aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2016-10-14cygwin, #533: FIX daemon launchingKostis Anagnostopoulos
+ Rework git-daemon launching with `with` resource-management. + cmd: add `is_cygwin` optional override kwd on `Git.polish_url()`. - Cygwin TCs failing: - PY2: err: 13, fail: 3 - PY3: err: 12, fail: 3
2016-10-14cygwin, #533: Try to make it work with Cygwin's Git.Kostis Anagnostopoulos
+ Make `Git.polish_url()` convert paths into Cygwin-friendly paths. + Add utility and soe TCs for funcs for detecting cygwin and converting abs-paths to `/cygdrive/c/...`. - Cygwin TCs failing: - PY2: err: 14, fail: 3 - PY3: err: 13, fail: 3
2016-10-14appveyor, #533: enable CYGWIN TCs without failingKostis Anagnostopoulos
- Cygwin TCs failing (start, no Cygwin specific code): - PY2: err: 44, fail: 0 - PY3: err: 13, fail: 0
2016-10-14Merge pull request #530 fixing some #525 Windows errorsKostis Anagnostopoulos
+ git-daemon: + Use git-daemon PORT above 10k; on Windows all below need Admin rights. + Used relative daemon-paths with `--base-pth`. + Simplify git-daemon start/stop/ex-hanlding. +FIXED git-daemon @with_rw_and_rw_remote_repo(): + "Polish" most remote & config urls, converting \-->/. + test_base.test_with_rw_remote_and_rw_repo() PASS. + Remote: + test_remote: apply polish-urls on `_do_test_fetch()` checking function. + test_remote.test_base() now freezes on Windows! (so still hidden win_err). pump fetch-infos instead of GIL-reading stderr. + Push-cmd also keep (and optionally raise) any error messages. + `cmd.handle_process_output()` accepts null-finalizer, to pump completely stderr before raising any errors. + test: Enable `TestGit.test_environment()` on Windows (to checks stderr consumption). + util: delete unused `absolute_project_path()`. + Control separately *freezing* TCs on Windows with `git.util.HIDE_WINDOWS_FREEZE_ERRORS` flag.
2016-10-14cmd, #525: Always include stdout+stderr in exceptionsKostis Anagnostopoulos
+ Ignore `with_extended_output` arg when reaising the exception, keep its behavior when `status==0`.
2016-10-14remote, #525: FIX BUG push-cmd misses error messagesKostis Anagnostopoulos
+ Bug discovered after enabling TC in prev commit and rework of fetch. + remote_tc: unitestize assertions. + util: DEL unused `_mktemp()`.
2016-10-13remote, #525: pump fetch-infos instead of GIL-read stderrKostis Anagnostopoulos
+ `handle_process_output()` accepts null-finalizer, to pump completely stderr before raising any errors. + test: Enable `TestGit.test_environment()` on Windows (to checks stderr consumption).
2016-10-13test, #525: allow disabling freeze errors separatelyKostis Anagnostopoulos
+ cmd: use DEVNULL for non PIPEs; no open-file. + TCs: some unitestize-assertions on base & remote TCs.
2016-10-12config, #525: polish more config-urlsKostis Anagnostopoulos
2016-10-12daemon, #525: simplify exception handlingKostis Anagnostopoulos
2016-10-12daemon, #525: FIX remote urls in config-filesKostis Anagnostopoulos
+ Parse most remote & config-urls \-->/. + Used relative daemon-paths. + Use git-daemon PORT above 10k; on Windows all below need Admin rights. +FIXED git-daemon @with_rw_and_rw_remote_repo(): + test_base.test_with_rw_remote_and_rw_repo() PASS. + test_remote.test_base() now freezes! (so still hidden win_err) + repo_test: minor finally delete test-repos created inside this repo. + util: delete unused `absolute_project_path()`.
2016-10-12repo, cmd: DROP UNEEDED Win path for chcwd & check for '~' homedirKostis Anagnostopoulos
+ Do not abspath twice when contructing cloned repo. + Add `git.repo.base` logger.
2016-10-12pumps: FIX don't pump when proc has no streamsKostis Anagnostopoulos
2016-10-12build: run codecov on AppveyorKostis Anagnostopoulos
[travisci skip]
2016-10-12Add Guyzmo into AUTHORSGuyzmo
2016-10-12remote, #528: Fix regression shadowing exceptionsGuyzmo
2016-10-11remote, #528: fix prev cmt, Git<2.7 miss `get-url`Kostis Anagnostopoulos
2016-10-11remote, #528: Improved way of listing URLsGuyzmo
+ Instead of using `git remote show` that may triggers connection to remote repo, use `git remote get-url --all` that works by only reading the `.git/config`. + Change should have no functional impact, so no test needed. + Works only with git -2.7+. Signed-off-by: Guyzmo <guyzmo+github@m0g.net>
2016-10-11hidden win_errors: mark also git-daemon errors failingKostis Anagnostopoulos
2016-10-11ci: print python/git versions before starting buildKostis Anagnostopoulos
2016-10-11remote: unfix fetch-infos paring of 8a2f7dce4(pydev fixes)Kostis Anagnostopoulos
+ Mark another TC failing when not in master.
2016-10-11fix(travis): increase ulimitSebastian Thiel
Now that performance tests are run, it appears we run into one particular failure on travis, possibly indicating a bug in python 3.3. Just bluntly increason the amount of handles might silence it... . Related to #524
2016-10-11imp(performance): execute performance tests on travisSebastian Thiel
Fixes #524
2016-10-11FIX #526: Do not depend on test-sourcesKostis Anagnostopoulos
+ Move `HIDE_WINDOWS_KNOWN_ERRORS` flag from `git.test.lib.helper-->git.util`; regular modules in main-sources folder also depend on that flag. + Use unittest.SkipTest instead of from non-standard `nose` lib.
2016-10-09doc(README): add codecov badgeSebastian Thiel
[skip ci]
2016-10-09Merge pull request #522 from yarikoptic/enh-codecovSebastian Thiel
RF: coveralls (not used/relied on really) -> codecov
2016-10-04config: FIX regression by prev commit "wrap all conf..."Kostis Anagnostopoulos
+ Bug appeared as last 5 TCs (test_commit & test_stream) said: OSError: [WinError 6] The handle is invalid
2016-10-04io: Wrap (probably) allconfig_writers in `with` blocksKostis Anagnostopoulos
2016-10-04src: No PyDev warningsKostis Anagnostopoulos
+ Mark all unused vars and other non-pep8 (PyDev) warnings + test_utils: + enable & fix forgotten IterableList looped path. + unittestize all assertions. + remote: minor fix progress dispatching unknown err-lines
2016-10-04hidden win-errs: Let leaking TCs run till end, then hideKostis Anagnostopoulos
+ Detect code breaking the body of TCs eventually hidden win-errors by raising SkipTest ALAP. + submodule.base.py: import classes from `git.objects` instead of `utils`. + had to ++ ulimit 100->110 for the extra code tested (more leaks :-) + Centralize is_win detection.
2016-10-03doc: Explain Windows compatibility status, mention #525 asking for helpKostis Anagnostopoulos
2016-10-03Appveyor: Set HIDE_WINDOWS_KNOWN_ERRORS=TrueKostis Anagnostopoulos
+ Update error-conditions for PY-versions. + The purpose is to have NO TC FAILURES (with the minimum possible conditions).
2016-10-03Appveyor: Add and set HIDE_WINDOWS_KNOWN_ERRORS=Falsewinerr_showKostis Anagnostopoulos
+ Collect all "acknowledged" failing TCs on Appveyor and use "HIDE_WINDOWS_KNOWN_ERRORS" var to hide them.
2016-10-03repo-TC: FIX HOME-expansion check, Appveyor has no HOME varKostis Anagnostopoulos
2016-10-03ci: restore verbosity for travis/appveyor, increase 96->100 ulimitKostis Anagnostopoulos
+ PY3.3 fails due to 'too many files open"
2016-10-03"Null" merge yarik/bf-rsubmodule-remove to close #521Kostis Anagnostopoulos
+ The actual commits have been re-written and rebased previously.
2016-10-03TST: Add test for removing submodule with remotes without refsYaroslav Halchenko
2016-10-03BF: Allow to remove a submodule with a remote without refsYaroslav Halchenko
2016-10-03FIX config-lock release early regression caused by #519Kostis Anagnostopoulos
+ Regression introduced in d84b960982b, by a wrong comment interpretation.
2016-10-03src, config_tc: replace deprecated `failUnlessRaises`Kostis Anagnostopoulos
2016-10-02BF: log.info is a function, just pass msg, no .write!Yaroslav Halchenko
2016-10-02BF: @with_rw_directory must return decorated callYaroslav Halchenko
As it was - many tests were simply not accounted/run at all
2016-10-02Merge pull request #523 from yarikoptic/enh-wrapsKostis Anagnostopoulos
RF: use @functools.wraps within decorators instead of manual __name__ reassignment
2016-10-02RF: use @functools.wraps within decorators instead of manual __name__ ↵Yaroslav Halchenko
reassignment @wraps does more and does it right ;)
2016-10-02RF: coveralls (not used/relied on really) -> codecovYaroslav Halchenko
codecov in our (datalad, etc) experience provides a better service, great support, and super-nice intergration with chromium and firefox for reviewing coverage of pull requests. In light of the @with_rw_directory fiasco detected/fixed in #521 I would strongly recommend to (re-)enable and use coverage reports
2016-10-02TST: finishing test for removing submodule with remotes without refsYaroslav Halchenko
originally draft committed by mistake in 31fd955dfcc8176fd65f92fa859374387d3e0095 sorry
2016-10-02BF: log.info is a function, just pass msg, no .write!Yaroslav Halchenko
2016-10-02BF: @with_rw_directory must return decorated callYaroslav Halchenko
As it was - many tests were simply not accounted/run at all
2016-10-02FIX regression by #519 on reading stdout/stderr of cmdsKostis Anagnostopoulos
2016-10-01BF: Allow to remove a submodule with a remote without refsYaroslav Halchenko