Commit graph

2294 commits

Author SHA1 Message Date
Robert Hensing
aba6eb348e libstore: Make sure that initNix has been called
Prevent bugs like https://github.com/cachix/cachix/pull/477
2022-12-24 14:39:30 +01:00
Philipp Otterbein
8af839f48c remove undefined function 2022-12-24 12:19:53 +01:00
Philipp Otterbein
a6e9d9cb2f remove function makeImmutableStringWithLen 2022-12-24 12:09:06 +01:00
mupdt
a33e45b60b primops storeDir test uses settings.nixStore 2022-12-21 07:01:57 -05:00
Eelco Dolstra
bda879170f EvalState::copyPathToStore(): Return a StorePath 2022-12-20 14:58:39 +01:00
Eelco Dolstra
8332ac6a1d
Merge pull request #7451 from edolstra/abstract-pos
Introduce AbstractPos
2022-12-20 12:55:13 +01:00
Solène Rapenne
d0660c6c0b
printMsg replacement by printInfo
Co-authored-by: Théophane Hufschmitt <7226587+thufschmitt@users.noreply.github.com>
2022-12-16 09:34:22 +01:00
Solène Rapenne
98e01da0b1 warnings: switch to info level when using a saved substituter 2022-12-14 14:11:21 +01:00
Eelco Dolstra
c9b0a85b08 Restore display of source lines for stdin/string inputs 2022-12-13 16:00:44 +01:00
Eelco Dolstra
aea97f07a3 Fix compilation 2022-12-13 15:23:12 +01:00
Eelco Dolstra
1315133b50
Improve cast safety
Co-authored-by: Théophane Hufschmitt <7226587+thufschmitt@users.noreply.github.com>
2022-12-13 12:38:33 +01:00
Théophane Hufschmitt
6e31d27cba
Merge pull request #6741 from Mindavi/nix-no-gc
support building with --enable-gc=no
2022-12-13 10:36:59 +01:00
Eelco Dolstra
b3fdab28a2 Introduce AbstractPos
This makes the position object used in exceptions abstract, with a
method getSource() to get the source code of the file in which the
error originated. This is needed for lazy trees because source files
don't necessarily exist in the filesystem, and we don't want to make
libutil depend on the InputAccessor type in libfetcher.
2022-12-13 00:50:43 +01:00
Rick van Schijndel
672ee88231 support building with --enable-gc=no
Some minor changes fixing the build without boehm.
Fixes NixOS#6250
2022-12-12 23:31:30 +01:00
John Ericson
7b122d43a4 Fix stack context notes to not rely on order
Make everything be in the form "while ..." (most things were already),
and in particular *don't* use other propositions that must go after or
before specific "while ..." clauses to make sense.
2022-12-12 18:41:00 +00:00
Eelco Dolstra
fd0ed75118 Support flake references in the old CLI
Fixes #7026.
2022-12-12 14:05:52 +01:00
Eelco Dolstra
ae5f62a894 Move isUri() and resolveUri() out of filetransfer.cc
These are purely related to NIX_PATH / -I command line parsing, so put
them in libexpr.
2022-12-12 14:05:35 +01:00
Eelco Dolstra
877ea1dab8
Use get_ptr()
Co-authored-by: Théophane Hufschmitt <7226587+thufschmitt@users.noreply.github.com>
2022-12-12 12:46:13 +01:00
Eelco Dolstra
f3d1e92856
Update URL
Co-authored-by: Valentin Gagarin <valentin.gagarin@tweag.io>
2022-12-12 12:37:55 +01:00
Eelco Dolstra
703d863a48 Trivial changes from the lazy-trees branch 2022-12-07 14:06:34 +01:00
Eelco Dolstra
3a8341f57e
Merge pull request #7395 from obsidiansystems/evaluating-to-calling
Change "while evaluating <fun>" to "while *calling*" in trace
2022-12-02 15:53:57 +01:00
John Ericson
19c5394971 Change "while evaluating <fun>" to "while *calling*" in trace
The old way was not correct.

Here is an example:
```
 $ nix-instantiate --eval --expr 'let x = a: throw "asdf"; in x 1' --show-trace
error: asdf

       … while evaluating 'x'

       at «string»:1:9:

            1| let x = a: throw "asdf"; in x 1
             |         ^

       … from call site

       at «string»:1:29:

            1| let x = a: throw "asdf"; in x 1
             |                             ^
```
and yet also:
```
 $ nix-instantiate --eval --expr 'let x = a: throw "asdf"; in x' --show-trace
<LAMBDA>
```

Here is the thing: in both cases we are evaluating `x`!

Nix is a higher-order languages, and functions are a sort of value. When
we write `x = a: ...`, `a: ...` is the expression that `x` is being
defined to be, and that is already a value. Therefore, we should *never*
get an trace that says "while evaluating `x`", because evaluating `a:
...` is *trival* and nothing happens during it!

What is actually happening here is we are applying `x` and evaluating
its *body* with arguments substituted for parameters. I think the
simplest way to say is just "while *calling* `x`", and so that is what I
changed it to.
2022-12-02 09:14:12 -05:00
Alex Ameen
ef524013aa doc: listToAttrs: add extra whitespace 2022-12-01 10:32:45 -06:00
Alex Ameen
ec18b7d09b doc: listToAttrs: fix line wrapping 2022-11-30 23:21:09 -06:00
Alex Ameen
ad46726546 doc: listToAttrs: document repeated keys 2022-11-30 22:53:41 -06:00
Eelco Dolstra
dbf78a7ada
Merge pull request #7313 from yorickvP/nlohmann-everywhere
Replace src/libutil/json.cc with nlohmann
2022-11-28 15:03:48 +01:00
Eelco Dolstra
0b4c4d7434 Don't use GC_STRNDUP
It calls strlen() on the input (rather than simply copying at most
`size` bytes), which can fail if the input is not zero-terminated and
is inefficient in any case.

Fixes #7347.
2022-11-25 22:30:56 +01:00
Yorick van Pelt
09f00dd4d0
Replace src/libutil/json.cc with nlohmann json generation 2022-11-16 16:50:50 +01:00
Théophane Hufschmitt
8b4352d79b Merge remote-tracking branch 'nixos/master' into readFile-scan-references 2022-11-14 15:00:05 +01:00
Naïm Favier
e7ed9ae0c7
Restrict readFile context to references that appear in the string
When calling `builtins.readFile` on a store path, the references of that
path are currently added to the resulting string's context.

This change makes those references the *possible* context of the string,
but filters them to keep only the references whose hash actually appears
in the string, similarly to what is done for determining the runtime
references of a path.
2022-11-11 13:04:34 +01:00
Théophane Hufschmitt
9550b1d519
Merge pull request #7258 from ncfavier/fix-eval-error-fmt
Fix printing of eval errors with two format placeholders
2022-11-10 18:41:16 +01:00
Naïm Favier
dad859ba0f
Fix printing of eval errors with two format placeholders 2022-11-04 12:41:38 +01:00
Eelco Dolstra
da2c61637b
Use _type
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
2022-10-25 16:48:31 +02:00
Guillaume Maudoux
d6f5734c63 Complete genericClosure tests 2022-10-25 01:57:18 +02:00
Guillaume Maudoux
e93bf69b44 Rework error throwing, and test it 2022-10-25 01:46:10 +02:00
Shea Levy
334fa81d08
Mark flakes with .type = "flake".
Fixes #7186
2022-10-23 06:54:11 -04:00
Guillaume Maudoux
8bd8583bc7 Try to please clang with convoluted templates 2022-10-23 00:11:44 +02:00
Guillaume Maudoux
8c3afd2d68 Introduce an Error builder to tackle complexity 2022-10-22 23:37:54 +02:00
Graham Bennett
c5fd34a14e Build with C++20 2022-10-22 14:24:25 +01:00
Guillaume Maudoux
4a909c142c Rollback unneeded throwFrameErrorWithTrace function 2022-10-20 14:25:11 +02:00
Guillaume Maudoux
31ce52a045 Fix context message being printed twice with forceStringNoCtx 2022-10-20 14:18:35 +02:00
Guillaume Maudoux
512f6be9b5 Reword incomparable types message 2022-10-20 14:17:05 +02:00
Guillaume Maudoux
520404f450 Revert custom position of 'if' blocks 2022-10-20 13:55:15 +02:00
Guillaume Maudoux
b945b844a9 Initial frames support 2022-10-17 03:05:02 +02:00
Guillaume Maudoux
3f9f6ae127 Merge remote-tracking branch 'origin/master' into coerce-string 2022-10-16 20:39:19 +02:00
sternenseemann
d0f2da214b primops: make nature of foldl' strictness clearer
* Clarify the documentation of foldl': That the arguments are forced
  before application (?) of `op` is necessarily true. What is important
  to stress is that we force every application of `op`, even when the
  value turns out to be unused.

* Move the example before the comment about strictness to make it less
  confusing: It is a general example and doesn't really showcase anything
  about foldl' strictness.

* Add test cases which nail down aspects of foldl' strictness:
  * The initial accumulator value is not forced unconditionally.
  * Applications of op are forced.
  * The list elements are not forced unconditionally.
2022-10-16 14:29:12 +02:00
Eelco Dolstra
f3193edd87
Merge pull request #7149 from amjoseph-nixpkgs/pr/intersectAttrs/values
parseDrvName: remove doc/impl discrepancy, add test covering the gap
2022-10-12 10:00:39 +02:00
Adam Joseph
7ef71cd21f src/libexpr/primops.cc: parseDrvName: make documentation follow implementation
The documentation for `parseDrvName` does not agree with the implementation when
the derivation name contains a dash which is followed by something that is
neither a letter nor a digit.  This commit corrects the documentation to agree
with the implementation.
2022-10-08 17:21:34 -07:00
Eelco Dolstra
6806f20157
Merge pull request #7089 from NixOS/mention-accept-flake-config-in-warning
Mention `--accept-flake-config` in the related warning
2022-09-26 11:13:02 +02:00
Adam Joseph
dc205c75a2 src/libexpr/primops.cc: correct definition for intersectAttrs
The current definition of `intersectAttrs` is incorrect:

> Return a set consisting of the attributes in the set e2 that also exist in the
> set e1.

Recall that (Nix manual, section 5.1):

> An attribute set is a collection of name-value-pairs (called attributes)

According to the existing description of `intersectAttrs`, the following should
evaluate to the empty set, since no key-value *pair* (i.e. attribute) exists in
both sets:

```
builtins.intersectAttrs { x=3; } {x="foo";}
```

And yet:

```
nix-repl> builtins.intersectAttrs { x=3; } {x="foo";}
{ x = "foo"; }
```

Clearly the intent here was for the *names* of the resulting attribute set to be
the intersection of the *names* of the two arguments, and for the values of the
resulting attribute set to be the values from the second argument.

This commit corrects the definition, making it match the implementation and intent.
2022-09-24 23:51:09 -07:00
Théophane Hufschmitt
371013c08d
Merge pull request #7010 from edolstra/ignore-shell
nix develop: Ignore stdenv's $SHELL
2022-09-24 12:55:53 +02:00
Théophane Hufschmitt
1a5d094be7
Mention --accept-flake-config in the related warning
Make sure that people who run Nix in non-interactive mode (and so don't have the possibility to interactively accept the individual flake configuration settings) are aware of this flag.

Fix #7086
2022-09-24 12:11:26 +02:00
Adam Joseph
fb985f855c fetchurl.nix: change other use of __impure 2022-09-16 01:54:24 -07:00
Adam Joseph
673fd21b7c
Update src/libexpr/fetchurl.nix
Co-authored-by: Eelco Dolstra <edolstra@gmail.com>
2022-09-16 08:51:14 +00:00
Adam Joseph
b3550d9179 libexpr/fetchurl.nix: allow __impure fetch
This commit adds an optional `__impure` parameter to fetchurl.nix, which allows
the caller to use `libfetcher`'s fetcher in an impure derivation.  This allows
nixpkgs' patch-normalizing fetcher (fetchpatch) to be rewritten to use nix's
internal fetchurl, thereby eliminating the awkward "you can't use fetchpatch
here" banners scattered all over the place.

See also: https://github.com/NixOS/nixpkgs/pull/188587
2022-09-16 00:48:01 -07:00
Guillaume Maudoux
96f2dd99d3 fix remaining foo stuff 2022-09-11 23:09:36 +02:00
Guillaume Maudoux
e412bb6d30 fix remaining fixtures 2022-09-11 22:58:59 +02:00
Guillaume Maudoux
c7b901fd33 Cleanup error strings rebase 2022-09-11 01:34:19 +02:00
Guillaume Maudoux
eb460a9529 WIP: broken merge but need a git checkpoint 2022-09-07 00:34:03 +02:00
Eelco Dolstra
ece12a97d9 lockfile -> lock file for consistency 2022-09-06 19:20:31 +02:00
Eelco Dolstra
a8b3d777fb Revert "Merge pull request #6621 from Kha/nested-follows"
This reverts commit c530cda345, reversing
changes made to 4adcdff5c1.
2022-09-01 15:26:19 +02:00
Sebastian Ullrich
6f65c11780 Fix overlapping flake follows 2022-08-28 11:50:25 +02:00
Sebastian Ullrich
2b9d381301 Fix nested flake input overrides 2022-08-28 11:47:25 +02:00
Eelco Dolstra
a17ce0a8a9 Fix evaluation cache
98e361ad4c introduced a regression where
previously stored attributes were replaced by placeholders. As a
result, a command like 'nix build nixpkgs#hello' had to be executed at
least twice to get caching.

This code does not seem necessary for suggestions to work.
2022-08-24 21:19:43 +02:00
Naïm Favier
062e4fcdde
JSON: print paths as strings without copying them to the store
Makes `printValueAsJSON` not copy paths to the store for `nix eval
--json`, `nix-instantiate --eval --json` and `nix-env --json`.

Fixes https://github.com/NixOS/nix/issues/5612
2022-08-22 15:01:35 +02:00
Erik Arvstedt
53833dfb40 libexpr/flake: remove FIXME
Line 593 checks that all overrides (i.e. all elements of
`lockFlags.inputOverrides`) are members of `overridesUsed`.
2022-08-04 09:45:29 +02:00
Erik Arvstedt
c9f446ede1 flakeref: fix comment 2022-08-04 09:45:28 +02:00
Eelco Dolstra
ccbd906c86 Fix NIX_COUNT_CALLS=1
Also, make the JSON writer support std::string_view.

Fixes #6857.
2022-08-03 17:46:51 +02:00
Eelco Dolstra
12df8885cc Simplify the check for overrides on non-existent inputs 2022-07-13 13:40:40 +02:00
Ikko Ashimine
694a9dc282
Fix typo in flake.cc
non-existant -> non-existent
2022-07-13 01:10:32 +09:00
Théophane Hufschmitt
2dbd5ed0b4
Merge pull request #6663 from Ma27/follows-invalid-input
flakes: throw an error if `follows`-declaration for an input is invalid
2022-07-12 16:44:22 +02:00
Eelco Dolstra
f6a434c8a4 Fix debug message 2022-07-12 11:53:34 +02:00
Maximilian Bosch
1f771065f1
Move follows-check into its own function 2022-07-12 11:25:33 +02:00
Maximilian Bosch
411111a3bc
Turn error for non-existant follows into a warning 2022-07-12 11:22:35 +02:00
Maximilian Bosch
c1c37f3200
flakes: throw an error if follows-declaration for an input is invalid
I recently got fairly confused why the following expression didn't have
any effect

    {
      description = "Foobar";
      inputs.sops-nix = {
        url = github:mic92/sops-nix;
        inputs.nixpkgs_22_05.follows = "nixpkgs";
      };
    }

until I found out that the input was called `nixpkgs-22_05` (please note
the dash vs. underscore).

IMHO it's not a good idea to not throw an error in that case and
probably leave end-users rather confused, so I implemented a small check
for that which basically checks whether `follows`-declaration from
overrides actually have corresponding inputs in the transitive flake.

In fact this was done by accident already in our own test-suite where
the removal of a `follows` was apparently forgotten[1].

Since the key of the `std::map` that holds the `overrides` is a vector
and we have to find the last element of each vector (i.e. the override)
this has to be done with a for loop in O(n) complexity with `n` being
the total amount of overrides (which shouldn't be that large though).

Please note that this doesn't work with nested expressions, i.e.

    inputs.fenix.inputs.nixpkgs.follows = "...";

which is a known problem[2].

For the expression demonstrated above, an error like this will be
thrown:

    error: sops-nix has a `follows'-declaration for a non-existant input nixpkgs_22_05!

[1] 2664a216e5
[2] https://github.com/NixOS/nix/issues/5790
2022-07-12 10:18:38 +02:00
Ben Burdette
cdcc349072
Merge branch 'master' into ignore-try 2022-07-11 11:29:22 -06:00
Ben Burdette
a3629ab0cc move ignore-try to EvalSettings 2022-07-11 10:47:09 -06:00
Ben Burdette
6ac8200ff5 use util.hh class instead of local 2022-07-11 10:21:40 -06:00
Ben Burdette
69ea265fd2 'tryEval' not 'try clause' 2022-07-11 10:21:12 -06:00
Gytis Ivaskevicius
b2703c73a4 builtins.traceVerbose: Post rebase fixes 2022-07-05 19:56:39 +03:00
Gytis Ivaskevicius
ba1fe85b65 Add builtins.traceVerbose
Co-Authored-By: Silvan Mosberger <contact@infinisil.com>

Add builtins.traceVerbose tests
2022-07-05 19:44:26 +03:00
Eelco Dolstra
455177cbe0 src/libexpr/tests/primops.cc: Quote Nix expressions
Otherwise they don't survive reformatting, see the failure in
https://github.com/NixOS/nix/pull/6721.
2022-06-29 11:37:25 +02:00
Guillaume Girol
e8109cf405 fetchGit: document shallow argument 2022-06-26 12:00:00 +00:00
Théophane Hufschmitt
f06b6fa4fa
Merge pull request #6698 from tweag/forbid-tilde-paths-in-pure-mode
Forbid the tilde expansion in pure eval mode
2022-06-24 10:02:40 +02:00
Rick van Schijndel
2beb929753 eval-cache: cast rowId to correct type
Prevents errors when running with UBSan:

/nix/store/j5vhrywqmz1ixwhsmmjjxa85fpwryzh0-gcc-11.3.0/include/c++/11.3.0/bits/stl_pair.h:353:4: runtime error: load of value 229, which is not a valid value for type 'AttrType'
2022-06-23 21:11:08 +02:00
Théophane Hufschmitt
983efdbde4 Forbid the tilde expansion in pure eval mode
Fix #6684
2022-06-21 14:08:18 +02:00
Alexander Bantyev
d82a3dc70d
flake.cc: Make non-flake overrides sticky
Overrides for inputs with flake=false were non-sticky, since they
changed the `original` in `flake.lock`. This fixes it, by using the same
locked original for both flake and non-flake inputs.
2022-06-13 20:49:16 +04:00
Ben Burdette
49ff4ef637 remove unused parameter 2022-06-10 12:22:36 -06:00
Eelco Dolstra
81a486c607
Shut up clang warnings 2022-06-02 21:19:54 +02:00
Ben Burdette
8cf6ae8664 use Counter class to count tryEval levels 2022-06-02 12:29:38 -06:00
Ben Burdette
bc0d41e9ba print message with exceptions in a try clause 2022-06-02 12:17:28 -06:00
Ben Burdette
9151dbff88 ignore-try flag 2022-06-02 10:26:46 -06:00
Eelco Dolstra
8e8e9d8705
Respect the outputSpecified attribute
E.g. 'nix build nixpkgs#libxml2.dev' will build the 'dev' output.
2022-05-30 11:34:47 +02:00
Eelco Dolstra
ec07a70979
Style fix 2022-05-27 11:25:05 +02:00
Eelco Dolstra
9acc770ce4
Remove pre-C++11 hackiness 2022-05-26 12:40:01 +02:00
Eelco Dolstra
762fa2b2ff
Merge pull request #6570 from edolstra/eof-in-string
Handle EOFs in string literals correctly
2022-05-26 08:13:39 +02:00
Ben Burdette
9a5ea6c359
Merge branch 'master' into debug-exploratory-PR 2022-05-25 10:41:10 -06:00
Ben Burdette
b4c24a29c6 back to ref<EvalState> in NixRepl 2022-05-25 10:21:20 -06:00
Eelco Dolstra
27ebb97d0a
Handle EOFs in string literals correctly
We can't return a STR token without setting a valid StringToken,
otherwise the parser will crash.

Fixes #6562.
2022-05-25 17:58:13 +02:00
Eelco Dolstra
91b7d5373a
Style tweaks 2022-05-25 12:32:22 +02:00