-
-
Notifications
You must be signed in to change notification settings - Fork 704
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Backport changes since v1.0 to Python 2.7 #756
Conversation
* drop python 2.7 support * switch to os.fsencode / os.fsdecode on Unixes * updated tests * remove Python 2.7 from CI * removed obsolete __future__ imports * removed Py 2.7 references from docs * removed Py 2.7 references from readme * add test for path with invalid bytes * add entries to changelog
To handle cases with lot of changes, this seems the highest safest value we can use. Note: it will fail with `ERROR_INVALID_PARAMETER` when it is greater than 64 KB and the application is monitoring a directory over the network. This is due to a packet size limitation with the underlying file sharing protocols. https://docs.microsoft.com/en-us/windows/win32/api/winbase/nf-winbase-readdirectorychangesw#remarks Also introduced `winapi.PATH_BUFFER_SIZE` (defaults to `2048`) to keep the old behavior with path-realted functions.
They hang and timeout after 6 hours for some unknown reason
* Add compatibility with old macOS versions * Define symbols only for older versions And not for older or equal, since the equal version is when the symbol was introduced. * Ensure version macros are available * Update changelog
* monitor our own root * properly handle root change event * test on all platforms that root deleted event is emitted ... ... and that emitter is stopped * basic test for root moved * winapi: emit DirDeletedEvent of removed root * removed leftover commented-out code * expect KeyError when unscheduling removed watch * propagate root deleted event in inotify * remove `test_move_self` * catch OSError when stopping watch in fixture * adapt test_inotify_c to stopped emitter after delete_self * don't emit root deleted for child watches
* Actually use the input in the workflow * Formatting * Update description for input
…#746) Since CPython 3.7 `PyEval_InitThreads()` does nothing and the function has been deprecated since CPython 3.9.
* add: inotify file close events * fix: NOWRITE_CLOSE doesn't modify dir * update changelog & version * add: test events * Revert "update changelog & version" This reverts commit 69d61d0. * update changelog * [gnu/linux] Fix ability to catch IN_CLOSE_WRITE, IN_CLOSE_NOWRITE events gorakhargosh#690 * [gnu/linux] Test IN_CLOSE_WRITE, IN_CLOSE_NOWRITE events * Fix typos * [gnu/linux] Fix test due to IN_CLOSE_NOWRITE event * [gnu/linux] Revert IN_CLOSE_NOWRITE support gorakhargosh#747 * [gnu/linux] Extend test for IN_CLOSE_WRITE support gorakhargosh#747 * fix typos * Update changelog gorakhargosh#747, gorakhargosh#690 Co-authored-by: Lukas Šupienis <l.supienis@ncs.lt> Co-authored-by: Lukas Šupienis <lukas.supienis@gmail.com>
* Add `is_coalesced` property to `NativeEvent` So that we can effectively decide if we need to perform additional system calls to figure out what really happened. * Replace `NativeEvent._event_type` with `repr()` support It's more pythonic, and the `_event_type` implementation wasn't quite usable anyway. NB: the representation is not truly copy/paste python code if there is a double quote inside event.path, but that should be a rare case so we don't add the expensive special case handling there. * Allow running tests with debugger attached Some Python debuggers create additional threads, so we shouldn't assume that there is only one. * Request notifications for watched root * Expect events on macOS instead of using `time.sleep()` It might be even better to check for the emitter class, as opposed to platform * Add exception handling to FSEventsEmitter Reduce the amount of 'silent breakage' * Use sentinel event when setting up tests on macOS So that we can avoid a race between test setup and fseventsd * Improve handling of coalesced events * Revert accidental platform check change * Fix renaming_top_level_directory test on macOS * Generate sub events for move operations * Remove `filesystem_view` again While the `filesystem_view` helps with filtering out additional `FileCreatedEvent`+`DirModifiedEvent` pairs then it also introduces a huge amount of edge cases for synthetic events caused by move and rename operations. On top of that, in order to properly resolve those edge cases we'd have to go back to a solution very similar to the old directory snapshots, with all the performance penalties they suffered from... As such I think it's better to acknowledge the behaviour for coalesced events instead, and thus remove the `filesystem_view` again. * Update Changelog
* drop support for macOS 10.12 and lower * record inodes for all events * use inode to match rename events * add flush method * update tests to new NativeEvent * simplify handling of created and removed events * added test for case-change * clean up loop over native events Co-authored-by: Mickaël Schoentgen <contact@tiger-222.fr> * added changelog entries * remove unneeded Py_INCREF calls * Revert "simplify handling of created and removed events" This reverts commit 75afbd2. * allow inode to be None this is the case for kFSEventStreamEventFlagRootChanged * join event_dispatcher in `test_event_dispatcher` * fix memory management for inode * better logs for macOS tests * show flags in log output * fix returning inode attribute * tweak error message for unsupported macOS Co-authored-by: Mickaël Schoentgen <contact@tiger-222.fr> * emit debug instead of info logs from `queue_events` Co-authored-by: Mickaël Schoentgen <contact@tiger-222.fr>
Backports almost all changes from the current 1.0.x to python-2.7. Notable exceptions are: - Wheels are not built yet as the CI pipeline hasn't been adapted yet - Non-file-system-encoded paths on UNIX might not be working as intended, need to test it.
We want to know about all failures.
This should get pretty close to what Python 3's `os.fsencode` / `os.fsdecode` functions do.
This reverts commit b1c1b28. The `surrogateescape` error handler doesn't exist until Python 3.
Unfortunately gorakhargosh#703 cannot easily be backported, so we won't have this functionality in the Python 2.7 / 3.5 branch unless someone reports this as an issue.
Outch, this is a huge one. Part of those changes are not Python 2.7 compatible. |
As stated in the description, this back ports all changes except for the commits in #703 and anything that happened since this PR was opened. :-) The CI for it is (mostly) passing, too, with exception of what appears to be an ordering issue in the PollingObserver test and some unknown pypy-2.7 issues on macOS. |
OK, I'll review then :) |
As pointed out in the PR review, this will be updated for Python 2.7 in the master branch
Thanks a lot for the review and spotting the few spots that I missed while back porting! :-) |
As per code review feedback
Using a more descriptive name as suggested in the PR review. Co-authored-by: Mickaël Schoentgen <contact@tiger-222.fr>
@CCP-Aporia I guess you tested the python-2.7 branch? Before doing a release, I would like you input :) |
Thanks for merging! :-) |
Should cover about everything, except for #703 because Python 2.7 doesn't provide required functionality.
A couple of notes: