colin/nixpkgs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #297585 from r-ryantm/auto-update/python312Packages.trafilatura

Browse Source 
python312Packages.trafilatura: 1.7.0 -> 1.8.0
This commit is contained in:
Fabian Affolter 2024-03-21 09:39:07 +01:00 committed by GitHub
commit 48895ecb59
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 22 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
pkgs/development/python-modules/trafilatura/default.nix
View File

@ -10,20 +10,25 @@
, justext
, lxml
, urllib3
, setuptools
}:
buildPythonPackage rec {
pname = "trafilatura";
version = "1.7.0";
format = "setuptools";
version = "1.8.0";
pyproject = true;
disabled = pythonOlder "3.6";
disabled = pythonOlder "3.9";
src = fetchPypi {
inherit pname version;
hash = "sha256-oWbmfwBaahLvGU9Ix8n6ThsONnVv3Stk4CRzw1aWLwQ=";
hash = "sha256-6lSHXtJPPq+vGZuKD4m1g1x880NzPDLvvEr50wV6j3I=";
};
nativeBuildInputs = [
setuptools
];
propagatedBuildInputs = [
certifi
charset-normalizer
@ -34,10 +39,12 @@ buildPythonPackage rec {
urllib3
];
nativeCheckInputs = [ pytestCheckHook ];
nativeCheckInputs = [
pytestCheckHook
];
# disable tests that require an internet connection
disabledTests = [
# Disable tests that require an internet connection
"test_download"
"test_fetch"
"test_redirection"
@ -51,18 +58,22 @@ buildPythonPackage rec {
# patch out gui cli because it is not supported in this packaging
# nixify path to the trafilatura binary in the test suite
postPatch = ''
substituteInPlace setup.py --replace '"trafilatura_gui=trafilatura.gui:main",' ""
substituteInPlace tests/cli_tests.py --replace "trafilatura_bin = 'trafilatura'" "trafilatura_bin = '$out/bin/trafilatura'"
substituteInPlace setup.py \
--replace-fail '"trafilatura_gui=trafilatura.gui:main",' ""
substituteInPlace tests/cli_tests.py \
--replace-fail "trafilatura_bin = 'trafilatura'" "trafilatura_bin = '$out/bin/trafilatura'"
'';
pythonImportsCheck = [ "trafilatura" ];
pythonImportsCheck = [
"trafilatura"
];
meta = with lib; {
description = "Python package and command-line tool designed to gather text on the Web";
mainProgram = "trafilatura";
homepage = "https://trafilatura.readthedocs.io";
changelog = "https://github.com/adbar/trafilatura/blob/v${version}/HISTORY.md";
license = licenses.gpl3Plus;
license = licenses.asl20;
maintainers = with maintainers; [ jokatzke ];
mainProgram = "trafilatura";
};
}