such a value could then not be used with `fileFilter` unless the internal representation is changed considerably
- (-) Could return the empty file set
- (+) This would be wrong in set theory
- (-) Inconsistent with `union` and `unions`
### Intersection base path
The base path of the result of an `intersection` is the longest base path of the arguments.
E.g. the base path of `intersection ./foo ./foo/bar` is `./foo/bar`.
Meanwhile `intersection ./foo ./bar` returns the empty file set without a base path.
Arguments:
- Alternative: Use the common prefix of all base paths as the resulting base path
- (-) This is unnecessarily strict, because the purpose of the base path is to track the directory under which files _could_ be in the file set. It should be as long as possible.
All files contained in `intersection ./foo ./foo/bar` will be under `./foo/bar` (never just under `./foo`), and `intersection ./foo ./bar` will never contain any files (never under `./.`).
This would lead to `toSource` having to unexpectedly throw errors for cases such as `toSource { root = ./foo; fileset = intersect ./foo base; }`, where `base` may be `./bar` or `./.`.
- (-) There is no benefit to the user, since base path is not directly exposed in the interface
### Empty directories
File sets can only represent a _set_ of local files.
Directories on their own are not representable.
Arguments:
- (+) There does not seem to be a sensible set of combinators when directories can be represented on their own.
Here's some possibilities:
- `./.` represents the files in `./.` _and_ the directory itself including its subdirectories, meaning that even if there's no files, the entire structure of `./.` is preserved
In that case, what should `fileFilter (file: false) ./.` return?
It could return the entire directory structure unchanged, but with all files removed, which would not be what one would expect.
Trying to have a filter function that also supports directories will lead to the question of:
What should the behavior be if `./foo` itself is excluded but all of its contents are included?
It leads to having to define when directories are recursed into, but then we're effectively back at how the `builtins.path`-based filters work.
- `./.` represents all files in `./.` _and_ the directory itself, but not its subdirectories, meaning that at least `./.` will be preserved even if it's empty.
In that case, `intersection ./. ./foo` should only include files and no directories themselves, since `./.` includes only `./.` as a directory, and same for `./foo`, so there's no overlap in directories.
But intuitively this operation should result in the same as `./foo` everything else is just confusing.
- (+) This matches how Git only supports files, so developers should already be used to it.
- (-) Empty directories (even if they contain nested directories) are neither representable nor preserved when coercing from paths.
- (+) It is very rare that empty directories are necessary.
- (+) We can implement a workaround, allowing `toSource` to take an extra argument for ensuring certain extra directories exist in the result.
- (-) It slows down store imports, since the evaluator needs to traverse the entire tree to remove any empty directories
- (+) This can still be optimized by introducing more Nix builtins if necessary
### String paths
File sets do not support Nix store paths in strings such as `"/nix/store/...-source"`.
Arguments:
- (+) Such paths are usually produced by derivations, which means `toSource` would either:
- Require [Import From Derivation](https://nixos.org/manual/nix/unstable/language/import-from-derivation) (IFD) if `builtins.path` is used as the underlying primitive
- Require importing the entire `root` into the store such that derivations can be used to do the filtering
- (+) The convenient path coercion like `union ./foo ./bar` wouldn't work for absolute paths, requiring more verbose alternate interfaces:
- `let root = "/nix/store/...-source"; in union "${root}/foo" "${root}/bar"`
Verbose and dangerous because if `root` was a path, the entire path would get imported into the store.
- `toSource { root = "/nix/store/...-source"; fileset = union "./foo" "./bar"; }`
Does not allow debug printing intermediate file set contents, since we don't know the paths contents before having a `root`.
- `let fs = lib.fileset.withRoot "/nix/store/...-source"; in fs.union "./foo" "./bar"`
Makes library functions impure since they depend on the contextual root path, questionable composability.
- (+) The point of the file set abstraction is to specify which files should get imported into the store.
This use case makes little sense for files that are already in the store.
This should be a separate abstraction as e.g. `pkgs.drvLayout` instead, which could have a similar interface but be specific to derivations.
Additional capabilities could be supported that can't be done at evaluation time, such as renaming files, creating new directories, setting executable bits, etc.
- (+) An API for filtering/transforming Nix store paths could be much more powerful,
because it's not limited to just what is possible at evaluation time with `builtins.path`.
Operations such as moving and adding files would be supported.
### Single files
File sets cannot add single files to the store, they can only import files under directories.
Arguments:
- (+) There's no point in using this library for a single file, since you can't do anything other than add it to the store or not.
And it would be unclear how the library should behave if the one file wouldn't be added to the store:
`toSource { root = ./file.nix; fileset = <empty>; }` has no reasonable result because returing an empty store path wouldn't match the file type, and there's no way to have an empty file store path, whatever that would mean.
### `fileFilter` takes a path
The `fileFilter` function takes a path, and not a file set, as its second argument.
- (-) Makes it harder to compose functions, since the file set type, the return value, can't be passed to the function itself like `fileFilter predicate fileset`
- (+) It's still possible to use `intersection` to filter on file sets: `intersection fileset (fileFilter predicate ./.)`
- (-) This does need an extra `./.` argument that's not obvious
- (+) This could always be `/.` or the project directory, `intersection` will make it lazy
- (+) In the future this will allow `fileFilter` to support a predicate property like `subpath` and/or `components` in a reproducible way.
This wouldn't be possible if it took a file set, because file sets don't have a predictable absolute path.
- (-) What about the base path?
- (+) That can change depending on which files are included, so if it's used for `fileFilter`
it would change the `subpath`/`components` value depending on which files are included.
- (+) If necessary, this restriction can be relaxed later, the opposite wouldn't be possible
### Strict path existence checking
Coercing paths that don't exist to file sets always gives an error.
- (-) Sometimes you want to remove a file that may not always exist using `difference ./. ./does-not-exist`,
but this does not work because coercion of `./does-not-exist` fails,
even though its existence would have no influence on the result.
- (+) This is dangerous, because you wouldn't be protected against typos anymore.
E.g. when trying to prevent `./secret` from being imported, a typo like `difference ./. ./sercet` would import it regardless.
- (+) `difference ./. (maybeMissing ./does-not-exist)` can be used to do this more explicitly.
- (+) `difference ./. (difference ./foo ./foo/bar)` should report an error when `./foo/bar` does not exist ("double negation"). Unfortunately, the current internal representation does not lend itself to a behavior where both `difference x ./does-not-exists` and double negation are handled and checked correctly.
This could be fixed, but would require significant changes to the internal representation that are not worth the effort and the risk of introducing implicit behavior.

140
lib/fileset/benchmark.sh Executable file
View file

@ -0,0 +1,140 @@
#!/usr/bin/env nix-shell
#!nix-shell -i bash -p sta jq bc nix -I nixpkgs=../..
# shellcheck disable=SC2016
# Benchmarks lib.fileset
# Run:
# [nixpkgs]$ lib/fileset/benchmark.sh HEAD
set -euo pipefail
shopt -s inherit_errexit dotglob
if (( $# == 0 )); then
echo "Usage: $0 HEAD"
echo "Benchmarks the current tree against the HEAD commit. Any git ref will work."
exit 1
fi
compareTo=$1
SCRIPT_FILE=$(readlink -f "${BASH_SOURCE[0]}")
SCRIPT_DIR=$(dirname "$SCRIPT_FILE")
nixpkgs=$(cd "$SCRIPT_DIR/../.."; pwd)
tmp="$(mktemp -d)"
clean_up() {
rm -rf "$tmp"
}
trap clean_up EXIT SIGINT SIGTERM
work="$tmp/work"
mkdir "$work"
cd "$work"
declare -a stats=(
".envs.elements"
".envs.number"
".gc.totalBytes"
".list.concats"
".list.elements"
".nrFunctionCalls"
".nrLookups"
".nrOpUpdates"
".nrPrimOpCalls"
".nrThunks"
".sets.elements"
".sets.number"
".symbols.number"
".values.number"
)
runs=10
run() {
# Empty the file
: > cpuTimes
for i in $(seq 0 "$runs"); do
NIX_PATH=nixpkgs=$1 NIX_SHOW_STATS=1 NIX_SHOW_STATS_PATH=$tmp/stats.json \
nix-instantiate --eval --strict --show-trace >/dev/null \
--expr 'with import <nixpkgs/lib>; with fileset; '"$2"
# Only measure the time after the first run, one is warmup
if (( i > 0 )); then
jq '.cpuTime' "$tmp/stats.json" >> cpuTimes
fi
done
# Compute mean and standard deviation
read -r mean sd < <(sta --mean --sd --brief <cpuTimes)
jq --argjson mean "$mean" --argjson sd "$sd" \
'.cpuTimeMean = $mean | .cpuTimeSd = $sd' \
"$tmp/stats.json"
}
bench() {
echo "Benchmarking expression $1" >&2
#echo "Running benchmark on index" >&2
run "$nixpkgs" "$1" > "$tmp/new.json"
(
#echo "Checking out $compareTo" >&2
git -C "$nixpkgs" worktree add --quiet "$tmp/worktree" "$compareTo"
trap 'git -C "$nixpkgs" worktree remove "$tmp/worktree"' EXIT
#echo "Running benchmark on $compareTo" >&2
run "$tmp/worktree" "$1" > "$tmp/old.json"
)
read -r oldMean oldSd newMean newSd percentageMean percentageSd < \
<(jq -rn --slurpfile old "$tmp/old.json" --slurpfile new "$tmp/new.json" \
' $old[0].cpuTimeMean as $om
| $old[0].cpuTimeSd as $os
| $new[0].cpuTimeMean as $nm
| $new[0].cpuTimeSd as $ns
| (100 / $om * $nm) as $pm
# Copied from https://github.com/sharkdp/hyperfine/blob/b38d550b89b1dab85139eada01c91a60798db9cc/src/benchmark/relative_speed.rs#L46-L53
| ($pm * pow(pow($ns / $nm; 2) + pow($os / $om; 2); 0.5)) as $ps
| [ $om, $os, $nm, $ns, $pm, $ps ]
| @sh')
echo -e "Mean CPU time $newMean (σ = $newSd) for $runs runs is \e[0;33m$percentageMean% (σ = $percentageSd%)\e[0m of the old value $oldMean (σ = $oldSd)" >&2
different=0
for stat in "${stats[@]}"; do
oldValue=$(jq "$stat" "$tmp/old.json")
newValue=$(jq "$stat" "$tmp/new.json")
if (( oldValue != newValue )); then
percent=$(bc <<< "scale=100; result = 100/$oldValue*$newValue; scale=4; result / 1")
if (( oldValue < newValue )); then
echo -e "Statistic $stat ($newValue) is \e[0;31m$percent% (+$(( newValue - oldValue )))\e[0m of the old value $oldValue" >&2
else
echo -e "Statistic $stat ($newValue) is \e[0;32m$percent% (-$(( oldValue - newValue )))\e[0m of the old value $oldValue" >&2
fi
(( different++ )) || true
fi
done
echo "$different stats differ between the current tree and $compareTo"
echo ""
}
# Create a fairly populated tree
touch f{0..5}
mkdir d{0..5}
mkdir e{0..5}
touch d{0..5}/f{0..5}
mkdir -p d{0..5}/d{0..5}
mkdir -p e{0..5}/e{0..5}
touch d{0..5}/d{0..5}/f{0..5}
mkdir -p d{0..5}/d{0..5}/d{0..5}
mkdir -p e{0..5}/e{0..5}/e{0..5}
touch d{0..5}/d{0..5}/d{0..5}/f{0..5}
mkdir -p d{0..5}/d{0..5}/d{0..5}/d{0..5}
mkdir -p e{0..5}/e{0..5}/e{0..5}/e{0..5}
touch d{0..5}/d{0..5}/d{0..5}/d{0..5}/f{0..5}
bench 'toSource { root = ./.; fileset = ./.; }'
rm -rf -- *
touch {0..1000}
bench 'toSource { root = ./.; fileset = unions (mapAttrsToList (name: value: ./. + "/${name}") (builtins.readDir ./.)); }'
rm -rf -- *

860
lib/fileset/default.nix Normal file
View file

@ -0,0 +1,860 @@
/*
<!-- This anchor is here for backwards compatibility -->
[]{#sec-fileset}
The [`lib.fileset`](#sec-functions-library-fileset) library allows you to work with _file sets_.
A file set is a (mathematical) set of local files that can be added to the Nix store for use in Nix derivations.
File sets are easy and safe to use, providing obvious and composable semantics with good error messages to prevent mistakes.
## Overview {#sec-fileset-overview}
Basics:
- [Implicit coercion from paths to file sets](#sec-fileset-path-coercion)
- [`lib.fileset.maybeMissing`](#function-library-lib.fileset.maybeMissing):
Create a file set from a path that may be missing.
- [`lib.fileset.trace`](#function-library-lib.fileset.trace)/[`lib.fileset.traceVal`](#function-library-lib.fileset.trace):
Pretty-print file sets for debugging.
- [`lib.fileset.toSource`](#function-library-lib.fileset.toSource):
Add files in file sets to the store to use as derivation sources.
- [`lib.fileset.toList`](#function-library-lib.fileset.toList):
The list of files contained in a file set.
Combinators:
- [`lib.fileset.union`](#function-library-lib.fileset.union)/[`lib.fileset.unions`](#function-library-lib.fileset.unions):
Create a larger file set from all the files in multiple file sets.
- [`lib.fileset.intersection`](#function-library-lib.fileset.intersection):
Create a smaller file set from only the files in both file sets.
- [`lib.fileset.difference`](#function-library-lib.fileset.difference):
Create a smaller file set containing all files that are in one file set, but not another one.
Filtering:
- [`lib.fileset.fileFilter`](#function-library-lib.fileset.fileFilter):
Create a file set from all files that satisisfy a predicate in a directory.
Utilities:
- [`lib.fileset.fromSource`](#function-library-lib.fileset.fromSource):
Create a file set from a `lib.sources`-based value.
- [`lib.fileset.gitTracked`](#function-library-lib.fileset.gitTracked)/[`lib.fileset.gitTrackedWith`](#function-library-lib.fileset.gitTrackedWith):
Create a file set from all tracked files in a local Git repository.
If you need more file set functions,
see [this issue](https://github.com/NixOS/nixpkgs/issues/266356) to request it.
## Implicit coercion from paths to file sets {#sec-fileset-path-coercion}
All functions accepting file sets as arguments can also accept [paths](https://nixos.org/manual/nix/stable/language/values.html#type-path) as arguments.
Such path arguments are implicitly coerced to file sets containing all files under that path:
- A path to a file turns into a file set containing that single file.
- A path to a directory turns into a file set containing all files _recursively_ in that directory.
If the path points to a non-existent location, an error is thrown.
::: {.note}
Just like in Git, file sets cannot represent empty directories.
Because of this, a path to a directory that contains no files (recursively) will turn into a file set containing no files.
:::
:::{.note}
File set coercion does _not_ add any of the files under the coerced paths to the store.
Only the [`toSource`](#function-library-lib.fileset.toSource) function adds files to the Nix store, and only those files contained in the `fileset` argument.
This is in contrast to using [paths in string interpolation](https://nixos.org/manual/nix/stable/language/values.html#type-path), which does add the entire referenced path to the store.
:::
### Example {#sec-fileset-path-coercion-example}
Assume we are in a local directory with a file hierarchy like this:
```
a/
x (file)
b/
  y (file)
c/
   d/
```
Here's a listing of which files get included when different path expressions get coerced to file sets:
- `./.` as a file set contains both `a/x` and `a/b/y` (`c/` does not contain any files and is therefore omitted).
- `./a` as a file set contains both `a/x` and `a/b/y`.
- `./a/x` as a file set contains only `a/x`.
- `./a/b` as a file set contains only `a/b/y`.
- `./c` as a file set is empty, since neither `c` nor `c/d` contain any files.
*/
{ lib }:
let
inherit (import ./internal.nix { inherit lib; })
_coerce
_singleton
_coerceMany
_toSourceFilter
_fromSourceFilter
_toList
_unionMany
_fileFilter
_printFileset
_intersection
_difference
_fromFetchGit
_fetchGitSubmodulesMinver
_emptyWithoutBase
;
inherit (builtins)
isBool
isList
isPath
pathExists
seq
typeOf
nixVersion
;
inherit (lib.lists)
elemAt
imap0
;
inherit (lib.path)
hasPrefix
splitRoot
;
inherit (lib.strings)
isStringLike
versionOlder
;
inherit (lib.filesystem)
pathType
;
inherit (lib.sources)
cleanSourceWith
;
inherit (lib.trivial)
isFunction
pipe
;
in {
/*
Create a file set from a path that may or may not exist:
- If the path does exist, the path is [coerced to a file set](#sec-fileset-path-coercion).
- If the path does not exist, a file set containing no files is returned.
Type:
maybeMissing :: Path -> FileSet
Example:
# All files in the current directory, but excluding main.o if it exists
difference ./. (maybeMissing ./main.o)
*/
maybeMissing =
path:
if ! isPath path then
if isStringLike path then
throw ''
lib.fileset.maybeMissing: Argument ("${toString path}") is a string-like value, but it should be a path instead.''
else
throw ''
lib.fileset.maybeMissing: Argument is of type ${typeOf path}, but it should be a path instead.''
else if ! pathExists path then
_emptyWithoutBase
else
_singleton path;
/*
Incrementally evaluate and trace a file set in a pretty way.
This function is only intended for debugging purposes.
The exact tracing format is unspecified and may change.
This function takes a final argument to return.
In comparison, [`traceVal`](#function-library-lib.fileset.traceVal) returns
the given file set argument.
This variant is useful for tracing file sets in the Nix repl.
Type:
trace :: FileSet -> Any -> Any
Example:
trace (unions [ ./Makefile ./src ./tests/run.sh ]) null
=>
trace: /home/user/src/myProject
trace: - Makefile (regular)
trace: - src (all files in directory)
trace: - tests
trace: - run.sh (regular)
null
*/
trace =
/*
The file set to trace.
This argument can also be a path,
which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
*/
fileset:
let
# "fileset" would be a better name, but that would clash with the argument name,
# and we cannot change that because of https://github.com/nix-community/nixdoc/issues/76
actualFileset = _coerce "lib.fileset.trace: Argument" fileset;
in
seq
(_printFileset actualFileset)
(x: x);
/*
Incrementally evaluate and trace a file set in a pretty way.
This function is only intended for debugging purposes.
The exact tracing format is unspecified and may change.
This function returns the given file set.
In comparison, [`trace`](#function-library-lib.fileset.trace) takes another argument to return.
This variant is useful for tracing file sets passed as arguments to other functions.
Type:
traceVal :: FileSet -> FileSet
Example:
toSource {
root = ./.;
fileset = traceVal (unions [
./Makefile
./src
./tests/run.sh
]);
}
=>
trace: /home/user/src/myProject
trace: - Makefile (regular)
trace: - src (all files in directory)
trace: - tests
trace: - run.sh (regular)
"/nix/store/...-source"
*/
traceVal =
/*
The file set to trace and return.
This argument can also be a path,
which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
*/
fileset:
let
# "fileset" would be a better name, but that would clash with the argument name,
# and we cannot change that because of https://github.com/nix-community/nixdoc/issues/76
actualFileset = _coerce "lib.fileset.traceVal: Argument" fileset;
in
seq
(_printFileset actualFileset)
# We could also return the original fileset argument here,
# but that would then duplicate work for consumers of the fileset, because then they have to coerce it again
actualFileset;
/*
Add the local files contained in `fileset` to the store as a single [store path](https://nixos.org/manual/nix/stable/glossary#gloss-store-path) rooted at `root`.
The result is the store path as a string-like value, making it usable e.g. as the `src` of a derivation, or in string interpolation:
```nix
stdenv.mkDerivation {
src = lib.fileset.toSource { ... };
# ...
}
```
The name of the store path is always `source`.
Type:
toSource :: {
root :: Path,
fileset :: FileSet,
} -> SourceLike
Example:
# Import the current directory into the store
# but only include files under ./src
toSource {
root = ./.;
fileset = ./src;
}
=> "/nix/store/...-source"
# Import the current directory into the store
# but only include ./Makefile and all files under ./src
toSource {
root = ./.;
fileset = union
./Makefile
./src;
}
=> "/nix/store/...-source"
# Trying to include a file outside the root will fail
toSource {
root = ./.;
fileset = unions [
./Makefile
./src
../LICENSE
];
}
=> <error>
# The root needs to point to a directory that contains all the files
toSource {
root = ../.;
fileset = unions [
./Makefile
./src
../LICENSE
];
}
=> "/nix/store/...-source"
# The root has to be a local filesystem path
toSource {
root = "/nix/store/...-source";
fileset = ./.;
}
=> <error>
*/
toSource = {
/*
(required) The local directory [path](https://nixos.org/manual/nix/stable/language/values.html#type-path) that will correspond to the root of the resulting store path.
Paths in [strings](https://nixos.org/manual/nix/stable/language/values.html#type-string), including Nix store paths, cannot be passed as `root`.
`root` has to be a directory.
:::{.note}
Changing `root` only affects the directory structure of the resulting store path, it does not change which files are added to the store.
The only way to change which files get added to the store is by changing the `fileset` attribute.
:::
*/
root,
/*
(required) The file set whose files to import into the store.
File sets can be created using other functions in this library.
This argument can also be a path,
which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
:::{.note}
If a directory does not recursively contain any file, it is omitted from the store path contents.
:::
*/
fileset,
}:
let
# We cannot rename matched attribute arguments, so let's work around it with an extra `let in` statement
filesetArg = fileset;
in
let
fileset = _coerce "lib.fileset.toSource: `fileset`" filesetArg;
rootFilesystemRoot = (splitRoot root).root;
filesetFilesystemRoot = (splitRoot fileset._internalBase).root;
sourceFilter = _toSourceFilter fileset;
in
if ! isPath root then
if root ? _isLibCleanSourceWith then
throw ''
lib.fileset.toSource: `root` is a `lib.sources`-based value, but it should be a path instead.
To use a `lib.sources`-based value, convert it to a file set using `lib.fileset.fromSource` and pass it as `fileset`.
Note that this only works for sources created from paths.''
else if isStringLike root then
throw ''
lib.fileset.toSource: `root` (${toString root}) is a string-like value, but it should be a path instead.
Paths in strings are not supported by `lib.fileset`, use `lib.sources` or derivations instead.''
else
throw ''
lib.fileset.toSource: `root` is of type ${typeOf root}, but it should be a path instead.''
# Currently all Nix paths have the same filesystem root, but this could change in the future.
# See also ../path/README.md
else if ! fileset._internalIsEmptyWithoutBase && rootFilesystemRoot != filesetFilesystemRoot then
throw ''
lib.fileset.toSource: Filesystem roots are not the same for `fileset` and `root` (${toString root}):
`root`: Filesystem root is "${toString rootFilesystemRoot}"
`fileset`: Filesystem root is "${toString filesetFilesystemRoot}"
Different filesystem roots are not supported.''
else if ! pathExists root then
throw ''
lib.fileset.toSource: `root` (${toString root}) is a path that does not exist.''
else if pathType root != "directory" then
throw ''
lib.fileset.toSource: `root` (${toString root}) is a file, but it should be a directory instead. Potential solutions:
- If you want to import the file into the store _without_ a containing directory, use string interpolation or `builtins.path` instead of this function.
- If you want to import the file into the store _with_ a containing directory, set `root` to the containing directory, such as ${toString (dirOf root)}, and set `fileset` to the file path.''
else if ! fileset._internalIsEmptyWithoutBase && ! hasPrefix root fileset._internalBase then
throw ''
lib.fileset.toSource: `fileset` could contain files in ${toString fileset._internalBase}, which is not under the `root` (${toString root}). Potential solutions:
- Set `root` to ${toString fileset._internalBase} or any directory higher up. This changes the layout of the resulting store path.
- Set `fileset` to a file set that cannot contain files outside the `root` (${toString root}). This could change the files included in the result.''
else
seq sourceFilter
cleanSourceWith {
name = "source";
src = root;
filter = sourceFilter;
};
/*
The list of file paths contained in the given file set.
:::{.note}
This function is strict in the entire file set.
This is in contrast with combinators [`lib.fileset.union`](#function-library-lib.fileset.union),
[`lib.fileset.intersection`](#function-library-lib.fileset.intersection) and [`lib.fileset.difference`](#function-library-lib.fileset.difference).
Thus it is recommended to call `toList` on file sets created using the combinators,
instead of doing list processing on the result of `toList`.
:::
The resulting list of files can be turned back into a file set using [`lib.fileset.unions`](#function-library-lib.fileset.unions).
Type:
toList :: FileSet -> [ Path ]
Example:
toList ./.
[ ./README.md ./Makefile ./src/main.c ./src/main.h ]
toList (difference ./. ./src)
[ ./README.md ./Makefile ]
*/
toList =
# The file set whose file paths to return.
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
fileset:
_toList (_coerce "lib.fileset.toList: Argument" fileset);
/*
The file set containing all files that are in either of two given file sets.
This is the same as [`unions`](#function-library-lib.fileset.unions),
but takes just two file sets instead of a list.
See also [Union (set theory)](https://en.wikipedia.org/wiki/Union_(set_theory)).
The given file sets are evaluated as lazily as possible,
with the first argument being evaluated first if needed.
Type:
union :: FileSet -> FileSet -> FileSet
Example:
# Create a file set containing the file `Makefile`
# and all files recursively in the `src` directory
union ./Makefile ./src
# Create a file set containing the file `Makefile`
# and the LICENSE file from the parent directory
union ./Makefile ../LICENSE
*/
union =
# The first file set.
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
fileset1:
# The second file set.
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
fileset2:
_unionMany
(_coerceMany "lib.fileset.union" [
{
context = "First argument";
value = fileset1;
}
{
context = "Second argument";
value = fileset2;
}
]);
/*
The file set containing all files that are in any of the given file sets.
This is the same as [`union`](#function-library-lib.fileset.unions),
but takes a list of file sets instead of just two.
See also [Union (set theory)](https://en.wikipedia.org/wiki/Union_(set_theory)).
The given file sets are evaluated as lazily as possible,
with earlier elements being evaluated first if needed.
Type:
unions :: [ FileSet ] -> FileSet
Example:
# Create a file set containing selected files
unions [
# Include the single file `Makefile` in the current directory
# This errors if the file doesn't exist
./Makefile
# Recursively include all files in the `src/code` directory
# If this directory is empty this has no effect
./src/code
# Include the files `run.sh` and `unit.c` from the `tests` directory
./tests/run.sh
./tests/unit.c
# Include the `LICENSE` file from the parent directory
../LICENSE
]
*/
unions =
# A list of file sets.
# The elements can also be paths,
# which get [implicitly coerced to file sets](#sec-fileset-path-coercion).
filesets:
if ! isList filesets then
throw ''
lib.fileset.unions: Argument is of type ${typeOf filesets}, but it should be a list instead.''
else
pipe filesets [
# Annotate the elements with context, used by _coerceMany for better errors
(imap0 (i: el: {
context = "Element ${toString i}";
value = el;
}))
(_coerceMany "lib.fileset.unions")
_unionMany
];
/*
The file set containing all files that are in both of two given file sets.
See also [Intersection (set theory)](https://en.wikipedia.org/wiki/Intersection_(set_theory)).
The given file sets are evaluated as lazily as possible,
with the first argument being evaluated first if needed.
Type:
intersection :: FileSet -> FileSet -> FileSet
Example:
# Limit the selected files to the ones in ./., so only ./src and ./Makefile
intersection ./. (unions [ ../LICENSE ./src ./Makefile ])
*/
intersection =
# The first file set.
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
fileset1:
# The second file set.
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
fileset2:
let
filesets = _coerceMany "lib.fileset.intersection" [
{
context = "First argument";
value = fileset1;
}
{
context = "Second argument";
value = fileset2;
}
];
in
_intersection
(elemAt filesets 0)
(elemAt filesets 1);
/*
The file set containing all files from the first file set that are not in the second file set.
See also [Difference (set theory)](https://en.wikipedia.org/wiki/Complement_(set_theory)#Relative_complement).
The given file sets are evaluated as lazily as possible,
with the first argument being evaluated first if needed.
Type:
union :: FileSet -> FileSet -> FileSet
Example:
# Create a file set containing all files from the current directory,
# except ones under ./tests
difference ./. ./tests
let
# A set of Nix-related files
nixFiles = unions [ ./default.nix ./nix ./tests/default.nix ];
in
# Create a file set containing all files under ./tests, except ones in `nixFiles`,
# meaning only without ./tests/default.nix
difference ./tests nixFiles
*/
difference =
# The positive file set.
# The result can only contain files that are also in this file set.
#
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
positive:
# The negative file set.
# The result will never contain files that are also in this file set.
#
# This argument can also be a path,
# which gets [implicitly coerced to a file set](#sec-fileset-path-coercion).
negative:
let
filesets = _coerceMany "lib.fileset.difference" [
{
context = "First argument (positive set)";
value = positive;
}
{
context = "Second argument (negative set)";
value = negative;
}
];
in
_difference
(elemAt filesets 0)
(elemAt filesets 1);
/*
Filter a file set to only contain files matching some predicate.
Type:
fileFilter ::
({
name :: String,
type :: String,
hasExt :: String -> Bool,
...
} -> Bool)
-> Path
-> FileSet
Example:
# Include all regular `default.nix` files in the current directory
fileFilter (file: file.name == "default.nix") ./.
# Include all non-Nix files from the current directory
fileFilter (file: ! file.hasExt "nix") ./.
# Include all files that start with a "." in the current directory
fileFilter (file: hasPrefix "." file.name) ./.
# Include all regular files (not symlinks or others) in the current directory
fileFilter (file: file.type == "regular") ./.
*/
fileFilter =
/*
The predicate function to call on all files contained in given file set.
A file is included in the resulting file set if this function returns true for it.
This function is called with an attribute set containing these attributes:
- `name` (String): The name of the file
- `type` (String, one of `"regular"`, `"symlink"` or `"unknown"`): The type of the file.
This matches result of calling [`builtins.readFileType`](https://nixos.org/manual/nix/stable/language/builtins.html#builtins-readFileType) on the file's path.
- `hasExt` (String -> Bool): Whether the file has a certain file extension.
`hasExt ext` is true only if `hasSuffix ".${ext}" name`.
This also means that e.g. for a file with name `.gitignore`,
`hasExt "gitignore"` is true.
Other attributes may be added in the future.
*/
predicate:
# The path whose files to filter
path:
if ! isFunction predicate then
throw ''
lib.fileset.fileFilter: First argument is of type ${typeOf predicate}, but it should be a function instead.''
else if ! isPath path then
if path._type or "" == "fileset" then
throw ''
lib.fileset.fileFilter: Second argument is a file set, but it should be a path instead.
If you need to filter files in a file set, use `intersection fileset (fileFilter pred ./.)` instead.''
else
throw ''
lib.fileset.fileFilter: Second argument is of type ${typeOf path}, but it should be a path instead.''
else if ! pathExists path then
throw ''
lib.fileset.fileFilter: Second argument (${toString path}) is a path that does not exist.''
else
_fileFilter predicate path;
/*
Create a file set with the same files as a `lib.sources`-based value.
This does not import any of the files into the store.
This can be used to gradually migrate from `lib.sources`-based filtering to `lib.fileset`.
A file set can be turned back into a source using [`toSource`](#function-library-lib.fileset.toSource).
:::{.note}
File sets cannot represent empty directories.
Turning the result of this function back into a source using `toSource` will therefore not preserve empty directories.
:::
Type:
fromSource :: SourceLike -> FileSet
Example:
# There's no cleanSource-like function for file sets yet,
# but we can just convert cleanSource to a file set and use it that way
toSource {
root = ./.;
fileset = fromSource (lib.sources.cleanSource ./.);
}
# Keeping a previous sourceByRegex (which could be migrated to `lib.fileset.unions`),
# but removing a subdirectory using file set functions
difference
(fromSource (lib.sources.sourceByRegex ./. [
"^README\.md$"
# This regex includes everything in ./doc
"^doc(/.*)?$"
])
./doc/generated
# Use cleanSource, but limit it to only include ./Makefile and files under ./src
intersection
(fromSource (lib.sources.cleanSource ./.))
(unions [
./Makefile
./src
]);
*/
fromSource = source:
let
# This function uses `._isLibCleanSourceWith`, `.origSrc` and `.filter`,
# which are technically internal to lib.sources,
# but we'll allow this since both libraries are in the same code base
# and this function is a bridge between them.
isFiltered = source ? _isLibCleanSourceWith;
path = if isFiltered then source.origSrc else source;
in
# We can only support sources created from paths
if ! isPath path then
if isStringLike path then
throw ''
lib.fileset.fromSource: The source origin of the argument is a string-like value ("${toString path}"), but it should be a path instead.
Sources created from paths in strings cannot be turned into file sets, use `lib.sources` or derivations instead.''
else
throw ''
lib.fileset.fromSource: The source origin of the argument is of type ${typeOf path}, but it should be a path instead.''
else if ! pathExists path then
throw ''
lib.fileset.fromSource: The source origin (${toString path}) of the argument is a path that does not exist.''
else if isFiltered then
_fromSourceFilter path source.filter
else
# If there's no filter, no need to run the expensive conversion, all subpaths will be included
_singleton path;
/*
Create a file set containing all [Git-tracked files](https://git-scm.com/book/en/v2/Git-Basics-Recording-Changes-to-the-Repository) in a repository.
This function behaves like [`gitTrackedWith { }`](#function-library-lib.fileset.gitTrackedWith) - using the defaults.
Type:
gitTracked :: Path -> FileSet
Example:
# Include all files tracked by the Git repository in the current directory
gitTracked ./.
# Include only files tracked by the Git repository in the parent directory
# that are also in the current directory
intersection ./. (gitTracked ../.)
*/
gitTracked =
/*
The [path](https://nixos.org/manual/nix/stable/language/values#type-path) to the working directory of a local Git repository.
This directory must contain a `.git` file or subdirectory.
*/
path:
_fromFetchGit
"gitTracked"
"argument"
path
{};
/*
Create a file set containing all [Git-tracked files](https://git-scm.com/book/en/v2/Git-Basics-Recording-Changes-to-the-Repository) in a repository.
The first argument allows configuration with an attribute set,
while the second argument is the path to the Git working tree.
`gitTrackedWith` does not perform any filtering when the path is a [Nix store path](https://nixos.org/manual/nix/stable/store/store-path.html#store-path) and not a repository.
In this way, it accommodates the use case where the expression that makes the `gitTracked` call does not reside in an actual git repository anymore,
and has presumably already been fetched in a way that excludes untracked files.
Fetchers with such equivalent behavior include `builtins.fetchGit`, `builtins.fetchTree` (experimental), and `pkgs.fetchgit` when used without `leaveDotGit`.
If you don't need the configuration,
you can use [`gitTracked`](#function-library-lib.fileset.gitTracked) instead.
This is equivalent to the result of [`unions`](#function-library-lib.fileset.unions) on all files returned by [`git ls-files`](https://git-scm.com/docs/git-ls-files)
(which uses [`--cached`](https://git-scm.com/docs/git-ls-files#Documentation/git-ls-files.txt--c) by default).
:::{.warning}
Currently this function is based on [`builtins.fetchGit`](https://nixos.org/manual/nix/stable/language/builtins.html#builtins-fetchGit)
As such, this function causes all Git-tracked files to be unnecessarily added to the Nix store,
without being re-usable by [`toSource`](#function-library-lib.fileset.toSource).
This may change in the future.
:::
Type:
gitTrackedWith :: { recurseSubmodules :: Bool ? false } -> Path -> FileSet
Example:
# Include all files tracked by the Git repository in the current directory
# and any submodules under it
gitTracked { recurseSubmodules = true; } ./.
*/
gitTrackedWith =
{
/*
(optional, default: `false`) Whether to recurse into [Git submodules](https://git-scm.com/book/en/v2/Git-Tools-Submodules) to also include their tracked files.
If `true`, this is equivalent to passing the [--recurse-submodules](https://git-scm.com/docs/git-ls-files#Documentation/git-ls-files.txt---recurse-submodules) flag to `git ls-files`.
*/
recurseSubmodules ? false,
}:
/*
The [path](https://nixos.org/manual/nix/stable/language/values#type-path) to the working directory of a local Git repository.
This directory must contain a `.git` file or subdirectory.
*/
path:
if ! isBool recurseSubmodules then
throw "lib.fileset.gitTrackedWith: Expected the attribute `recurseSubmodules` of the first argument to be a boolean, but it's a ${typeOf recurseSubmodules} instead."
else if recurseSubmodules && versionOlder nixVersion _fetchGitSubmodulesMinver then
throw "lib.fileset.gitTrackedWith: Setting the attribute `recurseSubmodules` to `true` is only supported for Nix version ${_fetchGitSubmodulesMinver} and after, but Nix version ${nixVersion} is used."
else
_fromFetchGit
"gitTrackedWith"
"second argument"
path
# This is the only `fetchGit` parameter that makes sense in this context.
# We can't just pass `submodules = recurseSubmodules` here because
# this would fail for Nix versions that don't support `submodules`.
(lib.optionalAttrs recurseSubmodules {
submodules = true;
});
}

958
lib/fileset/internal.nix Normal file
View file

@ -0,0 +1,958 @@
{ lib ? import ../. }:
let
inherit (builtins)
isAttrs
isPath
isString
nixVersion
pathExists
readDir
split
trace
typeOf
fetchGit
;
inherit (lib.attrsets)
attrNames
attrValues
mapAttrs
mapAttrsToList
optionalAttrs
zipAttrsWith
;
inherit (lib.filesystem)
pathType
;
inherit (lib.lists)
all
commonPrefix
concatLists
elemAt
filter
findFirst
findFirstIndex
foldl'
head
length
sublist
tail
;
inherit (lib.path)
append
splitRoot
hasStorePathPrefix
splitStorePath
;
inherit (lib.path.subpath)
components
join
;
inherit (lib.strings)
isStringLike
concatStringsSep
substring
stringLength
hasSuffix
versionAtLeast
;
inherit (lib.trivial)
inPureEvalMode
;
in
# Rare case of justified usage of rec:
# - This file is internal, so the return value doesn't matter, no need to make things overridable
# - The functions depend on each other
# - We want to expose all of these functions for easy testing
rec {
# If you change the internal representation, make sure to:
# - Increment this version
# - Add an additional migration function below
# - Update the description of the internal representation in ./README.md
_currentVersion = 3;
# Migrations between versions. The 0th element converts from v0 to v1, and so on
migrations = [
# Convert v0 into v1: Add the _internalBase{Root,Components} attributes
(
filesetV0:
let
parts = splitRoot filesetV0._internalBase;
in
filesetV0 // {
_internalVersion = 1;
_internalBaseRoot = parts.root;
_internalBaseComponents = components parts.subpath;
}
)
# Convert v1 into v2: filesetTree's can now also omit attributes to signal paths not being included
(
filesetV1:
# This change is backwards compatible (but not forwards compatible, so we still need a new version)
filesetV1 // {
_internalVersion = 2;
}
)
# Convert v2 into v3: filesetTree's now have a representation for an empty file set without a base path
(
filesetV2:
filesetV2 // {
# All v1 file sets are not the new empty file set
_internalIsEmptyWithoutBase = false;
_internalVersion = 3;
}
)
];
_noEvalMessage = ''
lib.fileset: Directly evaluating a file set is not supported.
To turn it into a usable source, use `lib.fileset.toSource`.
To pretty-print the contents, use `lib.fileset.trace` or `lib.fileset.traceVal`.'';
# The empty file set without a base path
_emptyWithoutBase = {
_type = "fileset";
_internalVersion = _currentVersion;
# The one and only!
_internalIsEmptyWithoutBase = true;
# Due to alphabetical ordering, this is evaluated last,
# which makes the nix repl output nicer than if it would be ordered first.
# It also allows evaluating it strictly up to this error, which could be useful
_noEval = throw _noEvalMessage;
};
# Create a fileset, see ./README.md#fileset
# Type: path -> filesetTree -> fileset
_create = base: tree:
let
# Decompose the base into its components
# See ../path/README.md for why we're not just using `toString`
parts = splitRoot base;
in
{
_type = "fileset";
_internalVersion = _currentVersion;
_internalIsEmptyWithoutBase = false;
_internalBase = base;
_internalBaseRoot = parts.root;
_internalBaseComponents = components parts.subpath;
_internalTree = tree;
# Due to alphabetical ordering, this is evaluated last,
# which makes the nix repl output nicer than if it would be ordered first.
# It also allows evaluating it strictly up to this error, which could be useful
_noEval = throw _noEvalMessage;
};
# Coerce a value to a fileset, erroring when the value cannot be coerced.
# The string gives the context for error messages.
# Type: String -> (fileset | Path) -> fileset
_coerce = context: value:
if value._type or "" == "fileset" then
if value._internalVersion > _currentVersion then
throw ''
${context} is a file set created from a future version of the file set library with a different internal representation:
- Internal version of the file set: ${toString value._internalVersion}
- Internal version of the library: ${toString _currentVersion}
Make sure to update your Nixpkgs to have a newer version of `lib.fileset`.''
else if value._internalVersion < _currentVersion then
let
# Get all the migration functions necessary to convert from the old to the current version
migrationsToApply = sublist value._internalVersion (_currentVersion - value._internalVersion) migrations;
in
foldl' (value: migration: migration value) value migrationsToApply
else
value
else if ! isPath value then
if value ? _isLibCleanSourceWith then
throw ''
${context} is a `lib.sources`-based value, but it should be a file set or a path instead.
To convert a `lib.sources`-based value to a file set you can use `lib.fileset.fromSource`.
Note that this only works for sources created from paths.''
else if isStringLike value then
throw ''
${context} ("${toString value}") is a string-like value, but it should be a file set or a path instead.
Paths represented as strings are not supported by `lib.fileset`, use `lib.sources` or derivations instead.''
else
throw ''
${context} is of type ${typeOf value}, but it should be a file set or a path instead.''
else if ! pathExists value then
throw ''
${context} (${toString value}) is a path that does not exist.
To create a file set from a path that may not exist, use `lib.fileset.maybeMissing`.''
else
_singleton value;
# Coerce many values to filesets, erroring when any value cannot be coerced,
# or if the filesystem root of the values doesn't match.
# Type: String -> [ { context :: String, value :: fileset | Path } ] -> [ fileset ]
_coerceMany = functionContext: list:
let
filesets = map ({ context, value }:
_coerce "${functionContext}: ${context}" value
) list;
# Find the first value with a base, there may be none!
firstWithBase = findFirst (fileset: ! fileset._internalIsEmptyWithoutBase) null filesets;
# This value is only accessed if first != null
firstBaseRoot = firstWithBase._internalBaseRoot;
# Finds the first element with a filesystem root different than the first element, if any
differentIndex = findFirstIndex (fileset:
# The empty value without a base doesn't have a base path
! fileset._internalIsEmptyWithoutBase
&& firstBaseRoot != fileset._internalBaseRoot
) null filesets;
in
# Only evaluates `differentIndex` if there are any elements with a base
if firstWithBase != null && differentIndex != null then
throw ''
${functionContext}: Filesystem roots are not the same:
${(head list).context}: Filesystem root is "${toString firstBaseRoot}"
${(elemAt list differentIndex).context}: Filesystem root is "${toString (elemAt filesets differentIndex)._internalBaseRoot}"
Different filesystem roots are not supported.''
else
filesets;
# Create a file set from a path.
# Type: Path -> fileset
_singleton = path:
let
type = pathType path;
in
if type == "directory" then
_create path type
else
# This turns a file path ./default.nix into a fileset with
# - _internalBase: ./.
# - _internalTree: {
# "default.nix" = <type>;
# }
# See ./README.md#single-files
_create (dirOf path)
{
${baseNameOf path} = type;
};
# Expand a directory representation to an equivalent one in attribute set form.
# All directory entries are included in the result.
# Type: Path -> filesetTree -> { <name> = filesetTree; }
_directoryEntries = path: value:
if value == "directory" then
readDir path
else
# Set all entries not present to null
mapAttrs (name: value: null) (readDir path)
// value;
/*
A normalisation of a filesetTree suitable filtering with `builtins.path`:
- Replace all directories that have no files with `null`.
This removes directories that would be empty
- Replace all directories with all files with `"directory"`.
This speeds up the source filter function
Note that this function is strict, it evaluates the entire tree
Type: Path -> filesetTree -> filesetTree
*/
_normaliseTreeFilter = path: tree:
if tree == "directory" || isAttrs tree then
let
entries = _directoryEntries path tree;
normalisedSubtrees = mapAttrs (name: _normaliseTreeFilter (path + "/${name}")) entries;
subtreeValues = attrValues normalisedSubtrees;
in
# This triggers either when all files in a directory are filtered out
# Or when the directory doesn't contain any files at all
if all isNull subtreeValues then
null
# Triggers when we have the same as a `readDir path`, so we can turn it back into an equivalent "directory".
else if all isString subtreeValues then
"directory"
else
normalisedSubtrees
else
tree;
/*
A minimal normalisation of a filesetTree, intended for pretty-printing:
- If all children of a path are recursively included or empty directories, the path itself is also recursively included
- If all children of a path are fully excluded or empty directories, the path itself is an empty directory
- Other empty directories are represented with the special "emptyDir" string
While these could be replaced with `null`, that would take another mapAttrs
Note that this function is partially lazy.
Type: Path -> filesetTree -> filesetTree (with "emptyDir"'s)
*/
_normaliseTreeMinimal = path: tree:
if tree == "directory" || isAttrs tree then
let
entries = _directoryEntries path tree;
normalisedSubtrees = mapAttrs (name: _normaliseTreeMinimal (path + "/${name}")) entries;
subtreeValues = attrValues normalisedSubtrees;
in
# If there are no entries, or all entries are empty directories, return "emptyDir".
# After this branch we know that there's at least one file
if all (value: value == "emptyDir") subtreeValues then
"emptyDir"
# If all subtrees are fully included or empty directories
# (both of which are coincidentally represented as strings), return "directory".
# This takes advantage of the fact that empty directories can be represented as included directories.
# Note that the tree == "directory" check allows avoiding recursion
else if tree == "directory" || all (value: isString value) subtreeValues then
"directory"
# If all subtrees are fully excluded or empty directories, return null.
# This takes advantage of the fact that empty directories can be represented as excluded directories
else if all (value: isNull value || value == "emptyDir") subtreeValues then
null
# Mix of included and excluded entries
else
normalisedSubtrees
else
tree;
# Trace a filesetTree in a pretty way when the resulting value is evaluated.
# This can handle both normal filesetTree's, and ones returned from _normaliseTreeMinimal
# Type: Path -> filesetTree (with "emptyDir"'s) -> Null
_printMinimalTree = base: tree:
let
treeSuffix = tree:
if isAttrs tree then
""
else if tree == "directory" then
" (all files in directory)"
else
# This does "leak" the file type strings of the internal representation,
# but this is the main reason these file type strings even are in the representation!
# TODO: Consider removing that information from the internal representation for performance.
# The file types can still be printed by querying them only during tracing
" (${tree})";
# Only for attribute set trees
traceTreeAttrs = prevLine: indent: tree:
foldl' (prevLine: name:
let
subtree = tree.${name};
# Evaluating this prints the line for this subtree
thisLine =
trace "${indent}- ${name}${treeSuffix subtree}" prevLine;
in
if subtree == null || subtree == "emptyDir" then
# Don't print anything at all if this subtree is empty
prevLine
else if isAttrs subtree then
# A directory with explicit entries
# Do print this node, but also recurse
traceTreeAttrs thisLine "${indent} " subtree
else
# Either a file, or a recursively included directory
# Do print this node but no further recursion needed
thisLine
) prevLine (attrNames tree);
# Evaluating this will print the first line
firstLine =
if tree == null || tree == "emptyDir" then
trace "(empty)" null
else
trace "${toString base}${treeSuffix tree}" null;
in
if isAttrs tree then
traceTreeAttrs firstLine "" tree
else
firstLine;
# Pretty-print a file set in a pretty way when the resulting value is evaluated
# Type: fileset -> Null
_printFileset = fileset:
if fileset._internalIsEmptyWithoutBase then
trace "(empty)" null
else
_printMinimalTree fileset._internalBase
(_normaliseTreeMinimal fileset._internalBase fileset._internalTree);
# Turn a fileset into a source filter function suitable for `builtins.path`
# Only directories recursively containing at least one files are recursed into
# Type: fileset -> (String -> String -> Bool)
_toSourceFilter = fileset:
let
# Simplify the tree, necessary to make sure all empty directories are null
# which has the effect that they aren't included in the result
tree = _normaliseTreeFilter fileset._internalBase fileset._internalTree;
# The base path as a string with a single trailing slash
baseString =
if fileset._internalBaseComponents == [] then
# Need to handle the filesystem root specially
"/"
else
"/" + concatStringsSep "/" fileset._internalBaseComponents + "/";
baseLength = stringLength baseString;
# Check whether a list of path components under the base path exists in the tree.
# This function is called often, so it should be fast.
# Type: [ String ] -> Bool
inTree = components:
let
recurse = index: localTree:
if isAttrs localTree then
# We have an attribute set, meaning this is a directory with at least one file
if index >= length components then
# The path may have no more components though, meaning the filter is running on the directory itself,
# so we always include it, again because there's at least one file in it.
true
else
# If we do have more components, the filter runs on some entry inside this directory, so we need to recurse
# We do +2 because builtins.split is an interleaved list of the inbetweens and the matches
recurse (index + 2) localTree.${elemAt components index}
else
# If it's not an attribute set it can only be either null (in which case it's not included)
# or a string ("directory" or "regular", etc.) in which case it's included
localTree != null;
in recurse 0 tree;
# Filter suited when there's no files
empty = _: _: false;
# Filter suited when there's some files
# This can't be used for when there's no files, because the base directory is always included
nonEmpty =
path: type:
let
# Add a slash to the path string, turning "/foo" to "/foo/",
# making sure to not have any false prefix matches below.
# Note that this would produce "//" for "/",
# but builtins.path doesn't call the filter function on the `path` argument itself,
# meaning this function can never receive "/" as an argument
pathSlash = path + "/";
in
(
# Same as `hasPrefix pathSlash baseString`, but more efficient.
# With base /foo/bar we need to include /foo:
# hasPrefix "/foo/" "/foo/bar/"
if substring 0 (stringLength pathSlash) baseString == pathSlash then
true
# Same as `! hasPrefix baseString pathSlash`, but more efficient.
# With base /foo/bar we need to exclude /baz
# ! hasPrefix "/baz/" "/foo/bar/"
else if substring 0 baseLength pathSlash != baseString then
false
else
# Same as `removePrefix baseString path`, but more efficient.
# From the above code we know that hasPrefix baseString pathSlash holds, so this is safe.
# We don't use pathSlash here because we only needed the trailing slash for the prefix matching.
# With base /foo and path /foo/bar/baz this gives
# inTree (split "/" (removePrefix "/foo/" "/foo/bar/baz"))
# == inTree (split "/" "bar/baz")
# == inTree [ "bar" "baz" ]
inTree (split "/" (substring baseLength (-1) path))
)
# This is a way have an additional check in case the above is true without any significant performance cost
&& (
# This relies on the fact that Nix only distinguishes path types "directory", "regular", "symlink" and "unknown",
# so everything except "unknown" is allowed, seems reasonable to rely on that
type != "unknown"
|| throw ''
lib.fileset.toSource: `fileset` contains a file that cannot be added to the store: ${path}
This file is neither a regular file nor a symlink, the only file types supported by the Nix store.
Therefore the file set cannot be added to the Nix store as is. Make sure to not include that file to avoid this error.''
);
in
# Special case because the code below assumes that the _internalBase is always included in the result
# which shouldn't be done when we have no files at all in the base
# This also forces the tree before returning the filter, leads to earlier error messages
if fileset._internalIsEmptyWithoutBase || tree == null then
empty
else
nonEmpty;
# Turn a builtins.filterSource-based source filter on a root path into a file set
# containing only files included by the filter.
# The filter is lazily called as necessary to determine whether paths are included
# Type: Path -> (String -> String -> Bool) -> fileset
_fromSourceFilter = root: sourceFilter:
let
# During the recursion we need to track both:
# - The path value such that we can safely call `readDir` on it
# - The path string value such that we can correctly call the `filter` with it
#
# While we could just recurse with the path value,
# this would then require converting it to a path string for every path,
# which is a fairly expensive operation
# Create a file set from a directory entry
fromDirEntry = path: pathString: type:
# The filter needs to run on the path as a string
if ! sourceFilter pathString type then
null
else if type == "directory" then
fromDir path pathString
else
type;
# Create a file set from a directory
fromDir = path: pathString:
mapAttrs
# This looks a bit funny, but we need both the path-based and the path string-based values
(name: fromDirEntry (path + "/${name}") (pathString + "/${name}"))
# We need to readDir on the path value, because reading on a path string
# would be unspecified if there are multiple filesystem roots
(readDir path);
rootPathType = pathType root;
# We need to convert the path to a string to imitate what builtins.path calls the filter function with.
# We don't want to rely on `toString` for this though because it's not very well defined, see ../path/README.md
# So instead we use `lib.path.splitRoot` to safely deconstruct the path into its filesystem root and subpath
# We don't need the filesystem root though, builtins.path doesn't expose that in any way to the filter.
# So we only need the components, which we then turn into a string as one would expect.
rootString = "/" + concatStringsSep "/" (components (splitRoot root).subpath);
in
if rootPathType == "directory" then
# We imitate builtins.path not calling the filter on the root path
_create root (fromDir root rootString)
else
# Direct files are always included by builtins.path without calling the filter
# But we need to lift up the base path to its parent to satisfy the base path invariant
_create (dirOf root)
{
${baseNameOf root} = rootPathType;
};
# Turns a file set into the list of file paths it includes.
# Type: fileset -> [ Path ]
_toList = fileset:
let
recurse = path: tree:
if isAttrs tree then
concatLists (mapAttrsToList (name: value:
recurse (path + "/${name}") value
) tree)
else if tree == "directory" then
recurse path (readDir path)
else if tree == null then
[ ]
else
[ path ];
in
if fileset._internalIsEmptyWithoutBase then
[ ]
else
recurse fileset._internalBase fileset._internalTree;
# Transforms the filesetTree of a file set to a shorter base path, e.g.
# _shortenTreeBase [ "foo" ] (_create /foo/bar null)
# => { bar = null; }
_shortenTreeBase = targetBaseComponents: fileset:
let
recurse = index:
# If we haven't reached the required depth yet
if index < length fileset._internalBaseComponents then
# Create an attribute set and recurse as the value, this can be lazily evaluated this way
{ ${elemAt fileset._internalBaseComponents index} = recurse (index + 1); }
else
# Otherwise we reached the appropriate depth, here's the original tree
fileset._internalTree;
in
recurse (length targetBaseComponents);
# Transforms the filesetTree of a file set to a longer base path, e.g.
# _lengthenTreeBase [ "foo" "bar" ] (_create /foo { bar.baz = "regular"; })
# => { baz = "regular"; }
_lengthenTreeBase = targetBaseComponents: fileset:
let
recurse = index: tree:
# If the filesetTree is an attribute set and we haven't reached the required depth yet
if isAttrs tree && index < length targetBaseComponents then
# Recurse with the tree under the right component (which might not exist)
recurse (index + 1) (tree.${elemAt targetBaseComponents index} or null)
else
# For all values here we can just return the tree itself:
# tree == null -> the result is also null, everything is excluded
# tree == "directory" -> the result is also "directory",
# because the base path is always a directory and everything is included
# isAttrs tree -> the result is `tree`
# because we don't need to recurse any more since `index == length longestBaseComponents`
tree;
in
recurse (length fileset._internalBaseComponents) fileset._internalTree;
# Computes the union of a list of filesets.
# The filesets must already be coerced and validated to be in the same filesystem root
# Type: [ Fileset ] -> Fileset
_unionMany = filesets:
let
# All filesets that have a base, aka not the ones that are the empty value without a base
filesetsWithBase = filter (fileset: ! fileset._internalIsEmptyWithoutBase) filesets;
# The first fileset that has a base.
# This value is only accessed if there are at all.
firstWithBase = head filesetsWithBase;
# To be able to union filesetTree's together, they need to have the same base path.
# Base paths can be unioned by taking their common prefix,
# e.g. such that `union /foo/bar /foo/baz` has the base path `/foo`
# A list of path components common to all base paths.
# Note that commonPrefix can only be fully evaluated,
# so this cannot cause a stack overflow due to a build-up of unevaluated thunks.
commonBaseComponents = foldl'
(components: el: commonPrefix components el._internalBaseComponents)
firstWithBase._internalBaseComponents
# We could also not do the `tail` here to avoid a list allocation,
# but then we'd have to pay for a potentially expensive
# but unnecessary `commonPrefix` call
(tail filesetsWithBase);
# The common base path assembled from a filesystem root and the common components
commonBase = append firstWithBase._internalBaseRoot (join commonBaseComponents);
# A list of filesetTree's that all have the same base path
# This is achieved by nesting the trees into the components they have over the common base path
# E.g. `union /foo/bar /foo/baz` has the base path /foo
# So the tree under `/foo/bar` gets nested under `{ bar = ...; ... }`,
# while the tree under `/foo/baz` gets nested under `{ baz = ...; ... }`
# Therefore allowing combined operations over them.
trees = map (_shortenTreeBase commonBaseComponents) filesetsWithBase;
# Folds all trees together into a single one using _unionTree
# We do not use a fold here because it would cause a thunk build-up
# which could cause a stack overflow for a large number of trees
resultTree = _unionTrees trees;
in
# If there's no values with a base, we have no files
if filesetsWithBase == [ ] then
_emptyWithoutBase
else
_create commonBase resultTree;
# The union of multiple filesetTree's with the same base path.
# Later elements are only evaluated if necessary.
# Type: [ filesetTree ] -> filesetTree
_unionTrees = trees:
let
stringIndex = findFirstIndex isString null trees;
withoutNull = filter (tree: tree != null) trees;
in
if stringIndex != null then
# If there's a string, it's always a fully included tree (dir or file),
# no need to look at other elements
elemAt trees stringIndex
else if withoutNull == [ ] then
# If all trees are null, then the resulting tree is also null
null
else
# The non-null elements have to be attribute sets representing partial trees
# We need to recurse into those
zipAttrsWith (name: _unionTrees) withoutNull;
# Computes the intersection of a list of filesets.
# The filesets must already be coerced and validated to be in the same filesystem root
# Type: Fileset -> Fileset -> Fileset
_intersection = fileset1: fileset2:
let
# The common base components prefix, e.g.
# (/foo/bar, /foo/bar/baz) -> /foo/bar
# (/foo/bar, /foo/baz) -> /foo
commonBaseComponentsLength =
# TODO: Have a `lib.lists.commonPrefixLength` function such that we don't need the list allocation from commonPrefix here
length (
commonPrefix
fileset1._internalBaseComponents
fileset2._internalBaseComponents
);
# To be able to intersect filesetTree's together, they need to have the same base path.
# Base paths can be intersected by taking the longest one (if any)
# The fileset with the longest base, if any, e.g.
# (/foo/bar, /foo/bar/baz) -> /foo/bar/baz
# (/foo/bar, /foo/baz) -> null
longestBaseFileset =
if commonBaseComponentsLength == length fileset1._internalBaseComponents then
# The common prefix is the same as the first path, so the second path is equal or longer
fileset2
else if commonBaseComponentsLength == length fileset2._internalBaseComponents then
# The common prefix is the same as the second path, so the first path is longer
fileset1
else
# The common prefix is neither the first nor the second path
# This means there's no overlap between the two sets
null;
# Whether the result should be the empty value without a base
resultIsEmptyWithoutBase =
# If either fileset is the empty fileset without a base, the intersection is too
fileset1._internalIsEmptyWithoutBase
|| fileset2._internalIsEmptyWithoutBase
# If there is no overlap between the base paths
|| longestBaseFileset == null;
# Lengthen each fileset's tree to the longest base prefix
tree1 = _lengthenTreeBase longestBaseFileset._internalBaseComponents fileset1;
tree2 = _lengthenTreeBase longestBaseFileset._internalBaseComponents fileset2;
# With two filesetTree's with the same base, we can compute their intersection
resultTree = _intersectTree tree1 tree2;
in
if resultIsEmptyWithoutBase then
_emptyWithoutBase
else
_create longestBaseFileset._internalBase resultTree;
# The intersection of two filesetTree's with the same base path
# The second element is only evaluated as much as necessary.
# Type: filesetTree -> filesetTree -> filesetTree
_intersectTree = lhs: rhs:
if isAttrs lhs && isAttrs rhs then
# Both sides are attribute sets, we can recurse for the attributes existing on both sides
mapAttrs
(name: _intersectTree lhs.${name})
(builtins.intersectAttrs lhs rhs)
else if lhs == null || isString rhs then
# If the lhs is null, the result should also be null
# And if the rhs is the identity element
# (a string, aka it includes everything), then it's also the lhs
lhs
else
# In all other cases it's the rhs
rhs;
# Compute the set difference between two file sets.
# The filesets must already be coerced and validated to be in the same filesystem root.
# Type: Fileset -> Fileset -> Fileset
_difference = positive: negative:
let
# The common base components prefix, e.g.
# (/foo/bar, /foo/bar/baz) -> /foo/bar
# (/foo/bar, /foo/baz) -> /foo
commonBaseComponentsLength =
# TODO: Have a `lib.lists.commonPrefixLength` function such that we don't need the list allocation from commonPrefix here
length (
commonPrefix
positive._internalBaseComponents
negative._internalBaseComponents
);
# We need filesetTree's with the same base to be able to compute the difference between them
# This here is the filesetTree from the negative file set, but for a base path that matches the positive file set.
# Examples:
# For `difference /foo /foo/bar`, `negativeTreeWithPositiveBase = { bar = "directory"; }`
# because under the base path of `/foo`, only `bar` from the negative file set is included
# For `difference /foo/bar /foo`, `negativeTreeWithPositiveBase = "directory"`
# because under the base path of `/foo/bar`, everything from the negative file set is included
# For `difference /foo /bar`, `negativeTreeWithPositiveBase = null`
# because under the base path of `/foo`, nothing from the negative file set is included
negativeTreeWithPositiveBase =
if commonBaseComponentsLength == length positive._internalBaseComponents then
# The common prefix is the same as the positive base path, so the second path is equal or longer.
# We need to _shorten_ the negative filesetTree to the same base path as the positive one
# E.g. for `difference /foo /foo/bar` the common prefix is /foo, equal to the positive file set's base
# So we need to shorten the base of the tree for the negative argument from /foo/bar to just /foo
_shortenTreeBase positive._internalBaseComponents negative
else if commonBaseComponentsLength == length negative._internalBaseComponents then
# The common prefix is the same as the negative base path, so the first path is longer.
# We need to lengthen the negative filesetTree to the same base path as the positive one.
# E.g. for `difference /foo/bar /foo` the common prefix is /foo, equal to the negative file set's base
# So we need to lengthen the base of the tree for the negative argument from /foo to /foo/bar
_lengthenTreeBase positive._internalBaseComponents negative
else
# The common prefix is neither the first nor the second path.
# This means there's no overlap between the two file sets,
# and nothing from the negative argument should get removed from the positive one
# E.g for `difference /foo /bar`, we remove nothing to get the same as `/foo`
null;
resultingTree =
_differenceTree
positive._internalBase
positive._internalTree
negativeTreeWithPositiveBase;
in
# If the first file set is empty, we can never have any files in the result
if positive._internalIsEmptyWithoutBase then
_emptyWithoutBase
# If the second file set is empty, nothing gets removed, so the result is just the first file set
else if negative._internalIsEmptyWithoutBase then
positive
else
# We use the positive file set base for the result,
# because only files from the positive side may be included,
# which is what base path is for
_create positive._internalBase resultingTree;
# Computes the set difference of two filesetTree's
# Type: Path -> filesetTree -> filesetTree
_differenceTree = path: lhs: rhs:
# If the lhs doesn't have any files, or the right hand side includes all files
if lhs == null || isString rhs then
# The result will always be empty
null
# If the right hand side has no files
else if rhs == null then
# The result is always the left hand side, because nothing gets removed
lhs
else
# Otherwise we always have two attribute sets to recurse into
mapAttrs (name: lhsValue:
_differenceTree (path + "/${name}") lhsValue (rhs.${name} or null)
) (_directoryEntries path lhs);
# Filters all files in a path based on a predicate
# Type: ({ name, type, ... } -> Bool) -> Path -> FileSet
_fileFilter = predicate: root:
let
# Check the predicate for a single file
# Type: String -> String -> filesetTree
fromFile = name: type:
if
predicate {
inherit name type;
hasExt = ext: hasSuffix ".${ext}" name;
# To ensure forwards compatibility with more arguments being added in the future,
# adding an attribute which can't be deconstructed :)
"lib.fileset.fileFilter: The predicate function passed as the first argument must be able to handle extra attributes for future compatibility. If you're using `{ name, file, hasExt }:`, use `{ name, file, hasExt, ... }:` instead." = null;
}
then
type
else
null;
# Check the predicate for all files in a directory
# Type: Path -> filesetTree
fromDir = path:
mapAttrs (name: type:
if type == "directory" then
fromDir (path + "/${name}")
else
fromFile name type
) (readDir path);
rootType = pathType root;
in
if rootType == "directory" then
_create root (fromDir root)
else
# Single files are turned into a directory containing that file or nothing.
_create (dirOf root) {
${baseNameOf root} =
fromFile (baseNameOf root) rootType;
};
# Support for `builtins.fetchGit` with `submodules = true` was introduced in 2.4
# https://github.com/NixOS/nix/commit/55cefd41d63368d4286568e2956afd535cb44018
_fetchGitSubmodulesMinver = "2.4";
# Support for `builtins.fetchGit` with `shallow = true` was introduced in 2.4
# https://github.com/NixOS/nix/commit/d1165d8791f559352ff6aa7348e1293b2873db1c
_fetchGitShallowMinver = "2.4";
# Mirrors the contents of a Nix store path relative to a local path as a file set.
# Some notes:
# - The store path is read at evaluation time.
# - The store path must not include files that don't exist in the respective local path.
#
# Type: Path -> String -> FileSet
_mirrorStorePath = localPath: storePath:
let
recurse = focusedStorePath:
mapAttrs (name: type:
if type == "directory" then
recurse (focusedStorePath + "/${name}")
else
type
) (builtins.readDir focusedStorePath);
in
_create localPath
(recurse storePath);
# Create a file set from the files included in the result of a fetchGit call
# Type: String -> String -> Path -> Attrs -> FileSet
_fromFetchGit = function: argument: path: extraFetchGitAttrs:
let
# The code path for when isStorePath is true
tryStorePath =
if pathExists (path + "/.git") then
# If there is a `.git` directory in the path,
# it means that the path was imported unfiltered into the Nix store.
# This function should throw in such a case, because
# - `fetchGit` doesn't generally work with `.git` directories in store paths
# - Importing the entire path could include Git-tracked files
throw ''
lib.fileset.${function}: The ${argument} (${toString path}) is a store path within a working tree of a Git repository.
This indicates that a source directory was imported into the store using a method such as `import "''${./.}"` or `path:.`.
This function currently does not support such a use case, since it currently relies on `builtins.fetchGit`.
You could make this work by using a fetcher such as `fetchGit` instead of copying the whole repository.
If you can't avoid copying the repo to the store, see https://github.com/NixOS/nix/issues/9292.''
else
# Otherwise we're going to assume that the path was a Git directory originally,
# but it was fetched using a method that already removed files not tracked by Git,
# such as `builtins.fetchGit`, `pkgs.fetchgit` or others.
# So we can just import the path in its entirety.
_singleton path;
# The code path for when isStorePath is false
tryFetchGit =
let
# This imports the files unnecessarily, which currently can't be avoided
# because `builtins.fetchGit` is the only function exposing which files are tracked by Git.
# With the [lazy trees PR](https://github.com/NixOS/nix/pull/6530),
# the unnecessarily import could be avoided.
# However a simpler alternative still would be [a builtins.gitLsFiles](https://github.com/NixOS/nix/issues/2944).
fetchResult = fetchGit ({
url = path;
}
# In older Nix versions, repositories were always assumed to be deep clones, which made `fetchGit` fail for shallow clones
# For newer versions this was fixed, but the `shallow` flag is required.
# The only behavioral difference is that for shallow clones, `fetchGit` doesn't return a `revCount`,
# which we don't need here, so it's fine to always pass it.
# Unfortunately this means older Nix versions get a poor error message for shallow repositories, and there's no good way to improve that.
# Checking for `.git/shallow` doesn't seem worth it, especially since that's more of an implementation detail,
# and would also require more code to handle worktrees where `.git` is a file.
// optionalAttrs (versionAtLeast nixVersion _fetchGitShallowMinver) { shallow = true; }
// extraFetchGitAttrs);
in
# We can identify local working directories by checking for .git,
# see https://git-scm.com/docs/gitrepository-layout#_description.
# Note that `builtins.fetchGit` _does_ work for bare repositories (where there's no `.git`),
# even though `git ls-files` wouldn't return any files in that case.
if ! pathExists (path + "/.git") then
throw "lib.fileset.${function}: Expected the ${argument} (${toString path}) to point to a local working tree of a Git repository, but it's not."
else
_mirrorStorePath path fetchResult.outPath;
in
if ! isPath path then
throw "lib.fileset.${function}: Expected the ${argument} to be a path, but it's a ${typeOf path} instead."
else if pathType path != "directory" then
throw "lib.fileset.${function}: Expected the ${argument} (${toString path}) to be a directory, but it's a file instead."
else if hasStorePathPrefix path then
tryStorePath
else
tryFetchGit;
}

View file

@ -0,0 +1,26 @@
# This overlay implements mocking of the lib.path.splitRoot function
# It pretends that the last component named "mock-root" is the root:
#
# splitRoot /foo/mock-root/bar/mock-root/baz
# => {
# root = /foo/mock-root/bar/mock-root;
# subpath = "./baz";
# }
self: super: {
path = super.path // {
splitRoot = path:
let
parts = super.path.splitRoot path;
components = self.path.subpath.components parts.subpath;
count = self.length components;
rootIndex = count - self.lists.findFirstIndex
(component: component == "mock-root")
(self.length components)
(self.reverseList components);
root = self.path.append parts.root (self.path.subpath.join (self.take rootIndex components));
subpath = self.path.subpath.join (self.drop rootIndex components);
in {
inherit root subpath;
};
};
}

1597
lib/fileset/tests.sh Executable file

File diff suppressed because it is too large Load diff

311
lib/filesystem.nix Normal file
View file

@ -0,0 +1,311 @@
/*
Functions for querying information about the filesystem
without copying any files to the Nix store.
*/
{ lib }:
# Tested in lib/tests/filesystem.sh
let
inherit (builtins)
readDir
pathExists
toString
;
inherit (lib.attrsets)
mapAttrs'
filterAttrs
;
inherit (lib.filesystem)
pathType
;
inherit (lib.strings)
hasSuffix
removeSuffix
;
in
{
/*
The type of a path. The path needs to exist and be accessible.
The result is either "directory" for a directory, "regular" for a regular file, "symlink" for a symlink, or "unknown" for anything else.
Type:
pathType :: Path -> String
Example:
pathType /.
=> "directory"
pathType /some/file.nix
=> "regular"
*/
pathType =
builtins.readFileType or
# Nix <2.14 compatibility shim
(path:
if ! pathExists path
# Fail irrecoverably to mimic the historic behavior of this function and
# the new builtins.readFileType
then abort "lib.filesystem.pathType: Path ${toString path} does not exist."
# The filesystem root is the only path where `dirOf / == /` and
# `baseNameOf /` is not valid. We can detect this and directly return
# "directory", since we know the filesystem root can't be anything else.
else if dirOf path == path
then "directory"
else (readDir (dirOf path)).${baseNameOf path}
);
/*
Whether a path exists and is a directory.
Type:
pathIsDirectory :: Path -> Bool
Example:
pathIsDirectory /.
=> true
pathIsDirectory /this/does/not/exist
=> false
pathIsDirectory /some/file.nix
=> false
*/
pathIsDirectory = path:
pathExists path && pathType path == "directory";
/*
Whether a path exists and is a regular file, meaning not a symlink or any other special file type.
Type:
pathIsRegularFile :: Path -> Bool
Example:
pathIsRegularFile /.
=> false
pathIsRegularFile /this/does/not/exist
=> false
pathIsRegularFile /some/file.nix
=> true
*/
pathIsRegularFile = path:
pathExists path && pathType path == "regular";
/*
A map of all haskell packages defined in the given path,
identified by having a cabal file with the same name as the
directory itself.
Type: Path -> Map String Path
*/
haskellPathsInDir =
# The directory within to search
root:
let # Files in the root
root-files = builtins.attrNames (builtins.readDir root);
# Files with their full paths
root-files-with-paths =
map (file:
{ name = file; value = root + "/${file}"; }
) root-files;
# Subdirectories of the root with a cabal file.
cabal-subdirs =
builtins.filter ({ name, value }:
builtins.pathExists (value + "/${name}.cabal")
) root-files-with-paths;
in builtins.listToAttrs cabal-subdirs;
/*
Find the first directory containing a file matching 'pattern'
upward from a given 'file'.
Returns 'null' if no directories contain a file matching 'pattern'.
Type: RegExp -> Path -> Nullable { path : Path; matches : [ MatchResults ]; }
*/
locateDominatingFile =
# The pattern to search for
pattern:
# The file to start searching upward from
file:
let go = path:
let files = builtins.attrNames (builtins.readDir path);
matches = builtins.filter (match: match != null)
(map (builtins.match pattern) files);
in
if builtins.length matches != 0
then { inherit path matches; }
else if path == /.
then null
else go (dirOf path);
parent = dirOf file;
isDir =
let base = baseNameOf file;
type = (builtins.readDir parent).${base} or null;
in file == /. || type == "directory";
in go (if isDir then file else parent);
/*
Given a directory, return a flattened list of all files within it recursively.
Type: Path -> [ Path ]
*/
listFilesRecursive =
# The path to recursively list
dir:
lib.flatten (lib.mapAttrsToList (name: type:
if type == "directory" then
lib.filesystem.listFilesRecursive (dir + "/${name}")
else
dir + "/${name}"
) (builtins.readDir dir));
/*
Transform a directory tree containing package files suitable for
`callPackage` into a matching nested attribute set of derivations.
For a directory tree like this:
```
my-packages
a.nix
b.nix
c
my-extra-feature.patch
package.nix
support-definitions.nix
my-namespace
d.nix
e.nix
f
package.nix
```
`packagesFromDirectoryRecursive` will produce an attribute set like this:
```nix
# packagesFromDirectoryRecursive {
# callPackage = pkgs.callPackage;
# directory = ./my-packages;
# }
{
a = pkgs.callPackage ./my-packages/a.nix { };
b = pkgs.callPackage ./my-packages/b.nix { };
c = pkgs.callPackage ./my-packages/c/package.nix { };
my-namespace = {
d = pkgs.callPackage ./my-packages/my-namespace/d.nix { };
e = pkgs.callPackage ./my-packages/my-namespace/e.nix { };
f = pkgs.callPackage ./my-packages/my-namespace/f/package.nix { };
};
}
```
In particular:
- If the input directory contains a `package.nix` file, then
`callPackage <directory>/package.nix { }` is returned.
- Otherwise, the input directory's contents are listed and transformed into
an attribute set.
- If a file name has the `.nix` extension, it is turned into attribute
where:
- The attribute name is the file name without the `.nix` extension
- The attribute value is `callPackage <file path> { }`
- Other files are ignored.
- Directories are turned into an attribute where:
- The attribute name is the name of the directory
- The attribute value is the result of calling
`packagesFromDirectoryRecursive { ... }` on the directory.
As a result, directories with no `.nix` files (including empty
directories) will be transformed into empty attribute sets.
Example:
packagesFromDirectoryRecursive {
inherit (pkgs) callPackage;
directory = ./my-packages;
}
=> { ... }
lib.makeScope pkgs.newScope (
self: packagesFromDirectoryRecursive {
callPackage = self.callPackage;
directory = ./my-packages;
}
)
=> { ... }
Type:
packagesFromDirectoryRecursive :: AttrSet -> AttrSet
*/
packagesFromDirectoryRecursive =
# Options.
{
/*
`pkgs.callPackage`
Type:
Path -> AttrSet -> a
*/
callPackage,
/*
The directory to read package files from
Type:
Path
*/
directory,
...
}:
let
# Determine if a directory entry from `readDir` indicates a package or
# directory of packages.
directoryEntryIsPackage = basename: type:
type == "directory" || hasSuffix ".nix" basename;
# List directory entries that indicate packages in the given `path`.
packageDirectoryEntries = path:
filterAttrs directoryEntryIsPackage (readDir path);
# Transform a directory entry (a `basename` and `type` pair) into a
# package.
directoryEntryToAttrPair = subdirectory: basename: type:
let
path = subdirectory + "/${basename}";
in
if type == "regular"
then
{
name = removeSuffix ".nix" basename;
value = callPackage path { };
}
else
if type == "directory"
then
{
name = basename;
value = packagesFromDirectory path;
}
else
throw
''
lib.filesystem.packagesFromDirectoryRecursive: Unsupported file type ${type} at path ${toString subdirectory}
'';
# Transform a directory into a package (if there's a `package.nix`) or
# set of packages (otherwise).
packagesFromDirectory = path:
let
defaultPackagePath = path + "/package.nix";
in
if pathExists defaultPackagePath
then callPackage defaultPackagePath { }
else mapAttrs'
(directoryEntryToAttrPair path)
(packageDirectoryEntries path);
in
packagesFromDirectory directory;
}

309
lib/fixed-points.nix Normal file
View file

@ -0,0 +1,309 @@
{ lib, ... }:
rec {
/*
`fix f` computes the fixed point of the given function `f`. In other words, the return value is `x` in `x = f x`.
`f` must be a lazy function.
This means that `x` must be a value that can be partially evaluated,
such as an attribute set, a list, or a function.
This way, `f` can use one part of `x` to compute another part.
**Relation to syntactic recursion**
This section explains `fix` by refactoring from syntactic recursion to a call of `fix` instead.
For context, Nix lets you define attributes in terms of other attributes syntactically using the [`rec { }` syntax](https://nixos.org/manual/nix/stable/language/constructs.html#recursive-sets).
```nix
nix-repl> rec {
foo = "foo";
bar = "bar";
foobar = foo + bar;
}
{ bar = "bar"; foo = "foo"; foobar = "foobar"; }
```
This is convenient when constructing a value to pass to a function for example,
but an equivalent effect can be achieved with the `let` binding syntax:
```nix
nix-repl> let self = {
foo = "foo";
bar = "bar";
foobar = self.foo + self.bar;
}; in self
{ bar = "bar"; foo = "foo"; foobar = "foobar"; }
```
But in general you can get more reuse out of `let` bindings by refactoring them to a function.
```nix
nix-repl> f = self: {
foo = "foo";
bar = "bar";
foobar = self.foo + self.bar;
}
```
This is where `fix` comes in, it contains the syntactic recursion that's not in `f` anymore.
```nix
nix-repl> fix = f:
let self = f self; in self;
```
By applying `fix` we get the final result.
```nix
nix-repl> fix f
{ bar = "bar"; foo = "foo"; foobar = "foobar"; }
```
Such a refactored `f` using `fix` is not useful by itself.
See [`extends`](#function-library-lib.fixedPoints.extends) for an example use case.
There `self` is also often called `final`.
Type: fix :: (a -> a) -> a
Example:
fix (self: { foo = "foo"; bar = "bar"; foobar = self.foo + self.bar; })
=> { bar = "bar"; foo = "foo"; foobar = "foobar"; }
fix (self: [ 1 2 (elemAt self 0 + elemAt self 1) ])
=> [ 1 2 3 ]
*/
fix = f: let x = f x; in x;
/*
A variant of `fix` that records the original recursive attribute set in the
result, in an attribute named `__unfix__`.
This is useful in combination with the `extends` function to
implement deep overriding.
*/
fix' = f: let x = f x // { __unfix__ = f; }; in x;
/*
Return the fixpoint that `f` converges to when called iteratively, starting
with the input `x`.
```
nix-repl> converge (x: x / 2) 16
0
```
Type: (a -> a) -> a -> a
*/
converge = f: x:
let
x' = f x;
in
if x' == x
then x
else converge f x';
/*
Extend a function using an overlay.
Overlays allow modifying and extending fixed-point functions, specifically ones returning attribute sets.
A fixed-point function is a function which is intended to be evaluated by passing the result of itself as the argument.
This is possible due to Nix's lazy evaluation.
A fixed-point function returning an attribute set has the form
```nix
final: { # attributes }
```
where `final` refers to the lazily evaluated attribute set returned by the fixed-point function.
An overlay to such a fixed-point function has the form
```nix
final: prev: { # attributes }
```
where `prev` refers to the result of the original function to `final`, and `final` is the result of the composition of the overlay and the original function.
Applying an overlay is done with `extends`:
```nix
let
f = final: { # attributes };
overlay = final: prev: { # attributes };
in extends overlay f;
```
To get the value of `final`, use `lib.fix`:
```nix
let
f = final: { # attributes };
overlay = final: prev: { # attributes };
g = extends overlay f;
in fix g
```
:::{.note}
The argument to the given fixed-point function after applying an overlay will *not* refer to its own return value, but rather to the value after evaluating the overlay function.
The given fixed-point function is called with a separate argument than if it was evaluated with `lib.fix`.
:::
:::{.example}
# Extend a fixed-point function with an overlay
Define a fixed-point function `f` that expects its own output as the argument `final`:
```nix-repl
f = final: {
# Constant value a
a = 1;
# b depends on the final value of a, available as final.a
b = final.a + 2;
}
```
Evaluate this using [`lib.fix`](#function-library-lib.fixedPoints.fix) to get the final result:
```nix-repl
fix f
=> { a = 1; b = 3; }
```
An overlay represents a modification or extension of such a fixed-point function.
Here's an example of an overlay:
```nix-repl
overlay = final: prev: {
# Modify the previous value of a, available as prev.a
a = prev.a + 10;
# Extend the attribute set with c, letting it depend on the final values of a and b
c = final.a + final.b;
}
```
Use `extends overlay f` to apply the overlay to the fixed-point function `f`.
This produces a new fixed-point function `g` with the combined behavior of `f` and `overlay`:
```nix-repl
g = extends overlay f
```
The result is a function, so we can't print it directly, but it's the same as:
```nix-repl
g' = final: {
# The constant from f, but changed with the overlay
a = 1 + 10;
# Unchanged from f
b = final.a + 2;
# Extended in the overlay
c = final.a + final.b;
}
```
Evaluate this using [`lib.fix`](#function-library-lib.fixedPoints.fix) again to get the final result:
```nix-repl
fix g
=> { a = 11; b = 13; c = 24; }
```
:::
Type:
extends :: (Attrs -> Attrs -> Attrs) # The overlay to apply to the fixed-point function
-> (Attrs -> Attrs) # A fixed-point function
-> (Attrs -> Attrs) # The resulting fixed-point function
Example:
f = final: { a = 1; b = final.a + 2; }
fix f
=> { a = 1; b = 3; }
fix (extends (final: prev: { a = prev.a + 10; }) f)
=> { a = 11; b = 13; }
fix (extends (final: prev: { b = final.a + 5; }) f)
=> { a = 1; b = 6; }
fix (extends (final: prev: { c = final.a + final.b; }) f)
=> { a = 1; b = 3; c = 4; }
*/
extends =
# The overlay to apply to the fixed-point function
overlay:
# The fixed-point function
f:
# Wrap with parenthesis to prevent nixdoc from rendering the `final` argument in the documentation
# The result should be thought of as a function, the argument of that function is not an argument to `extends` itself
(
final:
let
prev = f final;
in
prev // overlay final prev
);
/*
Compose two extending functions of the type expected by 'extends'
into one where changes made in the first are available in the
'super' of the second
*/
composeExtensions =
f: g: final: prev:
let fApplied = f final prev;
prev' = prev // fApplied;
in fApplied // g final prev';
/*
Compose several extending functions of the type expected by 'extends' into
one where changes made in preceding functions are made available to
subsequent ones.
```
composeManyExtensions : [packageSet -> packageSet -> packageSet] -> packageSet -> packageSet -> packageSet
^final ^prev ^overrides ^final ^prev ^overrides
```
*/
composeManyExtensions =
lib.foldr (x: y: composeExtensions x y) (final: prev: {});
/*
Create an overridable, recursive attribute set. For example:
```
nix-repl> obj = makeExtensible (self: { })
nix-repl> obj
{ __unfix__ = «lambda»; extend = «lambda»; }
nix-repl> obj = obj.extend (self: super: { foo = "foo"; })
nix-repl> obj
{ __unfix__ = «lambda»; extend = «lambda»; foo = "foo"; }
nix-repl> obj = obj.extend (self: super: { foo = super.foo + " + "; bar = "bar"; foobar = self.foo + self.bar; })
nix-repl> obj
{ __unfix__ = «lambda»; bar = "bar"; extend = «lambda»; foo = "foo + "; foobar = "foo + bar"; }
```
*/
makeExtensible = makeExtensibleWithCustomName "extend";
/*
Same as `makeExtensible` but the name of the extending attribute is
customized.
*/
makeExtensibleWithCustomName = extenderName: rattrs:
fix' (self: (rattrs self) // {
${extenderName} = f: makeExtensibleWithCustomName extenderName (extends f rattrs);
});
}

View file

@ -0,0 +1,20 @@
# This function produces a lib overlay to be used by the nixpkgs
# & nixpkgs/lib flakes to provide meaningful values for
# `lib.trivial.version` et al..
#
# Internal and subject to change, don't use this anywhere else!
# Instead, consider using a public interface, such as this flake here
# in this directory, `lib/`, or use the nixpkgs flake, which applies
# this logic for you in its `lib` output attribute.
self: # from the flake
finalLib: prevLib: # lib overlay
{
trivial = prevLib.trivial // {
versionSuffix =
".${finalLib.substring 0 8 (self.lastModifiedDate or "19700101")}.${self.shortRev or "dirty"}";
revisionWithDefault = default: self.rev or default;
};
}

10
lib/flake.nix Normal file
View file

@ -0,0 +1,10 @@
{
description = "Library of low-level helper functions for nix expressions.";
outputs = { self }:
let
lib0 = import ./.;
in {
lib = lib0.extend (import ./flake-version-info.nix self);
};
}

602
lib/generators.nix Normal file
View file

@ -0,0 +1,602 @@
/* Functions that generate widespread file
* formats from nix data structures.
*
* They all follow a similar interface:
* generator { config-attrs } data
*
* `config-attrs` are holes in the generators
* with sensible default implementations that
* can be overwritten. The default implementations
* are mostly generators themselves, called with
* their respective default values; they can be reused.
*
* Tests can be found in ./tests/misc.nix
* Documentation in the manual, #sec-generators
*/
{ lib }:
let
inherit (lib)
addErrorContext
assertMsg
attrNames
concatLists
concatMapStringsSep
concatStrings
concatStringsSep
const
elem
escape
filter
flatten
foldl
functionArgs # Note: not the builtin; considers `__functor` in attrsets.
gvariant
hasInfix
head
id
init
isAttrs
isBool
isDerivation
isFloat
isFunction # Note: not the builtin; considers `__functor` in attrsets.
isInt
isList
isPath
isString
last
length
mapAttrs
mapAttrsToList
optionals
recursiveUpdate
replaceStrings
reverseList
splitString
tail
toList
;
inherit (lib.strings)
escapeNixIdentifier
floatToString
match
split
toJSON
typeOf
;
## -- HELPER FUNCTIONS & DEFAULTS --
/* Convert a value to a sensible default string representation.
* The builtin `toString` function has some strange defaults,
* suitable for bash scripts but not much else.
*/
mkValueStringDefault = {}: v:
let err = t: v: abort
("generators.mkValueStringDefault: " +
"${t} not supported: ${toPretty {} v}");
in if isInt v then toString v
# convert derivations to store paths
else if isDerivation v then toString v
# we default to not quoting strings
else if isString v then v
# isString returns "1", which is not a good default
else if true == v then "true"
# here it returns to "", which is even less of a good default
else if false == v then "false"
else if null == v then "null"
# if you have lists you probably want to replace this
else if isList v then err "lists" v
# same as for lists, might want to replace
else if isAttrs v then err "attrsets" v
# functions cant be printed of course
else if isFunction v then err "functions" v
# Floats currently can't be converted to precise strings,
# condition warning on nix version once this isn't a problem anymore
# See https://github.com/NixOS/nix/pull/3480
else if isFloat v then floatToString v
else err "this value is" (toString v);
/* Generate a line of key k and value v, separated by
* character sep. If sep appears in k, it is escaped.
* Helper for synaxes with different separators.
*
* mkValueString specifies how values should be formatted.
*
* mkKeyValueDefault {} ":" "f:oo" "bar"
* > "f\:oo:bar"
*/
mkKeyValueDefault = {
mkValueString ? mkValueStringDefault {}
}: sep: k: v:
"${escape [sep] k}${sep}${mkValueString v}";
## -- FILE FORMAT GENERATORS --
/* Generate a key-value-style config file from an attrset.
*
* mkKeyValue is the same as in toINI.
*/
toKeyValue = {
mkKeyValue ? mkKeyValueDefault {} "=",
listsAsDuplicateKeys ? false,
indent ? ""
}:
let mkLine = k: v: indent + mkKeyValue k v + "\n";
mkLines = if listsAsDuplicateKeys
then k: v: map (mkLine k) (if isList v then v else [v])
else k: v: [ (mkLine k v) ];
in attrs: concatStrings (concatLists (mapAttrsToList mkLines attrs));
/* Generate an INI-style config file from an
* attrset of sections to an attrset of key-value pairs.
*
* generators.toINI {} {
* foo = { hi = "${pkgs.hello}"; ciao = "bar"; };
* baz = { "also, integers" = 42; };
* }
*
*> [baz]
*> also, integers=42
*>
*> [foo]
*> ciao=bar
*> hi=/nix/store/y93qql1p5ggfnaqjjqhxcw0vqw95rlz0-hello-2.10
*
* The mk* configuration attributes can generically change
* the way sections and key-value strings are generated.
*
* For more examples see the test cases in ./tests/misc.nix.
*/
toINI = {
# apply transformations (e.g. escapes) to section names
mkSectionName ? (name: escape [ "[" "]" ] name),
# format a setting line from key and value
mkKeyValue ? mkKeyValueDefault {} "=",
# allow lists as values for duplicate keys
listsAsDuplicateKeys ? false
}: attrsOfAttrs:
let
# map function to string for each key val
mapAttrsToStringsSep = sep: mapFn: attrs:
concatStringsSep sep
(mapAttrsToList mapFn attrs);
mkSection = sectName: sectValues: ''
[${mkSectionName sectName}]
'' + toKeyValue { inherit mkKeyValue listsAsDuplicateKeys; } sectValues;
in
# map input to ini sections
mapAttrsToStringsSep "\n" mkSection attrsOfAttrs;
/* Generate an INI-style config file from an attrset
* specifying the global section (no header), and an
* attrset of sections to an attrset of key-value pairs.
*
* generators.toINIWithGlobalSection {} {
* globalSection = {
* someGlobalKey = "hi";
* };
* sections = {
* foo = { hi = "${pkgs.hello}"; ciao = "bar"; };
* baz = { "also, integers" = 42; };
* }
*
*> someGlobalKey=hi
*>
*> [baz]
*> also, integers=42
*>
*> [foo]
*> ciao=bar
*> hi=/nix/store/y93qql1p5ggfnaqjjqhxcw0vqw95rlz0-hello-2.10
*
* The mk* configuration attributes can generically change
* the way sections and key-value strings are generated.
*
* For more examples see the test cases in ./tests/misc.nix.
*
* If you dont need a global section, you can also use
* `generators.toINI` directly, which only takes
* the part in `sections`.
*/
toINIWithGlobalSection = {
# apply transformations (e.g. escapes) to section names
mkSectionName ? (name: escape [ "[" "]" ] name),
# format a setting line from key and value
mkKeyValue ? mkKeyValueDefault {} "=",
# allow lists as values for duplicate keys
listsAsDuplicateKeys ? false
}: { globalSection, sections ? {} }:
( if globalSection == {}
then ""
else (toKeyValue { inherit mkKeyValue listsAsDuplicateKeys; } globalSection)
+ "\n")
+ (toINI { inherit mkSectionName mkKeyValue listsAsDuplicateKeys; } sections);
/* Generate a git-config file from an attrset.
*
* It has two major differences from the regular INI format:
*
* 1. values are indented with tabs
* 2. sections can have sub-sections
*
* generators.toGitINI {
* url."ssh://git@github.com/".insteadOf = "https://github.com";
* user.name = "edolstra";
* }
*
*> [url "ssh://git@github.com/"]
*> insteadOf = "https://github.com"
*>
*> [user]
*> name = "edolstra"
*/
toGitINI = attrs:
let
mkSectionName = name:
let
containsQuote = hasInfix ''"'' name;
sections = splitString "." name;
section = head sections;
subsections = tail sections;
subsection = concatStringsSep "." subsections;
in if containsQuote || subsections == [ ] then
name
else
''${section} "${subsection}"'';
mkValueString = v:
let
escapedV = ''
"${
replaceStrings [ "\n" " " ''"'' "\\" ] [ "\\n" "\\t" ''\"'' "\\\\" ] v
}"'';
in mkValueStringDefault { } (if isString v then escapedV else v);
# generation for multiple ini values
mkKeyValue = k: v:
let mkKeyValue = mkKeyValueDefault { inherit mkValueString; } " = " k;
in concatStringsSep "\n" (map (kv: "\t" + mkKeyValue kv) (toList v));
# converts { a.b.c = 5; } to { "a.b".c = 5; } for toINI
gitFlattenAttrs = let
recurse = path: value:
if isAttrs value && !isDerivation value then
mapAttrsToList (name: value: recurse ([ name ] ++ path) value) value
else if length path > 1 then {
${concatStringsSep "." (reverseList (tail path))}.${head path} = value;
} else {
${head path} = value;
};
in attrs: foldl recursiveUpdate { } (flatten (recurse [ ] attrs));
toINI_ = toINI { inherit mkKeyValue mkSectionName; };
in
toINI_ (gitFlattenAttrs attrs);
# mkKeyValueDefault wrapper that handles dconf INI quirks.
# The main differences of the format is that it requires strings to be quoted.
mkDconfKeyValue = mkKeyValueDefault { mkValueString = v: toString (gvariant.mkValue v); } "=";
# Generates INI in dconf keyfile style. See https://help.gnome.org/admin/system-admin-guide/stable/dconf-keyfiles.html.en
# for details.
toDconfINI = toINI { mkKeyValue = mkDconfKeyValue; };
withRecursion =
{
/* If this option is not null, the given value will stop evaluating at a certain depth */
depthLimit
/* If this option is true, an error will be thrown, if a certain given depth is exceeded */
, throwOnDepthLimit ? true
}:
assert isInt depthLimit;
let
specialAttrs = [
"__functor"
"__functionArgs"
"__toString"
"__pretty"
];
stepIntoAttr = evalNext: name:
if elem name specialAttrs
then id
else evalNext;
transform = depth:
if depthLimit != null && depth > depthLimit then
if throwOnDepthLimit
then throw "Exceeded maximum eval-depth limit of ${toString depthLimit} while trying to evaluate with `generators.withRecursion'!"
else const "<unevaluated>"
else id;
mapAny = depth: v:
let
evalNext = x: mapAny (depth + 1) (transform (depth + 1) x);
in
if isAttrs v then mapAttrs (stepIntoAttr evalNext) v
else if isList v then map evalNext v
else transform (depth + 1) v;
in
mapAny 0;
/* Pretty print a value, akin to `builtins.trace`.
* Should probably be a builtin as well.
* The pretty-printed string should be suitable for rendering default values
* in the NixOS manual. In particular, it should be as close to a valid Nix expression
* as possible.
*/
toPretty = {
/* If this option is true, attrsets like { __pretty = fn; val = ; }
will use fn to convert val to a pretty printed representation.
(This means fn is type Val -> String.) */
allowPrettyValues ? false,
/* If this option is true, the output is indented with newlines for attribute sets and lists */
multiline ? true,
/* Initial indentation level */
indent ? ""
}:
let
go = indent: v:
let introSpace = if multiline then "\n${indent} " else " ";
outroSpace = if multiline then "\n${indent}" else " ";
in if isInt v then toString v
# toString loses precision on floats, so we use toJSON instead. This isn't perfect
# as the resulting string may not parse back as a float (e.g. 42, 1e-06), but for
# pretty-printing purposes this is acceptable.
else if isFloat v then builtins.toJSON v
else if isString v then
let
lines = filter (v: ! isList v) (split "\n" v);
escapeSingleline = escape [ "\\" "\"" "\${" ];
escapeMultiline = replaceStrings [ "\${" "''" ] [ "''\${" "'''" ];
singlelineResult = "\"" + concatStringsSep "\\n" (map escapeSingleline lines) + "\"";
multilineResult = let
escapedLines = map escapeMultiline lines;
# The last line gets a special treatment: if it's empty, '' is on its own line at the "outer"
# indentation level. Otherwise, '' is appended to the last line.
lastLine = last escapedLines;
in "''" + introSpace + concatStringsSep introSpace (init escapedLines)
+ (if lastLine == "" then outroSpace else introSpace + lastLine) + "''";
in
if multiline && length lines > 1 then multilineResult else singlelineResult
else if true == v then "true"
else if false == v then "false"
else if null == v then "null"
else if isPath v then toString v
else if isList v then
if v == [] then "[ ]"
else "[" + introSpace
+ concatMapStringsSep introSpace (go (indent + " ")) v
+ outroSpace + "]"
else if isFunction v then
let fna = functionArgs v;
showFnas = concatStringsSep ", " (mapAttrsToList
(name: hasDefVal: if hasDefVal then name + "?" else name)
fna);
in if fna == {} then "<function>"
else "<function, args: {${showFnas}}>"
else if isAttrs v then
# apply pretty values if allowed
if allowPrettyValues && v ? __pretty && v ? val
then v.__pretty v.val
else if v == {} then "{ }"
else if v ? type && v.type == "derivation" then
"<derivation ${v.name or "???"}>"
else "{" + introSpace
+ concatStringsSep introSpace (mapAttrsToList
(name: value:
"${escapeNixIdentifier name} = ${
addErrorContext "while evaluating an attribute `${name}`"
(go (indent + " ") value)
};") v)
+ outroSpace + "}"
else abort "generators.toPretty: should never happen (v = ${v})";
in go indent;
# PLIST handling
toPlist = {}: v: let
expr = ind: x:
if x == null then "" else
if isBool x then bool ind x else
if isInt x then int ind x else
if isString x then str ind x else
if isList x then list ind x else
if isAttrs x then attrs ind x else
if isPath x then str ind (toString x) else
if isFloat x then float ind x else
abort "generators.toPlist: should never happen (v = ${v})";
literal = ind: x: ind + x;
bool = ind: x: literal ind (if x then "<true/>" else "<false/>");
int = ind: x: literal ind "<integer>${toString x}</integer>";
str = ind: x: literal ind "<string>${x}</string>";
key = ind: x: literal ind "<key>${x}</key>";
float = ind: x: literal ind "<real>${toString x}</real>";
indent = ind: expr "\t${ind}";
item = ind: concatMapStringsSep "\n" (indent ind);
list = ind: x: concatStringsSep "\n" [
(literal ind "<array>")
(item ind x)
(literal ind "</array>")
];
attrs = ind: x: concatStringsSep "\n" [
(literal ind "<dict>")
(attr ind x)
(literal ind "</dict>")
];
attr = let attrFilter = name: value: name != "_module" && value != null;
in ind: x: concatStringsSep "\n" (flatten (mapAttrsToList
(name: value: optionals (attrFilter name value) [
(key "\t${ind}" name)
(expr "\t${ind}" value)
]) x));
in ''<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple Computer//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
${expr "" v}
</plist>'';
/* Translate a simple Nix expression to Dhall notation.
* Note that integers are translated to Integer and never
* the Natural type.
*/
toDhall = { }@args: v:
let concatItems = concatStringsSep ", ";
in if isAttrs v then
"{ ${
concatItems (mapAttrsToList
(key: value: "${key} = ${toDhall args value}") v)
} }"
else if isList v then
"[ ${concatItems (map (toDhall args) v)} ]"
else if isInt v then
"${if v < 0 then "" else "+"}${toString v}"
else if isBool v then
(if v then "True" else "False")
else if isFunction v then
abort "generators.toDhall: cannot convert a function to Dhall"
else if v == null then
abort "generators.toDhall: cannot convert a null to Dhall"
else
toJSON v;
/*
Translate a simple Nix expression to Lua representation with occasional
Lua-inlines that can be constructed by mkLuaInline function.
Configuration:
* multiline - by default is true which results in indented block-like view.
* indent - initial indent.
* asBindings - by default generate single value, but with this use attrset to set global vars.
Attention:
Regardless of multiline parameter there is no trailing newline.
Example:
generators.toLua {}
{
cmd = [ "typescript-language-server" "--stdio" ];
settings.workspace.library = mkLuaInline ''vim.api.nvim_get_runtime_file("", true)'';
}
->
{
["cmd"] = {
"typescript-language-server",
"--stdio"
},
["settings"] = {
["workspace"] = {
["library"] = (vim.api.nvim_get_runtime_file("", true))
}
}
}
Type:
toLua :: AttrSet -> Any -> String
*/
toLua = {
/* If this option is true, the output is indented with newlines for attribute sets and lists */
multiline ? true,
/* Initial indentation level */
indent ? "",
/* Interpret as variable bindings */
asBindings ? false,
}@args: v:
let
innerIndent = "${indent} ";
introSpace = if multiline then "\n${innerIndent}" else " ";
outroSpace = if multiline then "\n${indent}" else " ";
innerArgs = args // {
indent = if asBindings then indent else innerIndent;
asBindings = false;
};
concatItems = concatStringsSep ",${introSpace}";
isLuaInline = { _type ? null, ... }: _type == "lua-inline";
generatedBindings =
assert assertMsg (badVarNames == []) "Bad Lua var names: ${toPretty {} badVarNames}";
concatStrings (
mapAttrsToList (key: value: "${indent}${key} = ${toLua innerArgs value}\n") v
);
# https://en.wikibooks.org/wiki/Lua_Programming/variable#Variable_names
matchVarName = match "[[:alpha:]_][[:alnum:]_]*(\\.[[:alpha:]_][[:alnum:]_]*)*";
badVarNames = filter (name: matchVarName name == null) (attrNames v);
in
if asBindings then
generatedBindings
else if v == null then
"nil"
else if isInt v || isFloat v || isString v || isBool v then
toJSON v
else if isList v then
(if v == [ ] then "{}" else
"{${introSpace}${concatItems (map (value: "${toLua innerArgs value}") v)}${outroSpace}}")
else if isAttrs v then
(
if isLuaInline v then
"(${v.expr})"
else if v == { } then
"{}"
else if isDerivation v then
''"${toString v}"''
else
"{${introSpace}${concatItems (
mapAttrsToList (key: value: "[${toJSON key}] = ${toLua innerArgs value}") v
)}${outroSpace}}"
)
else
abort "generators.toLua: type ${typeOf v} is unsupported";
/*
Mark string as Lua expression to be inlined when processed by toLua.
Type:
mkLuaInline :: String -> AttrSet
*/
mkLuaInline = expr: { _type = "lua-inline"; inherit expr; };
in
# Everything in this attrset is the public interface of the file.
{
inherit
mkDconfKeyValue
mkKeyValueDefault
mkLuaInline
mkValueStringDefault
toDconfINI
toDhall
toGitINI
toINI
toINIWithGlobalSection
toKeyValue
toLua
toPlist
toPretty
withRecursion
;
/* Generates JSON from an arbitrary (non-function) value.
* For more information see the documentation of the builtin.
*/
toJSON = {}: toJSON;
/* YAML has been a strict superset of JSON since 1.2, so we
* use toJSON. Before it only had a few differences referring
* to implicit typing rules, so it should work with older
* parsers as well.
*/
toYAML = {}: toJSON;
}

291
lib/gvariant.nix Normal file
View file

@ -0,0 +1,291 @@
/*
A partial and basic implementation of GVariant formatted strings.
See [GVariant Format Strings](https://docs.gtk.org/glib/gvariant-format-strings.html) for details.
:::{.warning}
This API is not considered fully stable and it might therefore
change in backwards incompatible ways without prior notice.
:::
*/
# This file is based on https://github.com/nix-community/home-manager
# Copyright (c) 2017-2022 Home Manager contributors
{ lib }:
let
inherit (lib)
concatMapStringsSep concatStrings escape head replaceStrings;
mkPrimitive = t: v: {
_type = "gvariant";
type = t;
value = v;
__toString = self: "@${self.type} ${toString self.value}"; # https://docs.gtk.org/glib/gvariant-text.html
};
type = {
arrayOf = t: "a${t}";
maybeOf = t: "m${t}";
tupleOf = ts: "(${concatStrings ts})";
dictionaryEntryOf = nameType: valueType: "{${nameType}${valueType}}";
string = "s";
boolean = "b";
uchar = "y";
int16 = "n";
uint16 = "q";
int32 = "i";
uint32 = "u";
int64 = "x";
uint64 = "t";
double = "d";
variant = "v";
};
/* Check if a value is a GVariant value
Type:
isGVariant :: Any -> Bool
*/
isGVariant = v: v._type or "" == "gvariant";
in
rec {
inherit type isGVariant;
/* Returns the GVariant value that most closely matches the given Nix value.
If no GVariant value can be found unambiguously then error is thrown.
Type:
mkValue :: Any -> gvariant
*/
mkValue = v:
if builtins.isBool v then
mkBoolean v
else if builtins.isFloat v then
mkDouble v
else if builtins.isString v then
mkString v
else if builtins.isList v then
mkArray v
else if isGVariant v then
v
else
throw "The GVariant type of ${v} can't be inferred.";
/* Returns the GVariant array from the given type of the elements and a Nix list.
Type:
mkArray :: [Any] -> gvariant
Example:
# Creating a string array
lib.gvariant.mkArray [ "a" "b" "c" ]
*/
mkArray = elems:
let
vs = map mkValue (lib.throwIf (elems == [ ]) "Please create empty array with mkEmptyArray." elems);
elemType = lib.throwIfNot (lib.all (t: (head vs).type == t) (map (v: v.type) vs))
"Elements in a list should have same type."
(head vs).type;
in
mkPrimitive (type.arrayOf elemType) vs // {
__toString = self:
"@${self.type} [${concatMapStringsSep "," toString self.value}]";
};
/* Returns the GVariant array from the given empty Nix list.
Type:
mkEmptyArray :: gvariant.type -> gvariant
Example:
# Creating an empty string array
lib.gvariant.mkEmptyArray (lib.gvariant.type.string)
*/
mkEmptyArray = elemType: mkPrimitive (type.arrayOf elemType) [ ] // {
__toString = self: "@${self.type} []";
};
/* Returns the GVariant variant from the given Nix value. Variants are containers
of different GVariant type.
Type:
mkVariant :: Any -> gvariant
Example:
lib.gvariant.mkArray [
(lib.gvariant.mkVariant "a string")
(lib.gvariant.mkVariant (lib.gvariant.mkInt32 1))
]
*/
mkVariant = elem:
let gvarElem = mkValue elem;
in mkPrimitive type.variant gvarElem // {
__toString = self: "<${toString self.value}>";
};
/* Returns the GVariant dictionary entry from the given key and value.
Type:
mkDictionaryEntry :: String -> Any -> gvariant
Example:
# A dictionary describing an Epiphanys search provider
[
(lib.gvariant.mkDictionaryEntry "url" (lib.gvariant.mkVariant "https://duckduckgo.com/?q=%s&t=epiphany"))
(lib.gvariant.mkDictionaryEntry "bang" (lib.gvariant.mkVariant "!d"))
(lib.gvariant.mkDictionaryEntry "name" (lib.gvariant.mkVariant "DuckDuckGo"))
]
*/
mkDictionaryEntry =
# The key of the entry
name:
# The value of the entry
value:
let
name' = mkValue name;
value' = mkValue value;
dictionaryType = type.dictionaryEntryOf name'.type value'.type;
in
mkPrimitive dictionaryType { inherit name value; } // {
__toString = self: "@${self.type} {${name'},${value'}}";
};
/* Returns the GVariant maybe from the given element type.
Type:
mkMaybe :: gvariant.type -> Any -> gvariant
*/
mkMaybe = elemType: elem:
mkPrimitive (type.maybeOf elemType) elem // {
__toString = self:
if self.value == null then
"@${self.type} nothing"
else
"just ${toString self.value}";
};
/* Returns the GVariant nothing from the given element type.
Type:
mkNothing :: gvariant.type -> gvariant
*/
mkNothing = elemType: mkMaybe elemType null;
/* Returns the GVariant just from the given Nix value.
Type:
mkJust :: Any -> gvariant
*/
mkJust = elem: let gvarElem = mkValue elem; in mkMaybe gvarElem.type gvarElem;
/* Returns the GVariant tuple from the given Nix list.
Type:
mkTuple :: [Any] -> gvariant
*/
mkTuple = elems:
let
gvarElems = map mkValue elems;
tupleType = type.tupleOf (map (e: e.type) gvarElems);
in
mkPrimitive tupleType gvarElems // {
__toString = self:
"@${self.type} (${concatMapStringsSep "," toString self.value})";
};
/* Returns the GVariant boolean from the given Nix bool value.
Type:
mkBoolean :: Bool -> gvariant
*/
mkBoolean = v:
mkPrimitive type.boolean v // {
__toString = self: if self.value then "true" else "false";
};
/* Returns the GVariant string from the given Nix string value.
Type:
mkString :: String -> gvariant
*/
mkString = v:
let sanitize = s: replaceStrings [ "\n" ] [ "\\n" ] (escape [ "'" "\\" ] s);
in mkPrimitive type.string v // {
__toString = self: "'${sanitize self.value}'";
};
/* Returns the GVariant object path from the given Nix string value.
Type:
mkObjectpath :: String -> gvariant
*/
mkObjectpath = v:
mkPrimitive type.string v // {
__toString = self: "objectpath '${escape [ "'" ] self.value}'";
};
/* Returns the GVariant uchar from the given Nix int value.
Type:
mkUchar :: Int -> gvariant
*/
mkUchar = mkPrimitive type.uchar;
/* Returns the GVariant int16 from the given Nix int value.
Type:
mkInt16 :: Int -> gvariant
*/
mkInt16 = mkPrimitive type.int16;
/* Returns the GVariant uint16 from the given Nix int value.
Type:
mkUint16 :: Int -> gvariant
*/
mkUint16 = mkPrimitive type.uint16;
/* Returns the GVariant int32 from the given Nix int value.
Type:
mkInt32 :: Int -> gvariant
*/
mkInt32 = v:
mkPrimitive type.int32 v // {
__toString = self: toString self.value;
};
/* Returns the GVariant uint32 from the given Nix int value.
Type:
mkUint32 :: Int -> gvariant
*/
mkUint32 = mkPrimitive type.uint32;
/* Returns the GVariant int64 from the given Nix int value.
Type:
mkInt64 :: Int -> gvariant
*/
mkInt64 = mkPrimitive type.int64;
/* Returns the GVariant uint64 from the given Nix int value.
Type:
mkUint64 :: Int -> gvariant
*/
mkUint64 = mkPrimitive type.uint64;
/* Returns the GVariant double from the given Nix float value.
Type:
mkDouble :: Float -> gvariant
*/
mkDouble = v:
mkPrimitive type.double v // {
__toString = self: toString self.value;
};
}

28
lib/kernel.nix Normal file
View file

@ -0,0 +1,28 @@
{ lib }:
let
inherit (lib) mkIf versionAtLeast versionOlder;
in
{
# Keeping these around in case we decide to change this horrible implementation :)
option = x:
x // { optional = true; };
yes = { tristate = "y"; optional = false; };
no = { tristate = "n"; optional = false; };
module = { tristate = "m"; optional = false; };
unset = { tristate = null; optional = false; };
freeform = x: { freeform = x; optional = false; };
# Common patterns/legacy used in common-config/hardened/config.nix
whenHelpers = version: {
whenAtLeast = ver: mkIf (versionAtLeast version ver);
whenOlder = ver: mkIf (versionOlder version ver);
# range is (inclusive, exclusive)
whenBetween = verLow: verHigh: mkIf (versionAtLeast version verLow && versionOlder version verHigh);
};
}

1312
lib/licenses.nix Normal file

File diff suppressed because it is too large Load diff

1901
lib/lists.nix Normal file

File diff suppressed because it is too large Load diff

189
lib/meta.nix Normal file
View file

@ -0,0 +1,189 @@
/* Some functions for manipulating meta attributes, as well as the
name attribute. */
{ lib }:
let
inherit (lib) matchAttrs any all isDerivation getBin assertMsg;
inherit (builtins) isString match typeOf;
in
rec {
/* Add to or override the meta attributes of the given
derivation.
Example:
addMetaAttrs {description = "Bla blah";} somePkg
*/
addMetaAttrs = newAttrs: drv:
drv // { meta = (drv.meta or {}) // newAttrs; };
/* Disable Hydra builds of given derivation.
*/
dontDistribute = drv: addMetaAttrs { hydraPlatforms = []; } drv;
/* Change the symbolic name of a package for presentation purposes
(i.e., so that nix-env users can tell them apart).
*/
setName = name: drv: drv // {inherit name;};
/* Like `setName`, but takes the previous name as an argument.
Example:
updateName (oldName: oldName + "-experimental") somePkg
*/
updateName = updater: drv: drv // {name = updater (drv.name);};
/* Append a suffix to the name of a package (before the version
part). */
appendToName = suffix: updateName (name:
let x = builtins.parseDrvName name; in "${x.name}-${suffix}-${x.version}");
/* Apply a function to each derivation and only to derivations in an attrset.
*/
mapDerivationAttrset = f: set: lib.mapAttrs (name: pkg: if lib.isDerivation pkg then (f pkg) else pkg) set;
/* Set the nix-env priority of the package.
*/
setPrio = priority: addMetaAttrs { inherit priority; };
/* Decrease the nix-env priority of the package, i.e., other
versions/variants of the package will be preferred.
*/
lowPrio = setPrio 10;
/* Apply lowPrio to an attrset with derivations
*/
lowPrioSet = set: mapDerivationAttrset lowPrio set;
/* Increase the nix-env priority of the package, i.e., this
version/variant of the package will be preferred.
*/
hiPrio = setPrio (-10);
/* Apply hiPrio to an attrset with derivations
*/
hiPrioSet = set: mapDerivationAttrset hiPrio set;
/* Check to see if a platform is matched by the given `meta.platforms`
element.
A `meta.platform` pattern is either
1. (legacy) a system string.
2. (modern) a pattern for the entire platform structure (see `lib.systems.inspect.platformPatterns`).
3. (modern) a pattern for the platform `parsed` field (see `lib.systems.inspect.patterns`).
We can inject these into a pattern for the whole of a structured platform,
and then match that.
Example:
lib.meta.platformMatch { system = "aarch64-darwin"; } "aarch64-darwin"
=> true
*/
platformMatch = platform: elem: (
# Check with simple string comparison if elem was a string.
#
# The majority of comparisons done with this function will be against meta.platforms
# which contains a simple platform string.
#
# Avoiding an attrset allocation results in significant performance gains (~2-30) across the board in OfBorg
# because this is a hot path for nixpkgs.
if isString elem then platform ? system && elem == platform.system
else matchAttrs (
# Normalize platform attrset.
if elem ? parsed then elem
else { parsed = elem; }
) platform
);
/* Check if a package is available on a given platform.
A package is available on a platform if both
1. One of `meta.platforms` pattern matches the given
platform, or `meta.platforms` is not present.
2. None of `meta.badPlatforms` pattern matches the given platform.
Example:
lib.meta.availableOn { system = "aarch64-darwin"; } pkg.zsh
=> true
*/
availableOn = platform: pkg:
((!pkg?meta.platforms) || any (platformMatch platform) pkg.meta.platforms) &&
all (elem: !platformMatch platform elem) (pkg.meta.badPlatforms or []);
/* Get the corresponding attribute in lib.licenses
from the SPDX ID.
For SPDX IDs, see
https://spdx.org/licenses
Type:
getLicenseFromSpdxId :: str -> AttrSet
Example:
lib.getLicenseFromSpdxId "MIT" == lib.licenses.mit
=> true
lib.getLicenseFromSpdxId "mIt" == lib.licenses.mit
=> true
lib.getLicenseFromSpdxId "MY LICENSE"
=> trace: warning: getLicenseFromSpdxId: No license matches the given SPDX ID: MY LICENSE
=> { shortName = "MY LICENSE"; }
*/
getLicenseFromSpdxId =
let
spdxLicenses = lib.mapAttrs (id: ls: assert lib.length ls == 1; builtins.head ls)
(lib.groupBy (l: lib.toLower l.spdxId) (lib.filter (l: l ? spdxId) (lib.attrValues lib.licenses)));
in licstr:
spdxLicenses.${ lib.toLower licstr } or (
lib.warn "getLicenseFromSpdxId: No license matches the given SPDX ID: ${licstr}"
{ shortName = licstr; }
);
/* Get the path to the main program of a package based on meta.mainProgram
Type: getExe :: package -> string
Example:
getExe pkgs.hello
=> "/nix/store/g124820p9hlv4lj8qplzxw1c44dxaw1k-hello-2.12/bin/hello"
getExe pkgs.mustache-go
=> "/nix/store/am9ml4f4ywvivxnkiaqwr0hyxka1xjsf-mustache-go-1.3.0/bin/mustache"
*/
getExe = x: getExe' x (x.meta.mainProgram or (
# This could be turned into an error when 23.05 is at end of life
lib.warn "getExe: Package ${lib.strings.escapeNixIdentifier x.meta.name or x.pname or x.name} does not have the meta.mainProgram attribute. We'll assume that the main program has the same name for now, but this behavior is deprecated, because it leads to surprising errors when the assumption does not hold. If the package has a main program, please set `meta.mainProgram` in its definition to make this warning go away. Otherwise, if the package does not have a main program, or if you don't control its definition, use getExe' to specify the name to the program, such as lib.getExe' foo \"bar\"."
lib.getName
x
));
/* Get the path of a program of a derivation.
Type: getExe' :: derivation -> string -> string
Example:
getExe' pkgs.hello "hello"
=> "/nix/store/g124820p9hlv4lj8qplzxw1c44dxaw1k-hello-2.12/bin/hello"
getExe' pkgs.imagemagick "convert"
=> "/nix/store/5rs48jamq7k6sal98ymj9l4k2bnwq515-imagemagick-7.1.1-15/bin/convert"
*/
getExe' = x: y:
assert assertMsg (isDerivation x)
"lib.meta.getExe': The first argument is of type ${typeOf x}, but it should be a derivation instead.";
assert assertMsg (isString y)
"lib.meta.getExe': The second argument is of type ${typeOf y}, but it should be a string instead.";
assert assertMsg (match ".*\/.*" y == null)
"lib.meta.getExe': The second argument \"${y}\" is a nested path with a \"/\" character, but it should just be the name of the executable instead.";
"${getBin x}/bin/${y}";
}

2
lib/minver.nix Normal file
View file

@ -0,0 +1,2 @@
# Expose the minimum required version for evaluating Nixpkgs
"2.3"

1438
lib/modules.nix Normal file

File diff suppressed because it is too large Load diff

473
lib/options.nix Normal file
View file

@ -0,0 +1,473 @@
/* Nixpkgs/NixOS option handling. */
{ lib }:
let
inherit (lib)
all
collect
concatLists
concatMap
concatMapStringsSep
filter
foldl'
head
tail
isAttrs
isBool
isDerivation
isFunction
isInt
isList
isString
length
mapAttrs
optional
optionals
take
;
inherit (lib.attrsets)
attrByPath
optionalAttrs
;
inherit (lib.strings)
concatMapStrings
concatStringsSep
;
inherit (lib.types)
mkOptionType
;
inherit (lib.lists)
last
;
prioritySuggestion = ''
Use `lib.mkForce value` or `lib.mkDefault value` to change the priority on any of these definitions.
'';
in
rec {
/* Returns true when the given argument is an option
Type: isOption :: a -> bool
Example:
isOption 1 // => false
isOption (mkOption {}) // => true
*/
isOption = lib.isType "option";
/* Creates an Option attribute set. mkOption accepts an attribute set with the following keys:
All keys default to `null` when not given.
Example:
mkOption { } // => { _type = "option"; }
mkOption { default = "foo"; } // => { _type = "option"; default = "foo"; }
*/
mkOption =
{
# Default value used when no definition is given in the configuration.
default ? null,
# Textual representation of the default, for the manual.
defaultText ? null,
# Example value used in the manual.
example ? null,
# String describing the option.
description ? null,
# Related packages used in the manual (see `genRelatedPackages` in ../nixos/lib/make-options-doc/default.nix).
relatedPackages ? null,
# Option type, providing type-checking and value merging.
type ? null,
# Function that converts the option value to something else.
apply ? null,
# Whether the option is for NixOS developers only.
internal ? null,
# Whether the option shows up in the manual. Default: true. Use false to hide the option and any sub-options from submodules. Use "shallow" to hide only sub-options.
visible ? null,
# Whether the option can be set only once
readOnly ? null,
} @ attrs:
attrs // { _type = "option"; };
/* Creates an Option attribute set for a boolean value option i.e an
option to be toggled on or off:
Example:
mkEnableOption "foo"
=> { _type = "option"; default = false; description = "Whether to enable foo."; example = true; type = { ... }; }
*/
mkEnableOption =
# Name for the created option
name: mkOption {
default = false;
example = true;
description = "Whether to enable ${name}.";
type = lib.types.bool;
};
/* Creates an Option attribute set for an option that specifies the
package a module should use for some purpose.
The package is specified in the third argument under `default` as a list of strings
representing its attribute path in nixpkgs (or another package set).
Because of this, you need to pass nixpkgs itself (usually `pkgs` in a module;
alternatively to nixpkgs itself, another package set) as the first argument.
If you pass another package set you should set the `pkgsText` option.
This option is used to display the expression for the package set. It is `"pkgs"` by default.
If your expression is complex you should parenthesize it, as the `pkgsText` argument
is usually immediately followed by an attribute lookup (`.`).
The second argument may be either a string or a list of strings.
It provides the display name of the package in the description of the generated option
(using only the last element if the passed value is a list)
and serves as the fallback value for the `default` argument.
To include extra information in the description, pass `extraDescription` to
append arbitrary text to the generated description.
You can also pass an `example` value, either a literal string or an attribute path.
The `default` argument can be omitted if the provided name is
an attribute of pkgs (if `name` is a string) or a valid attribute path in pkgs (if `name` is a list).
You can also set `default` to just a string in which case it is interpreted as an attribute name
(a singleton attribute path, if you will).
If you wish to explicitly provide no default, pass `null` as `default`.
If you want users to be able to set no package, pass `nullable = true`.
In this mode a `default = null` will not be interpreted as no default and is interpreted literally.
Type: mkPackageOption :: pkgs -> (string|[string]) -> { nullable? :: bool, default? :: string|[string], example? :: null|string|[string], extraDescription? :: string, pkgsText? :: string } -> option
Example:
mkPackageOption pkgs "hello" { }
=> { ...; default = pkgs.hello; defaultText = literalExpression "pkgs.hello"; description = "The hello package to use."; type = package; }
Example:
mkPackageOption pkgs "GHC" {
default = [ "ghc" ];
example = "pkgs.haskell.packages.ghc92.ghc.withPackages (hkgs: [ hkgs.primes ])";
}
=> { ...; default = pkgs.ghc; defaultText = literalExpression "pkgs.ghc"; description = "The GHC package to use."; example = literalExpression "pkgs.haskell.packages.ghc92.ghc.withPackages (hkgs: [ hkgs.primes ])"; type = package; }
Example:
mkPackageOption pkgs [ "python3Packages" "pytorch" ] {
extraDescription = "This is an example and doesn't actually do anything.";
}
=> { ...; default = pkgs.python3Packages.pytorch; defaultText = literalExpression "pkgs.python3Packages.pytorch"; description = "The pytorch package to use. This is an example and doesn't actually do anything."; type = package; }
Example:
mkPackageOption pkgs "nushell" {
nullable = true;
}
=> { ...; default = pkgs.nushell; defaultText = literalExpression "pkgs.nushell"; description = "The nushell package to use."; type = nullOr package; }
Example:
mkPackageOption pkgs "coreutils" {
default = null;
}
=> { ...; description = "The coreutils package to use."; type = package; }
Example:
mkPackageOption pkgs "dbus" {
nullable = true;
default = null;
}
=> { ...; default = null; description = "The dbus package to use."; type = nullOr package; }
Example:
mkPackageOption pkgs.javaPackages "OpenJFX" {
default = "openjfx20";
pkgsText = "pkgs.javaPackages";
}
=> { ...; default = pkgs.javaPackages.openjfx20; defaultText = literalExpression "pkgs.javaPackages.openjfx20"; description = "The OpenJFX package to use."; type = package; }
*/
mkPackageOption =
# Package set (an instantiation of nixpkgs such as pkgs in modules or another package set)
pkgs:
# Name for the package, shown in option description
name:
{
# Whether the package can be null, for example to disable installing a package altogether (defaults to false)
nullable ? false,
# The attribute path where the default package is located (may be omitted, in which case it is copied from `name`)
default ? name,
# A string or an attribute path to use as an example (may be omitted)
example ? null,
# Additional text to include in the option description (may be omitted)
extraDescription ? "",
# Representation of the package set passed as pkgs (defaults to `"pkgs"`)
pkgsText ? "pkgs"
}:
let
name' = if isList name then last name else name;
default' = if isList default then default else [ default ];
defaultText = concatStringsSep "." default';
defaultValue = attrByPath default'
(throw "${defaultText} cannot be found in ${pkgsText}") pkgs;
defaults = if default != null then {
default = defaultValue;
defaultText = literalExpression ("${pkgsText}." + defaultText);
} else optionalAttrs nullable {
default = null;
};
in mkOption (defaults // {
description = "The ${name'} package to use."
+ (if extraDescription == "" then "" else " ") + extraDescription;
type = with lib.types; (if nullable then nullOr else lib.id) package;
} // optionalAttrs (example != null) {
example = literalExpression
(if isList example then "${pkgsText}." + concatStringsSep "." example else example);
});
/* Alias of mkPackageOption. Previously used to create options with markdown
documentation, which is no longer required.
*/
mkPackageOptionMD = mkPackageOption;
/* This option accepts anything, but it does not produce any result.
This is useful for sharing a module across different module sets
without having to implement similar features as long as the
values of the options are not accessed. */
mkSinkUndeclaredOptions = attrs: mkOption ({
internal = true;
visible = false;
default = false;
description = "Sink for option definitions.";
type = mkOptionType {
name = "sink";
check = x: true;
merge = loc: defs: false;
};
apply = x: throw "Option value is not readable because the option is not declared.";
} // attrs);
mergeDefaultOption = loc: defs:
let list = getValues defs; in
if length list == 1 then head list
else if all isFunction list then x: mergeDefaultOption loc (map (f: f x) list)
else if all isList list then concatLists list
else if all isAttrs list then foldl' lib.mergeAttrs {} list
else if all isBool list then foldl' lib.or false list
else if all isString list then lib.concatStrings list
else if all isInt list && all (x: x == head list) list then head list
else throw "Cannot merge definitions of `${showOption loc}'. Definition values:${showDefs defs}";
/*
Require a single definition.
WARNING: Does not perform nested checks, as this does not run the merge function!
*/
mergeOneOption = mergeUniqueOption { message = ""; };
/*
Require a single definition.
NOTE: When the type is not checked completely by check, pass a merge function for further checking (of sub-attributes, etc).
*/
mergeUniqueOption = args@{
message,
# WARNING: the default merge function assumes that the definition is a valid (option) value. You MUST pass a merge function if the return value needs to be
# - type checked beyond what .check does (which should be very litte; only on the value head; not attribute values, etc)
# - if you want attribute values to be checked, or list items
# - if you want coercedTo-like behavior to work
merge ? loc: defs: (head defs).value }:
loc: defs:
if length defs == 1
then merge loc defs
else
assert length defs > 1;
throw "The option `${showOption loc}' is defined multiple times while it's expected to be unique.\n${message}\nDefinition values:${showDefs defs}\n${prioritySuggestion}";
/* "Merge" option definitions by checking that they all have the same value. */
mergeEqualOption = loc: defs:
if defs == [] then abort "This case should never happen."
# Return early if we only have one element
# This also makes it work for functions, because the foldl' below would try
# to compare the first element with itself, which is false for functions
else if length defs == 1 then (head defs).value
else (foldl' (first: def:
if def.value != first.value then
throw "The option `${showOption loc}' has conflicting definition values:${showDefs [ first def ]}\n${prioritySuggestion}"
else
first) (head defs) (tail defs)).value;
/* Extracts values of all "value" keys of the given list.
Type: getValues :: [ { value :: a; } ] -> [a]
Example:
getValues [ { value = 1; } { value = 2; } ] // => [ 1 2 ]
getValues [ ] // => [ ]
*/
getValues = map (x: x.value);
/* Extracts values of all "file" keys of the given list
Type: getFiles :: [ { file :: a; } ] -> [a]
Example:
getFiles [ { file = "file1"; } { file = "file2"; } ] // => [ "file1" "file2" ]
getFiles [ ] // => [ ]
*/
getFiles = map (x: x.file);
# Generate documentation template from the list of option declaration like
# the set generated with filterOptionSets.
optionAttrSetToDocList = optionAttrSetToDocList' [];
optionAttrSetToDocList' = _: options:
concatMap (opt:
let
name = showOption opt.loc;
docOption = {
loc = opt.loc;
inherit name;
description = opt.description or null;
declarations = filter (x: x != unknownModule) opt.declarations;
internal = opt.internal or false;
visible =
if (opt?visible && opt.visible == "shallow")
then true
else opt.visible or true;
readOnly = opt.readOnly or false;
type = opt.type.description or "unspecified";
}
// optionalAttrs (opt ? example) {
example =
builtins.addErrorContext "while evaluating the example of option `${name}`" (
renderOptionValue opt.example
);
}
// optionalAttrs (opt ? defaultText || opt ? default) {
default =
builtins.addErrorContext "while evaluating the ${if opt?defaultText then "defaultText" else "default value"} of option `${name}`" (
renderOptionValue (opt.defaultText or opt.default)
);
}
// optionalAttrs (opt ? relatedPackages && opt.relatedPackages != null) { inherit (opt) relatedPackages; };
subOptions =
let ss = opt.type.getSubOptions opt.loc;
in if ss != {} then optionAttrSetToDocList' opt.loc ss else [];
subOptionsVisible = docOption.visible && opt.visible or null != "shallow";
in
# To find infinite recursion in NixOS option docs:
# builtins.trace opt.loc
[ docOption ] ++ optionals subOptionsVisible subOptions) (collect isOption options);
/* This function recursively removes all derivation attributes from
`x` except for the `name` attribute.
This is to make the generation of `options.xml` much more
efficient: the XML representation of derivations is very large
(on the order of megabytes) and is not actually used by the
manual generator.
This function was made obsolete by renderOptionValue and is kept for
compatibility with out-of-tree code.
*/
scrubOptionValue = x:
if isDerivation x then
{ type = "derivation"; drvPath = x.name; outPath = x.name; name = x.name; }
else if isList x then map scrubOptionValue x
else if isAttrs x then mapAttrs (n: v: scrubOptionValue v) (removeAttrs x ["_args"])
else x;
/* Ensures that the given option value (default or example) is a `_type`d string
by rendering Nix values to `literalExpression`s.
*/
renderOptionValue = v:
if v ? _type && v ? text then v
else literalExpression (lib.generators.toPretty {
multiline = true;
allowPrettyValues = true;
} v);
/* For use in the `defaultText` and `example` option attributes. Causes the
given string to be rendered verbatim in the documentation as Nix code. This
is necessary for complex values, e.g. functions, or values that depend on
other values or packages.
*/
literalExpression = text:
if ! isString text then throw "literalExpression expects a string."
else { _type = "literalExpression"; inherit text; };
literalExample = lib.warn "lib.literalExample is deprecated, use lib.literalExpression instead, or use lib.literalMD for a non-Nix description." literalExpression;
/* Transition marker for documentation that's already migrated to markdown
syntax. Has been a no-op for some while and been removed from nixpkgs.
Kept here to alert downstream users who may not be aware of the migration's
completion that it should be removed from modules.
*/
mdDoc = lib.warn "lib.mdDoc will be removed from nixpkgs in 24.11. Option descriptions are now in Markdown by default; you can remove any remaining uses of lib.mdDoc.";
/* For use in the `defaultText` and `example` option attributes. Causes the
given MD text to be inserted verbatim in the documentation, for when
a `literalExpression` would be too hard to read.
*/
literalMD = text:
if ! isString text then throw "literalMD expects a string."
else { _type = "literalMD"; inherit text; };
# Helper functions.
/* Convert an option, described as a list of the option parts to a
human-readable version.
Example:
(showOption ["foo" "bar" "baz"]) == "foo.bar.baz"
(showOption ["foo" "bar.baz" "tux"]) == "foo.\"bar.baz\".tux"
(showOption ["windowManager" "2bwm" "enable"]) == "windowManager.\"2bwm\".enable"
Placeholders will not be quoted as they are not actual values:
(showOption ["foo" "*" "bar"]) == "foo.*.bar"
(showOption ["foo" "<name>" "bar"]) == "foo.<name>.bar"
*/
showOption = parts: let
escapeOptionPart = part:
let
# We assume that these are "special values" and not real configuration data.
# If it is real configuration data, it is rendered incorrectly.
specialIdentifiers = [
"<name>" # attrsOf (submodule {})
"*" # listOf (submodule {})
"<function body>" # functionTo
];
in if builtins.elem part specialIdentifiers
then part
else lib.strings.escapeNixIdentifier part;
in (concatStringsSep ".") (map escapeOptionPart parts);
showFiles = files: concatStringsSep " and " (map (f: "`${f}'") files);
showDefs = defs: concatMapStrings (def:
let
# Pretty print the value for display, if successful
prettyEval = builtins.tryEval
(lib.generators.toPretty { }
(lib.generators.withRecursion { depthLimit = 10; throwOnDepthLimit = false; } def.value));
# Split it into its lines
lines = filter (v: ! isList v) (builtins.split "\n" prettyEval.value);
# Only display the first 5 lines, and indent them for better visibility
value = concatStringsSep "\n " (take 5 lines ++ optional (length lines > 5) "...");
result =
# Don't print any value if evaluating the value strictly fails
if ! prettyEval.success then ""
# Put it on a new line if it consists of multiple
else if length lines > 1 then ":\n " + value
else ": " + value;
in "\n- In `${def.file}'${result}"
) defs;
showOptionWithDefLocs = opt: ''
${showOption opt.loc}, with values defined in:
${concatMapStringsSep "\n" (defFile: " - ${defFile}") opt.files}
'';
unknownModule = "<unknown-file>";
}

217
lib/path/README.md Normal file
View file

@ -0,0 +1,217 @@
# Path library
This document explains why the `lib.path` library is designed the way it is.
The purpose of this library is to process [filesystem paths]. It does not read files from the filesystem.
It exists to support the native Nix [path value type] with extra functionality.
[filesystem paths]: https://en.m.wikipedia.org/wiki/Path_(computing)
[path value type]: https://nixos.org/manual/nix/stable/language/values.html#type-path
As an extension of the path value type, it inherits the same intended use cases and limitations:
- Only use paths to access files at evaluation time, such as the local project source.
- Paths cannot point to derivations, so they are unfit to represent dependencies.
- A path implicitly imports the referenced files into the Nix store when interpolated to a string. Therefore paths are not suitable to access files at build- or run-time, as you risk importing the path from the evaluation system instead.
Overall, this library works with two types of paths:
- Absolute paths are represented with the Nix [path value type]. Nix automatically normalises these paths.
- Subpaths are represented with the [string value type] since path value types don't support relative paths. This library normalises these paths as safely as possible. Absolute paths in strings are not supported.
A subpath refers to a specific file or directory within an absolute base directory.
It is a stricter form of a relative path, notably [without support for `..` components][parents] since those could escape the base directory.
[string value type]: https://nixos.org/manual/nix/stable/language/values.html#type-string
This library is designed to be as safe and intuitive as possible, throwing errors when operations are attempted that would produce surprising results, and giving the expected result otherwise.
This library is designed to work well as a dependency for the `lib.filesystem` and `lib.sources` library components. Contrary to these library components, `lib.path` does not read any paths from the filesystem.
This library makes only these assumptions about paths and no others:
- `dirOf path` returns the path to the parent directory of `path`, unless `path` is the filesystem root, in which case `path` is returned.
- There can be multiple filesystem roots: `p == dirOf p` and `q == dirOf q` does not imply `p == q`.
- While there's only a single filesystem root in stable Nix, the [lazy trees feature](https://github.com/NixOS/nix/pull/6530) introduces [additional filesystem roots](https://github.com/NixOS/nix/pull/6530#discussion_r1041442173).
- `path + ("/" + string)` returns the path to the `string` subdirectory in `path`.
- If `string` contains no `/` characters, then `dirOf (path + ("/" + string)) == path`.
- If `string` contains no `/` characters, then `baseNameOf (path + ("/" + string)) == string`.
- `path1 == path2` returns `true` only if `path1` points to the same filesystem path as `path2`.
Notably we do not make the assumption that we can turn paths into strings using `toString path`.
## Design decisions
Each subsection here contains a decision along with arguments and counter-arguments for (+) and against (-) that decision.
### Leading dots for relative paths
[leading-dots]: #leading-dots-for-relative-paths
Observing: Since subpaths are a form of relative paths, they can have a leading `./` to indicate it being a relative path, this is generally not necessary for tools though.
Considering: Paths should be as explicit, consistent and unambiguous as possible.
Decision: Returned subpaths should always have a leading `./`.
<details>
<summary>Arguments</summary>
- (+) In shells, just running `foo` as a command wouldn't execute the file `foo`, whereas `./foo` would execute the file. In contrast, `foo/bar` does execute that file without the need for `./`. This can lead to confusion about when a `./` needs to be prefixed. If a `./` is always included, this becomes a non-issue. This effectively then means that paths don't overlap with command names.
- (+) Prepending with `./` makes the subpaths always valid as relative Nix path expressions.
- (+) Using paths in command line arguments could give problems if not escaped properly, e.g. if a path was `--version`. This is not a problem with `./--version`. This effectively then means that paths don't overlap with GNU-style command line options.
- (-) `./` is not required to resolve relative paths, resolution always has an implicit `./` as prefix.
- (-) It's less noisy without the `./`, e.g. in error messages.
- (+) But similarly, it could be confusing whether something was even a path.
e.g. `foo` could be anything, but `./foo` is more clearly a path.
- (+) Makes it more uniform with absolute paths (those always start with `/`).
- (-) That is not relevant for practical purposes.
- (+) `find` also outputs results with `./`.
- (-) But only if you give it an argument of `.`. If you give it the argument `some-directory`, it won't prefix that.
- (-) `realpath --relative-to` doesn't prefix relative paths with `./`.
- (+) There is no need to return the same result as `realpath`.
</details>
### Representation of the current directory
[curdir]: #representation-of-the-current-directory
Observing: The subpath that produces the base directory can be represented with `.` or `./` or `./.`.
Considering: Paths should be as consistent and unambiguous as possible.
Decision: It should be `./.`.
<details>
<summary>Arguments</summary>
- (+) `./` would be inconsistent with [the decision to not persist trailing slashes][trailing-slashes].
- (-) `.` is how `realpath` normalises paths.
- (+) `.` can be interpreted as a shell command (it's a builtin for sourcing files in `bash` and `zsh`).
- (+) `.` would be the only path without a `/`. It could not be used as a Nix path expression, since those require at least one `/` to be parsed as such.
- (-) `./.` is rather long.
- (-) We don't require users to type this though, as it's only output by the library.
As inputs all three variants are supported for subpaths (and we can't do anything about absolute paths)
- (-) `builtins.dirOf "foo" == "."`, so `.` would be consistent with that.
- (+) `./.` is consistent with the [decision to have leading `./`][leading-dots].
- (+) `./.` is a valid Nix path expression, although this property does not hold for every relative path or subpath.
</details>
### Subpath representation
[relrepr]: #subpath-representation
Observing: Subpaths such as `foo/bar` can be represented in various ways:
- string: `"foo/bar"`
- list with all the components: `[ "foo" "bar" ]`
- attribute set: `{ type = "relative-path"; components = [ "foo" "bar" ]; }`
Considering: Paths should be as safe to use as possible. We should generate string outputs in the library and not encourage users to do that themselves.
Decision: Paths are represented as strings.
<details>
<summary>Arguments</summary>
- (+) It's simpler for the users of the library. One doesn't have to convert a path a string before it can be used.
- (+) Naively converting the list representation to a string with `concatStringsSep "/"` would break for `[]`, requiring library users to be more careful.
- (+) It doesn't encourage people to do their own path processing and instead use the library.
With a list representation it would seem easy to just use `lib.lists.init` to get the parent directory, but then it breaks for `.`, which would be represented as `[ ]`.
- (+) `+` is convenient and doesn't work on lists and attribute sets.
- (-) Shouldn't use `+` anyways, we export safer functions for path manipulation.
</details>
### Parent directory
[parents]: #parent-directory
Observing: Relative paths can have `..` components, which refer to the parent directory.
Considering: Paths should be as safe and unambiguous as possible.
Decision: `..` path components in string paths are not supported, neither as inputs nor as outputs. Hence, string paths are called subpaths, rather than relative paths.
<details>
<summary>Arguments</summary>
- (+) If we wanted relative paths to behave according to the "physical" interpretation (as a directory tree with relations between nodes), it would require resolving symlinks, since e.g. `foo/..` would not be the same as `.` if `foo` is a symlink.
- (-) The "logical" interpretation is also valid (treating paths as a sequence of names), and is used by some software. It is simpler, and not using symlinks at all is safer.
- (+) Mixing both models can lead to surprises.
- (+) We can't resolve symlinks without filesystem access.
- (+) Nix also doesn't support reading symlinks at evaluation time.
- (-) We could just not handle such cases, e.g. `equals "foo" "foo/bar/.. == false`. The paths are different, we don't need to check whether the paths point to the same thing.
- (+) Assume we said `relativeTo /foo /bar == "../bar"`. If this is used like `/bar/../foo` in the end, and `bar` turns out to be a symlink to somewhere else, this won't be accurate.
- (-) We could decide to not support such ambiguous operations, or mark them as such, e.g. the normal `relativeTo` will error on such a case, but there could be `extendedRelativeTo` supporting that.
- (-) `..` are a part of paths, a path library should therefore support it.
- (+) If we can convincingly argue that all such use cases are better done e.g. with runtime tools, the library not supporting it can nudge people towards using those.
- (-) We could allow "..", but only in the prefix.
- (+) Then we'd have to throw an error for doing `append /some/path "../foo"`, making it non-composable.
- (+) The same is for returning paths with `..`: `relativeTo /foo /bar => "../bar"` would produce a non-composable path.
- (+) We argue that `..` is not needed at the Nix evaluation level, since we'd always start evaluation from the project root and don't go up from there.
- (+) `..` is supported in Nix paths, turning them into absolute paths.
- (-) This is ambiguous in the presence of symlinks.
- (+) If you need `..` for building or runtime, you can use build-/run-time tooling to create those (e.g. `realpath` with `--relative-to`), or use absolute paths instead.
This also gives you the ability to correctly handle symlinks.
</details>
### Trailing slashes
[trailing-slashes]: #trailing-slashes
Observing: Subpaths can contain trailing slashes, like `foo/`, indicating that the path points to a directory and not a file.
Considering: Paths should be as consistent as possible, there should only be a single normalisation for the same path.
Decision: All functions remove trailing slashes in their results.
<details>
<summary>Arguments</summary>
- (+) It allows normalisations to be unique, in that there's only a single normalisation for the same path. If trailing slashes were preserved, both `foo/bar` and `foo/bar/` would be valid but different normalisations for the same path.
- Comparison to other frameworks to figure out the least surprising behavior:
- (+) Nix itself doesn't support trailing slashes when parsing and doesn't preserve them when appending paths.
- (-) [Rust's std::path](https://doc.rust-lang.org/std/path/index.html) does preserve them during [construction](https://doc.rust-lang.org/std/path/struct.Path.html#method.new).
- (+) Doesn't preserve them when returning individual [components](https://doc.rust-lang.org/std/path/struct.Path.html#method.components).
- (+) Doesn't preserve them when [canonicalizing](https://doc.rust-lang.org/std/path/struct.Path.html#method.canonicalize).
- (+) [Python 3's pathlib](https://docs.python.org/3/library/pathlib.html#module-pathlib) doesn't preserve them during [construction](https://docs.python.org/3/library/pathlib.html#pathlib.PurePath).
- Notably it represents the individual components as a list internally.
- (-) [Haskell's filepath](https://hackage.haskell.org/package/filepath-1.4.100.0) has [explicit support](https://hackage.haskell.org/package/filepath-1.4.100.0/docs/System-FilePath.html#g:6) for handling trailing slashes.
- (-) Does preserve them for [normalisation](https://hackage.haskell.org/package/filepath-1.4.100.0/docs/System-FilePath.html#v:normalise).
- (-) [NodeJS's Path library](https://nodejs.org/api/path.html) preserves trailing slashes for [normalisation](https://nodejs.org/api/path.html#pathnormalizepath).
- (+) For [parsing a path](https://nodejs.org/api/path.html#pathparsepath) into its significant elements, trailing slashes are not preserved.
- (+) Nix's builtin function `dirOf` gives an unexpected result for paths with trailing slashes: `dirOf "foo/bar/" == "foo/bar"`.
Inconsistently, `baseNameOf` works correctly though: `baseNameOf "foo/bar/" == "bar"`.
- (-) We are writing a path library to improve handling of paths though, so we shouldn't use these functions and discourage their use.
- (-) Unexpected result when normalising intermediate paths, like `relative.normalise ("foo" + "/") + "bar" == "foobar"`.
- (+) This is not a practical use case though.
- (+) Don't use `+` to append paths, this library has a `join` function for that.
- (-) Users might use `+` out of habit though.
- (+) The `realpath` command also removes trailing slashes.
- (+) Even with a trailing slash, the path is the same, it's only an indication that it's a directory.
</details>
### Prefer returning subpaths over components
[subpath-preference]: #prefer-returning-subpaths-over-components
Observing: Functions could return subpaths or lists of path component strings.
Considering: Subpaths are used as inputs for some functions. Using them for outputs, too, makes the library more consistent and composable.
Decision: Subpaths should be preferred over list of path component strings.
<details>
<summary>Arguments</summary>
- (+) It is consistent with functions accepting subpaths, making the library more composable
- (-) It is less efficient when the components are needed, because after creating the normalised subpath string, it will have to be parsed into components again
- (+) If necessary, we can still make it faster by adding builtins to Nix
- (+) Alternatively if necessary, versions of these functions that return components could later still be introduced.
- (+) It makes the path library simpler because there's only two types (paths and subpaths). Only `lib.path.subpath.components` can be used to get a list of components.
And once we have a list of component strings, `lib.lists` and `lib.strings` can be used to operate on them.
For completeness, `lib.path.subpath.join` allows converting the list of components back to a subpath.
</details>
## Other implementations and references
- [Rust](https://doc.rust-lang.org/std/path/struct.Path.html)
- [Python](https://docs.python.org/3/library/pathlib.html)
- [Haskell](https://hackage.haskell.org/package/filepath-1.4.100.0/docs/System-FilePath.html)
- [Nodejs](https://nodejs.org/api/path.html)
- [POSIX.1-2017](https://pubs.opengroup.org/onlinepubs/9699919799/nframe.html)

641
lib/path/default.nix Normal file
View file

@ -0,0 +1,641 @@
/* Functions for working with path values. */
# See ./README.md for internal docs
{ lib }:
let
inherit (builtins)
isString
isPath
split
match
typeOf
storeDir
;
inherit (lib.lists)
length
head
last
genList
elemAt
all
concatMap
foldl'
take
drop
;
listHasPrefix = lib.lists.hasPrefix;
inherit (lib.strings)
concatStringsSep
substring
;
inherit (lib.asserts)
assertMsg
;
inherit (lib.path.subpath)
isValid
;
# Return the reason why a subpath is invalid, or `null` if it's valid
subpathInvalidReason = value:
if ! isString value then
"The given value is of type ${builtins.typeOf value}, but a string was expected"
else if value == "" then
"The given string is empty"
else if substring 0 1 value == "/" then
"The given string \"${value}\" starts with a `/`, representing an absolute path"
# We don't support ".." components, see ./path.md#parent-directory
else if match "(.*/)?\\.\\.(/.*)?" value != null then
"The given string \"${value}\" contains a `..` component, which is not allowed in subpaths"
else null;
# Split and normalise a relative path string into its components.
# Error for ".." components and doesn't include "." components
splitRelPath = path:
let
# Split the string into its parts using regex for efficiency. This regex
# matches patterns like "/", "/./", "/././", with arbitrarily many "/"s
# together. These are the main special cases:
# - Leading "./" gets split into a leading "." part
# - Trailing "/." or "/" get split into a trailing "." or ""
# part respectively
#
# These are the only cases where "." and "" parts can occur
parts = split "/+(\\./+)*" path;
# `split` creates a list of 2 * k + 1 elements, containing the k +
# 1 parts, interleaved with k matches where k is the number of
# (non-overlapping) matches. This calculation here gets the number of parts
# back from the list length
# floor( (2 * k + 1) / 2 ) + 1 == floor( k + 1/2 ) + 1 == k + 1
partCount = length parts / 2 + 1;
# To assemble the final list of components we want to:
# - Skip a potential leading ".", normalising "./foo" to "foo"
# - Skip a potential trailing "." or "", normalising "foo/" and "foo/." to
# "foo". See ./path.md#trailing-slashes
skipStart = if head parts == "." then 1 else 0;
skipEnd = if last parts == "." || last parts == "" then 1 else 0;
# We can now know the length of the result by removing the number of
# skipped parts from the total number
componentCount = partCount - skipEnd - skipStart;
in
# Special case of a single "." path component. Such a case leaves a
# componentCount of -1 due to the skipStart/skipEnd not verifying that
# they don't refer to the same character
if path == "." then []
# Generate the result list directly. This is more efficient than a
# combination of `filter`, `init` and `tail`, because here we don't
# allocate any intermediate lists
else genList (index:
# To get to the element we need to add the number of parts we skip and
# multiply by two due to the interleaved layout of `parts`
elemAt parts ((skipStart + index) * 2)
) componentCount;
# Join relative path components together
joinRelPath = components:
# Always return relative paths with `./` as a prefix (./path.md#leading-dots-for-relative-paths)
"./" +
# An empty string is not a valid relative path, so we need to return a `.` when we have no components
(if components == [] then "." else concatStringsSep "/" components);
# Type: Path -> { root :: Path, components :: [ String ] }
#
# Deconstruct a path value type into:
# - root: The filesystem root of the path, generally `/`
# - components: All the path's components
#
# This is similar to `splitString "/" (toString path)` but safer
# because it can distinguish different filesystem roots
deconstructPath =
let
recurse = components: base:
# If the parent of a path is the path itself, then it's a filesystem root
if base == dirOf base then { root = base; inherit components; }
else recurse ([ (baseNameOf base) ] ++ components) (dirOf base);
in recurse [];
# The components of the store directory, typically [ "nix" "store" ]
storeDirComponents = splitRelPath ("./" + storeDir);
# The number of store directory components, typically 2
storeDirLength = length storeDirComponents;
# Type: [ String ] -> Bool
#
# Whether path components have a store path as a prefix, according to
# https://nixos.org/manual/nix/stable/store/store-path.html#store-path.
componentsHaveStorePathPrefix = components:
# path starts with the store directory (typically /nix/store)
listHasPrefix storeDirComponents components
# is not the store directory itself, meaning there's at least one extra component
&& storeDirComponents != components
# and the first component after the store directory has the expected format.
# NOTE: We could change the hash regex to be [0-9a-df-np-sv-z],
# because these are the actual ASCII characters used by Nix's base32 implementation,
# but this is not fully specified, so let's tie this too much to the currently implemented concept of store paths.
# Similar reasoning applies to the validity of the name part.
# We care more about discerning store path-ness on realistic values. Making it airtight would be fragile and slow.
&& match ".{32}-.+" (elemAt components storeDirLength) != null;
in /* No rec! Add dependencies on this file at the top. */ {
/*
Append a subpath string to a path.
Like `path + ("/" + string)` but safer, because it errors instead of returning potentially surprising results.
More specifically, it checks that the first argument is a [path value type](https://nixos.org/manual/nix/stable/language/values.html#type-path"),
and that the second argument is a [valid subpath string](#function-library-lib.path.subpath.isValid).
Laws:
- Not influenced by subpath [normalisation](#function-library-lib.path.subpath.normalise):
append p s == append p (subpath.normalise s)
Type:
append :: Path -> String -> Path
Example:
append /foo "bar/baz"
=> /foo/bar/baz
# subpaths don't need to be normalised
append /foo "./bar//baz/./"
=> /foo/bar/baz
# can append to root directory
append /. "foo/bar"
=> /foo/bar
# first argument needs to be a path value type
append "/foo" "bar"
=> <error>
# second argument needs to be a valid subpath string
append /foo /bar
=> <error>
append /foo ""
=> <error>
append /foo "/bar"
=> <error>
append /foo "../bar"
=> <error>
*/
append =
# The absolute path to append to
path:
# The subpath string to append
subpath:
assert assertMsg (isPath path) ''
lib.path.append: The first argument is of type ${builtins.typeOf path}, but a path was expected'';
assert assertMsg (isValid subpath) ''
lib.path.append: Second argument is not a valid subpath string:
${subpathInvalidReason subpath}'';
path + ("/" + subpath);
/*
Whether the first path is a component-wise prefix of the second path.
Laws:
- `hasPrefix p q` is only true if [`q == append p s`](#function-library-lib.path.append) for some [subpath](#function-library-lib.path.subpath.isValid) `s`.
- `hasPrefix` is a [non-strict partial order](https://en.wikipedia.org/wiki/Partially_ordered_set#Non-strict_partial_order) over the set of all path values.
Type:
hasPrefix :: Path -> Path -> Bool
Example:
hasPrefix /foo /foo/bar
=> true
hasPrefix /foo /foo
=> true
hasPrefix /foo/bar /foo
=> false
hasPrefix /. /foo
=> true
*/
hasPrefix =
path1:
assert assertMsg
(isPath path1)
"lib.path.hasPrefix: First argument is of type ${typeOf path1}, but a path was expected";
let
path1Deconstructed = deconstructPath path1;
in
path2:
assert assertMsg
(isPath path2)
"lib.path.hasPrefix: Second argument is of type ${typeOf path2}, but a path was expected";
let
path2Deconstructed = deconstructPath path2;
in
assert assertMsg
(path1Deconstructed.root == path2Deconstructed.root) ''
lib.path.hasPrefix: Filesystem roots must be the same for both paths, but paths with different roots were given:
first argument: "${toString path1}" with root "${toString path1Deconstructed.root}"
second argument: "${toString path2}" with root "${toString path2Deconstructed.root}"'';
take (length path1Deconstructed.components) path2Deconstructed.components == path1Deconstructed.components;
/*
Remove the first path as a component-wise prefix from the second path.
The result is a [normalised subpath string](#function-library-lib.path.subpath.normalise).
Laws:
- Inverts [`append`](#function-library-lib.path.append) for [normalised subpath string](#function-library-lib.path.subpath.normalise):
removePrefix p (append p s) == subpath.normalise s
Type:
removePrefix :: Path -> Path -> String
Example:
removePrefix /foo /foo/bar/baz
=> "./bar/baz"
removePrefix /foo /foo
=> "./."
removePrefix /foo/bar /foo
=> <error>
removePrefix /. /foo
=> "./foo"
*/
removePrefix =
path1:
assert assertMsg
(isPath path1)
"lib.path.removePrefix: First argument is of type ${typeOf path1}, but a path was expected.";
let
path1Deconstructed = deconstructPath path1;
path1Length = length path1Deconstructed.components;
in
path2:
assert assertMsg
(isPath path2)
"lib.path.removePrefix: Second argument is of type ${typeOf path2}, but a path was expected.";
let
path2Deconstructed = deconstructPath path2;
success = take path1Length path2Deconstructed.components == path1Deconstructed.components;
components =
if success then
drop path1Length path2Deconstructed.components
else
throw ''
lib.path.removePrefix: The first path argument "${toString path1}" is not a component-wise prefix of the second path argument "${toString path2}".'';
in
assert assertMsg
(path1Deconstructed.root == path2Deconstructed.root) ''
lib.path.removePrefix: Filesystem roots must be the same for both paths, but paths with different roots were given:
first argument: "${toString path1}" with root "${toString path1Deconstructed.root}"
second argument: "${toString path2}" with root "${toString path2Deconstructed.root}"'';
joinRelPath components;
/*
Split the filesystem root from a [path](https://nixos.org/manual/nix/stable/language/values.html#type-path).
The result is an attribute set with these attributes:
- `root`: The filesystem root of the path, meaning that this directory has no parent directory.
- `subpath`: The [normalised subpath string](#function-library-lib.path.subpath.normalise) that when [appended](#function-library-lib.path.append) to `root` returns the original path.
Laws:
- [Appending](#function-library-lib.path.append) the `root` and `subpath` gives the original path:
p ==
append
(splitRoot p).root
(splitRoot p).subpath
- Trying to get the parent directory of `root` using [`readDir`](https://nixos.org/manual/nix/stable/language/builtins.html#builtins-readDir) returns `root` itself:
dirOf (splitRoot p).root == (splitRoot p).root
Type:
splitRoot :: Path -> { root :: Path, subpath :: String }
Example:
splitRoot /foo/bar
=> { root = /.; subpath = "./foo/bar"; }
splitRoot /.
=> { root = /.; subpath = "./."; }
# Nix neutralises `..` path components for all path values automatically
splitRoot /foo/../bar
=> { root = /.; subpath = "./bar"; }
splitRoot "/foo