rpms/python-cramjam
4
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Compare commits

merge into: rpms:rawhide
Branches Tags
rpms:rawhide
rpms:epel10
rpms:f43
rpms:main
rpms:epel10.1
rpms:epel10.0
rpms:f41
rpms:f42
rpms:f40
rpms:f39
rpms:f38
...
pull from: rpms:f41
Branches Tags
rpms:epel10
rpms:f43
rpms:main
rpms:rawhide
rpms:epel10.1
rpms:epel10.0
rpms:f41
rpms:f42
rpms:f40
rpms:f39
rpms:f38
Sign in to create a new pull request.

6 commits
rawhide ... f41

Author SHA1 Message Date
Benjamin A. Beasley
655812e7c9 Remove conditional for F42+; the branches diverged 2025-05-04 06:51:29 -04:00
Benjamin A. Beasley
1c354a3120 Report and skip test that fails with miniz_oxide>0.8.5 2025-05-04 06:50:51 -04:00
Benjamin A. Beasley
2015ea4d69 Add a tomcli version bound, matching F42+ 
[skip changelog]
 2025-05-04 06:48:54 -04:00
Benjamin A. Beasley
48e07eecc6 Expect maturin to handle license files 2025-05-04 06:45:31 -04:00
Benjamin A. Beasley
af1aea506d Don’t need to manually export RUSTFLAGS 
[skip changelog]
 2025-05-04 06:44:55 -04:00
Benjamin A. Beasley
e0d4568742 No longer need to patch pyproject.toml to keep maturin from stripping binaries 
We were failing to adjust profile.release.strip correctly anyway; it’s in Cargo.toml
 2025-05-04 06:44:49 -04:00
1 changed files with 6 additions and 21 deletions
Download patch file Download diff file Expand all files Collapse all files

27
python-cramjam.spec
View file

@ -38,7 +38,7 @@ Source1: get_source
Patch: %{url}/pull/196.patch
BuildRequires: python3-devel
BuildRequires: tomcli
BuildRequires: tomcli >= 0.8.0
BuildRequires: cargo-rpm-macros >= 24
%if %{with tests}
@ -81,11 +81,6 @@ License: %{shrink:
%prep
%autosetup -n cramjam-%{srcversion}
# Do not strip the compiled Python module; we need useful debuginfo. Upstream
# set this intentionally, so this makes sense to keep downstream-only.
tomcli set pyproject.toml false 'tool.maturin.strip'
tomcli set pyproject.toml false 'profile.release.strip'
# Downstream-only: patch out the generate-import-lib feature, which is only
# relevant on Windows, and which depends on the corresponding pyo3 feature –
# which is not packaged for that reason.
@ -136,7 +131,6 @@ done
%build
export RUSTFLAGS='%{build_rustflags}'
%cargo_license_summary
%{cargo_license} > LICENSES.dependencies
%pyproject_wheel
@ -144,7 +138,7 @@ export RUSTFLAGS='%{build_rustflags}'
%install
%pyproject_install
%pyproject_save_files cramjam
%pyproject_save_files -l cramjam
%check
@ -173,25 +167,16 @@ k="${k-}${k+ and }not test_variants_different_dtypes[blosc2]"
# https://github.com/milesgranger/cramjam/issues/190
k="${k-}${k+ and }not (test_variants and [blosc2)"
%if 0%{?fedora} > 41
# Test failures in test_variants_decompress_into with recent hypothesis
# versions
# https://github.com/milesgranger/cramjam/issues/201
#
# It is hard to be really sure what is going on here. The failures are
# concerning, and might (or might not) reflect a serious problem. Nevertheless,
# since the problem appears to be linked to newer hypothesis versions, there’s
# not reason to believe that the package has *new* problems. It *might* have
# newly *revealed* problems. This merits further investigation.
k="${k-}${k+ and }not test_variants_decompress_into"
%endif
# Regression with miniz_oxide>0.8.5
# https://github.com/milesgranger/cramjam/issues/211
# Failed: DID NOT RAISE <class 'cramjam.DecompressionError'>
k="${k-}${k+ and }not test_variants_raise_exception[deflate]"
%pytest -k "${k-}" -v -n auto
%endif
%files -n python3-cramjam -f %{pyproject_files}
%license LICENSE LICENSES.dependencies
%doc README.md