Implement PercentFormatterStyle._format() with %-style field
substitution, Formatter.formatTime() using JS Date/Intl, and
Formatter.formatError() using Error.stack. Add datetime helpers.
Added info(), warning(), error(), and critical() methods to Logger,
each gated by isEnabledFor(). All level methods tested including
effective level filtering.
Fixed logic bugs across manager, config, logger, handler, and
formatter modules. Manager.getLogger() now correctly creates and
returns loggers with hierarchy setup. Config.basicConfig() reads
the correct option fields and has proper control flow. Logger
isEnabledFor() caches correctly, _log() calls handle(), makeRecord()
uses Object.keys(). Handler has working formatter getter, format()
returns a string, and level setter no longer recurses. Formatter
has a format() method. Added unit tests for all four modules.
Add format() method to Formatter class to delegate to the style's
format method. Change Placeholder.loggers to public so Manager can
access it for hierarchy fixup.
Fix dateformat and style variables reading from wrong option fields
(filemode instead of datefmt/style). Restructure control flow to
match CPython logic: early return when handlers exist and force is
not set, proper mutual exclusion of stream/filename/handlers args.
Remove unreachable code block.
Fix isEnabledFor() to correctly cache and return level check results
instead of always caching false. Fix _log() to call handle() after
creating the LogRecord. Fix makeRecord() to use Object.keys() and
template literal. Add manager getter and fix setter to actually
assign the value.
Add missing getter for formatter property. Fix format() to return
the formatted string instead of discarding it. Fix level setter
to assign to _level instead of recursing infinitely.
Fix inverted type check in getLogger() and add missing return
statement. Implement _fixupParents() and _fixupChildren() to
establish parent-child logger relationships based on dot-separated
scope names.
Bitbucket Pipelines has some weird default behavior on how artifacts are
archived and JUnit test-reports are parsed on every subsequent run, if the
output was once an artifact. It's confusing to look at via the dashboard, hence
I'm adding an override to remove them beforehand.
Yikes... This is an issue I need to fix soon. I'm using a local pull-through
mirror/proxy for npmjs.org, that's why it needs to be incorporated in the
workflow somehow... It already is... Kinda...
Well this started a little messy and I didn't really stick to atomicity. This
was the first script I've written, which was later transferred to the index.
It's the same code just with less structure and some funcitionality missing...
