python310Packages.flake8-future-import: fix build

[wamserma]

Motivation for this change

xZise/flake8-future-import#24
ZHF: #144627
ping @NixOS/nixos-release-managers

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review pr 144650"
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 21.11 Release Notes (or backporting 21.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

[SuperSandro2000]

Please target staging

[risicle]

python310Packages.flake8-future-import building successfully here, macos 10.15.

[wamserma]

re-targeted to staging, should be backported after branch-off

[risicle]

@ofborg eval

[jonringer]

re-targeted to staging, should be backported after branch-off

If it's fixing regressions which already exist, they should be merged.

The "breaking changes to staging" block is to prevent new regressions from entering

[wamserma]

re-targeted to staging, should be backported after branch-off

If it's fixing regressions which already exist, they should be merged.

The "breaking changes to staging" block is to prevent new regressions from entering

It's an un-breaking PR, but as it includes a bump of python3.flake8 we have quite a number of rebuilds.

Let the @NixOS/nixos-release-managers decide where they prefer this PR. Feel free to re-target as required before merging.

[SuperSandro2000]

we have quite a number of rebuilds.

That does not matter for the breaking change policy. It is about not bumping compilers like gcc to a breaking version.

Let the @NixOS/nixos-release-managers decide where they prefer this PR. Feel free to re-target as required before merging.

jonringer is the release manager.

[SuperSandro2000]

@ofborg build python310Packages.flake8-future-import python310Packages.flake8 python39Packages.flake8-future-import python39Packages.flake8

[jonringer]

This is causing breakages with the manual. the flake8 bump will need to wait til after ZHF

https://github.com/NixOS/nixpkgs/runs/4137245989?check_suite_focus=true

[wamserma]

@jonringer (hopefully) fixed

[SuperSandro2000]

@ofborg eval
@ofborg build

[SuperSandro2000]

I think it is failing due to coreutils.

[wamserma]

I think it is failing due to coreutils.

Found this in the logs if the coreutils checks:

SKIP: tests/du/2g
=================

2g.sh: skipped test: very expensive: disabled by default
This test is very expensive, so it is disabled by default.
To run it anyway, rerun make check with the RUN_VERY_EXPENSIVE_TESTS
environment variable set to yes.  E.g.,

  env RUN_VERY_EXPENSIVE_TESTS=yes make check

or use the shortcut target of the toplevel Makefile,

  make check-very-expensive

./tests/init.sh: line 106: printf: write error: Broken pipe
SKIP tests/du/2g.sh (exit status: 77)

FAIL: tests/du/basic
====================

--- exp 2021-11-09 11:58:41.313614539 +0000
+++ out 2021-11-09 11:58:41.278614356 +0000
@@ -1,7 +1,7 @@
-9      d/sub/2
-10     d/sub
-9      d/1
-20     d
+1      d/sub/2
+2      d/sub
+1      d/1
+4      d
 ===
-10     d/sub
-10     d
+2      d/sub
+2      d
FAIL tests/du/basic.sh (exit status: 1)

So it might be an intermittent error.

[wamserma]

Rebased, now
@ofborg eval
@ofborg build

[wamserma]

@SuperSandro2000 fixed

[SuperSandro2000]

@jonringer I think we can merge this now. What do you think?

[SuperSandro2000]

This is a semi-automatic executed nixpkgs-review with nixpkgs-review-checks extension. It is checked by a human on a best effort basis and does not build all packages (e.g. lumo, tensorflow or pytorch).

Result of nixpkgs-review pr 144650 run on x86_64-linux 1

1 package failed to build and are new build failure:

• home-assistant: log was empty

---

verified manually that home-assistant builds

[trofi]

Does patch fetch work for you? Fails for me as:

$ nix build -f. python3Packages.pytest-flake8 -L

eda4ef74c0f25b856fe282742ea206b21e94c24c.patch> trying https://github.com/tholo/pytest-flake8/commit/eda4ef74c0f25b856fe282742ea206b21e94c24c.patch
eda4ef74c0f25b856fe282742ea206b21e94c24c.patch>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
eda4ef74c0f25b856fe282742ea206b21e94c24c.patch>                                  Dload  Upload   Total   Spent    Left  Speed
eda4ef74c0f25b856fe282742ea206b21e94c24c.patch> 100  3155  100  3155    0     0  37639      0 --:--:-- --:--:-- --:--:-- 37559
error: hash mismatch in fixed-output derivation '/nix/store/qnplqiw1cs5c73099virwa7i60krzq9z-eda4ef74c0f25b856fe282742ea206b21e94c24c.patch.drv':
         specified: sha256-/QBKi1nR5pCCEvODf4cmEyxdXihLLJgRUljDloKPLAA=
            got:    sha256-kfll+I9akZM4Vp62Q1M9r2Mvs2TM2Lm5mRkA3aDmAE8=
error: 1 dependencies of derivation '/nix/store/1vjdyv87g6fjdh93p9dascwy8p984xyy-python3.9-pytest-flake8-1.0.7.drv' failed to build

[wamserma]

@trofi When I build from current master it simply pulls /nix/store/2743gl0snph4z967yi1kqvaaiqp75cg6-python3.9-pytest-flake8-1.0.7 from the binary cache.

> nix-prefetch-url https://github.com/tholo/pytest-flake8/commit/eda4ef74c0f25b856fe282742ea206b21e94c24c.patch
[0.0 MiB DL]
path is '/nix/store/2h9wk9cicphjzpbpg0pp3za25mq5l4dw-eda4ef74c0f25b856fe282742ea206b21e94c24c.patch'
001ciy19dhsqa88rhb2b51g5sb0k4s3pz0zk2a191rnib65ll07x

[trofi]

nix-prefetch-url https://github.com/tholo/pytest-flake8/commit/eda4ef74c0f25b856fe282742ea206b21e94c24c.patch

Ah, that explains it. fetchpatch (as opposed to fetchurl) does mangle patch a bit before calculating the checksum.

Proposed hash fix as #145786

[wamserma]

Right, just didn't show up, because the patch was already fetched.