j5basic

Settings | Report Duplicate

0

I Use This!

Very Low Activity

Commits : Listings

Analyzed about 19 hours ago. based on code collected about 19 hours ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Feb 13, 2025 — Feb 13, 2026 Showing page 6 of 15 Search / Filter on:
Be a bit smarter about stripping attributes: just get each node, and drop the attributes if they exist. (See #8984)	roryg	More...				over 12 years ago
Handle stripping out of lang and xml:lang attributes. (See #8984)	roryg	More...				over 12 years ago
Add some debug logging for the database lock.	matth	More...				almost 13 years ago
Update to handle cleaning in one pass, and so that we don't completely strip out style attributes. (See #8984)	roryg	More...				almost 13 years ago
Moved changes over to CleanXHTML, fixed an import, and made sure the output is consistent. (See #8984)	roryg	More...				almost 13 years ago
Better log message which is unlikely to panic anyone See #9076	davidm	More...				almost 13 years ago
Catch a RuntimeError (size change during iteration) in purge, so purge can fail gracefully See #9076	davidm	More...				almost 13 years ago
If we're trying to resolve an attribute of a module that exists, then it's alright to log a debug message and carry on. Otherwise, this is an attempt to import a module that's failing - we shouldn't swallow that	davidf	More...				almost 13 years ago
Moved the encoding and unencoding to before the interaction with tidylib - which gets very confused with utf-8 encoded strings instead of unicode strings. Also made tidylib required. It's caused us trouble in the past when tidylib failed to import with only a warning being logged - leading to dirty data in the database that broke j5. Moved the actual cleaning into a unit-testable method.	matth	More...				almost 13 years ago
Previously, the cleaning of rich text would result in extra tags and things getting added (e.g. an extra div around all of the content, if there wasn't a single containing node, and extra <p> tags around plain text strings). This changes the implementation to wrap the content in a div, before passing it to lxml to clean, and then unwrapping again afterwards. Also removed support for cleaning xml objects instead of html fragments.	matth	More...				almost 13 years ago
Remove the bytes import, and replace the use of it with `str`.	roryg	More...				about 13 years ago
Override lxmls' Cleaner so that we don't end up breaking our richtext tags. (See #8984)	roryg	More...				about 13 years ago
Fix possible memory leak when not accessing Time Cache Rows	steveh	More...				about 13 years ago
Rename the ContextLogging to LoggingTest and move it into its own module	davidm	More...				about 13 years ago
Replace our instrumented logging with way more reliable MemoryHandler - hopefully this will stop the errors on jenkins with not getting certain logs	davidm	More...				about 13 years ago
Allow a bit more time in our timedLock tests, to account for slow jenkins	davidm	More...				about 13 years ago
Added an alias to 'no_database_writes' called 'self_managed_database_lock'. It makes it more expressive, when we are managing the lock explicitly.	matth	More...				over 13 years ago
If startup is not called when DatabaseWriteLock is asked for, assume we're in a test environment	davidm	More...				over 13 years ago
Reimplement a safer safestrptime, although this may block for a while if another thread hogs the import lock for ages	davidm	More...				over 13 years ago
Duplicate the setup_method and teardown_method calls in nose-style	davidm	More...				over 13 years ago
Add an option to assert_dicts_equal to allow comparing naive and aware datetimes safely	davidm	More...				over 13 years ago
Add an error log if the database release occurs but we're not the thread with the lock, and test for it	davidm	More...				over 13 years ago
Don't kill threads, just rob them. (We discovered that killing was unreliable and nasty; we can just take the database lock but warn in the same way) Also, the instrumented_logging class didn't know about critical or debug messages...	davidm	More...				over 13 years ago
Use the same timeout check to decide max timeout - i.e. base it on the length of the busy operation, not the amount of time the next-in-line has waited (since starting). Make kill and warn checks independent	davidm	More...				over 13 years ago
Compare on the basis of `start_time` rather, and don't keep track of the `busy_op` unneccessarily	davidm	More...				over 13 years ago
Record any exceptions that have already happened when we try to join, and add a TODO	davidm	More...				over 13 years ago
Use non-virtual time for timing thread operations	davidm	More...				over 13 years ago
Change tests to handle threads by object rather than identifier	davidm	More...				over 13 years ago
We don't actually need to store the `thread_id`, as its only used in exceptional circumstances. Rename `thread_instance` to `thread` too...	davidm	More...				over 13 years ago
Added some documentation, and changed the busyop index constants to use an object rather	davidm	More...				over 13 years ago

←
1
2
3
4
5
6
7
8
9
10
…
14
15
→