Get Gentoo!
gentoo.org sites
gentoo.org Wiki Bugs Forums Packages
Planet Archives Sources
Infra Status
summaryrefslogtreecommitdiff
path: root/www-client/qutebrowser/qutebrowser-9999.ebuild
Commit message (Collapse)AuthorAgeFilesLines
* www-client/qutebrowser: add gherkin-official upper bound for nowIonen Wolkens5 days1-1/+4
| | | | | | Haven't had time to look into this yet, and mask is annoying. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: drop obsolete pytest-bdd upper bound in liveIonen Wolkens2024-12-061-2/+1
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* Rename dev-python/{PyQt6 → pyqt6}Michał Górny2024-11-231-2/+2
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* Rename dev-python/{PyQt6-WebEngine → pyqt6-webengine}Michał Górny2024-11-231-1/+1
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* www-client/qutebrowser: add issue link for pytest-bdd upper boundIonen Wolkens2024-11-141-0/+1
| | | | | | | | fwiw there is a PR that may fix this, albeit no urgency given no other packages use pytest-bdd, so a upper bound is pretty harmless until upstream merge&release. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: depend on <pytest-bdd-8Ionen Wolkens2024-11-141-1/+1
| | | | | | | | pytest-bdd changed its parser, and syntax used here is completly broken with it, upstream keeps up with pytest-bdd changes so it should be fixed sooner or later. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: add missing pillow test depIonen Wolkens2024-11-131-0/+1
| | | | | | | Meant to add this alongside 3.13 compat when noticed it was missing, but overlooked change wasn't queue'd when pushed. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: enable py3.13Ionen Wolkens2024-11-131-2/+1
| | | | | | | | Looking upstream, the only issue left is with some specific tests that we're not running and the failure does not indicate a runtime problem (it's the test itself that has random issues). Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* Rename dev-python/{jinja → jinja2}Michał Górny2024-11-111-1/+1
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* www-client/qutebrowser: note link for py3.13 supportIonen Wolkens2024-09-031-0/+1
| | | | | | Next release has some fixes, albeit unsure if complete yet. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: adjust tests skipping for Qt 6.7.2Ionen Wolkens2024-06-181-3/+5
| | | | | | | | | | | | | | version.py newly checks for chromium security version which fails given it does not know about 6.7.2's, and so just skip the whole suite given not very meaningful downstream. Also, Qt 6.7.2 introduced new warnings about failing to initialize vulkan and qutebrowser's tests fail if it sees unrecognized warnings. It might work if had mesa's lavapipe enabled in Xvfb but new warnings been a recurring issue and little sense in worrying further about these downstream. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: update liveIonen Wolkens2024-06-031-5/+2
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: limit pdfjs to <4.1 for nowIonen Wolkens2024-04-141-1/+4
| | | | | | | | | | Incompatible with the non-legacy sdist of 4.1.329 and will fail to load any pdf with type errors. Not that 4.1.329 is in-tree yet, just doing it before it breaks (revbumps are cheap for this package) and will revisit later. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: force-enable Qt6 in liveIonen Wolkens2024-03-271-55/+8
| | | | | | | | | | | | | | | | Still supported upstream, but qtwebengine:5 is slowly on its way out in Gentoo and offers a poor browsing experience with modern websites. Some may prefer Qt5 for better DE integration at the moment, but considering that these issues throne over that. Also drop the transition warning, we can rely on the warning that qutebrowser itself will do (formerly was missing, thus the ebuild warning was more useful). Plus given there is no USE to disable Qt6 anymore going back the other way is unlikely. Qutebrowser's default is Qt6 so we can also drop the overrides. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: warn about qtwebengine[bindist]Ionen Wolkens2024-02-091-0/+8
| | | | | | | | With the binhost gaining popularity, it may be good to warn about it. Not an issue for most packages, but losing e.g. h264 in a full web browser is not great and may leave users confused. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: fix casing typo in commentIonen Wolkens2024-01-091-1/+1
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: fix tests due to python-utils changesIonen Wolkens2024-01-081-3/+11
| | | | | | | | | | | Several tests fail due to `-o tmp_path_retention_policy=failed` from python-utils.eclass since commit b77499537549, so set back to the default (=all). The other setting (tmp_path_retention_count=0) does not seem to be a problem. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: drop x86 keywordIonen Wolkens2023-11-301-1/+1
| | | | | | | | Same issue as calibre, qutebrowser and qtwebengine are masked on x86 -- so let's just drop the keyword rather than pretend that it's still supported. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: revert to back to old verify-sig handlingIonen Wolkens2023-11-011-9/+1
| | | | | | | This was done for BROOT, but if BROOT no longer need to be passed then redefining src_unpack is unnecessary. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* Remove BROOT from VERIFY_SIG_OPENPGP_KEY_PATHMike Gilbert2023-10-311-1/+1
| | | | Signed-off-by: Mike Gilbert <floppym@gentoo.org>
* www-client/qutebrowser: fix VariableScopeIonen Wolkens2023-10-101-1/+9
| | | | | | Would probably be simpler if the eclass could handle BROOT by itself. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: skip chromium version testIonen Wolkens2023-09-161-0/+2
| | | | | | | | | | | | | Checks if the chromium version matches known versions recorded in the test. Upstream updates as needed but that means failure for any older qutebrowser version whenever there's a new major qtwebengine (aka, currently fails with 6.6.9999 which reports 112 rather than 108, but otherwise qutebrowser works just fine with it and upstream has updated this in qutebrowser-9999). Not really useful for us, let's just skip it rather than update. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: remove widevine workaround for prefixIonen Wolkens2023-09-051-12/+0
| | | | | | | | | | | | | | | Now prexified in qtwebengine:6 itself, albeit the rare users actually using this on prefix will need to manually rebuild qtwebengine to get the change if <=qtwebengine-6.5.2. Given this messed with tests did not want to keep this in qutebrowser, and it can benefit others. Technically don't need USE=widevine anymore, but prefer this over a optfeature at the moment and qutebrowser is fast to "rebuild" being python-only. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: enable py3.12Ionen Wolkens2023-08-211-4/+1
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: note prefix qtwebengine reminderIonen Wolkens2023-08-181-0/+1
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: note issue link for deleted testIonen Wolkens2023-08-181-1/+1
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: verify-sig supportIonen Wolkens2023-08-181-2/+12
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: more minor live adjustmentsIonen Wolkens2023-08-181-12/+17
| | | | | | | wrt qt6 ewarn, at this point don't want it to sound like Qt6 is not stable and Qt5 removal is not happening anytime soon. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: update/tidy liveIonen Wolkens2023-08-161-38/+44
| | | | | | | | | | | | | | No notable changes, but clarify/reword a few comments, call python_gen_cond_dep less times, skip a failing test with Qt5, and uncomment the Qt6 warning. Rest is just style. This is in preparation for the imminent qutebrowser-3 release, tests pass for both Qt5 and Qt6 in current state. May still be other small adjustments to do depending on how the release tarball is created. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: add note wrt py3.12 in liveIonen Wolkens2023-07-241-1/+4
| | | | | | | | | | | | | Last blocker is pyside2+6 to enable py3.12 on QtPy which is needed by pytest-qt, but tests on pyside are restricted and I do not use/follow either of these to be able to know if it works. If ignore the pyside deps and force enable 3.12 just to be able to run qutebrowser's tests, it passes fine. Leaving this on hold for now so don't need to restrict tests. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: depend on qtsql/qtbase[sqlite]Ionen Wolkens2023-07-231-1/+2
| | | | | | | | Fairly rare that users have anything else enabled (e.g. mysql), but sqlite is specifically needed to read qtwebengine/chromium's sqlite databases like history.sqlite. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: update liveIonen Wolkens2023-07-221-7/+8
| | | | | | | | | | | | | | Now defaults to Qt6 and auto-fallbacks to Qt5 if missing or broken, but there is little reason to use automagic in our case and it should use the backend ebuild set dependencies for (users can still override this at runtime if really wanted). Using sed is what machinery.py suggests. On a side-note, selected qt6 tests currently pass and qt5 has one failure. Will be reviewed further on release. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: update liveIonen Wolkens2023-03-181-9/+51
| | | | | | | qt6-v2 was merged in the main branch, so this syncs 9999 with the old 3.0.0_pre9999 ebuild plus a few adjustments Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* Revert "www-client/qutebrowser: install apparmor profile"Ionen Wolkens2023-03-111-3/+0
| | | | | | | | | | | | | | | | | | | This reverts commit de3188d3b33e3826a278b3426d5376311439e3dc. Adding this was a request but profile is apparently unmaintained upstream and installing it does not seem like a good idea if it's going to potentially crash the browser wrt bug #900861. From a quick look it seems like it'd need adjustment for the paths we uses too, but not limited to these (handling both slots of qtwebengine and gpu-related things it does seems non-trivial, lot of this profile feels like wishful guesswork). Users that really want this are invited to install/modify it manually for now. Bug: https://bugs.gentoo.org/900861 Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: install apparmor profileSergey Alirzaev2023-03-051-0/+3
| | | | | | Signed-off-by: Sergey Alirzaev <l29ah@riseup.net> Closes: https://github.com/gentoo/gentoo/pull/29898 Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: sync liveIonen Wolkens2023-02-171-28/+35
| | | | | | | | | | Somewhat sync 9999 with 3.0.0_pre9999 minus qt6. Reverted xdist usage since getting random failures, perhaps they are qt5-only but not going to worry about it given need to bump/test this too rarely to need the test suite to be speedy. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: cleanup py3.8 and importlib_resources depIonen Wolkens2023-01-141-3/+2
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: adjust asciidoc call for liveIonen Wolkens2022-12-171-1/+7
| | | | | | | | | | | | | Upstream now calls `python3 -m asciidoc` to simplify some venv/pip handling but, given asciidoc is single target, matching pythons over this would be inconvenient (aka fails if asciidoc is 3.10 and qutebrowser is 3.11). Alternatively could make asciidoc multi-target but this works just fine if just swap it out for ["asciidoc"]. May review this solution eventually but won't worry too much if it's for live-only. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: add IUSE=pdf to handle pdfjsIonen Wolkens2022-11-041-4/+8
| | | | | | | | | | | Unfortunately qtwebengine:5 doesn't fare so well with new pdfjs-3, while :6 has no issues. As a optfeature, handling this is rather unintuitive and older pdfjs may get dropped without a dependency check. This is just python and builds quickly, so add IUSE=pdf to control versions and also make content.pdfjs a default if enabled. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: Enable pytest-xvfb explicitlyMichał Górny2022-09-071-1/+1
| | | | Signed-off-by: Michał Górny <mgorny@gentoo.org>
* www-client/qutebrowser: force pyqt5 for testsIonen Wolkens2022-07-201-0/+2
| | | | | | | | | | Doing this now so don't forget, will be needed when PyQt6 is in the tree and technically ensures pyside2 is not used too. Version 3.0.0_pre will be able to "pyqt$(usex qt6 6 5)" for tests, this already works with local Qt6 ebuilds. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: enable py3.11 againIonen Wolkens2022-07-121-2/+1
| | | | | | No known issues with >=python-3.11.0_beta4 (randomly crash with beta3). Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: sync liveIonen Wolkens2022-06-221-0/+2
| | | | Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: disable py3.11Ionen Wolkens2022-06-111-1/+2
| | | | | | | | Rare but can randomly crash with 3.11b3 which can be quite annoying, believed to be a cpython beta bug rather than in qutebrowser (see ebuild comment). Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: enable py3.11Ionen Wolkens2022-06-081-1/+1
| | | | | | | | There was a known issue with PyQt+py3.11_beta1 but does not seem to occur anymore with 3.11_beta3. Still feels early to do this but tests pass, and seems to run fine. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: sync liveIonen Wolkens2022-05-261-3/+3
| | | | | | Minor style nitpicks Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: tidy live ebuild and use pep517Ionen Wolkens2022-01-201-14/+12
| | | | | | | | | | | | | | | QtWebKit test was fixed to properly ignore that we don't have it, albeit may still be present in a potential 2.4.1 release on a separate branch. Remove typing-extensions, was only needed for <python-3.8. Also move asciidoc dep to live-only (it is getting tempting to just maintain live and releases separately given the increasing amount of PV == 9999 blocks, which may get worse given considering adding verify-sig support too). Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: remove mention of youtube-dlIonen Wolkens2021-11-201-1/+1
| | | | | | | mpv now defaults to yt-dlp without further setting up, and yt-dlp also provides a youtube-dl wrapper either way. Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: soften pyyaml version restrictionIonen Wolkens2021-11-201-1/+1
| | | | | | | | Upstream added a test fix for 6.0 and a requirement, but it does not actually harm use with <6.0 (no related code change beside the test, which pass with 5.4.1). Signed-off-by: Ionen Wolkens <ionen@gentoo.org>
* www-client/qutebrowser: skip test_custom_env posix userscript testIonen Wolkens2021-10-221-0/+2
| | | | | Closes: https://bugs.gentoo.org/819393 Signed-off-by: Ionen Wolkens <ionen@gentoo.org>