Changelog¶

All notable changes to this project will be documented in this file.

0.2.1¶

README examples are easier to read: one call per line, grouped by task, and without >>> prompts.

The to_* converters in camina.convert that the README described.
drop_dunders and drop_privates for sets and tuples.
A full test suite (including tests that run the README examples), docs pages, and typed Args in all docstrings.

Updated to snickerdoodle v0.3.3: uv and hatchling replace pdm, workflows were consolidated, and mypy and pydoclint are used in CI.
Dropped support for Python 3.10. Python 3.11 or later is required.
miller is no longer a dependency. It depended on camina, which made a circular dependency.
listify and tuplify convert non-str collections (tuples, sets, generators, etc.) and no longer split strings.
drop_dunders only drops names that both start and end with double underscores.
iterify returns its input if it is already iterable.

The container classes (Bunch, Dictionary, Catalog, ChainDictionary, Repository, Listing, Hybrid, Proxy, Descriptor, and Name) and their helpers (resolve_default, set_key_namer, get_key_namer, set_method_namer, and get_method_namer). They now live in the separate bunches package.

import camina failed because camina.__init__ imported a nonexistent Base class.
how_soon_is_now raised an error (tz_info argument), and timer could not be used as a decorator.
cleave split incorrectly when return_last was False; uniquify and drop_suffix (with a divider) returned wrong results.
numify raised on None and truncated floats; tuplify split strings; pathlibify failed on strings; windowify did not validate arguments until iterated.
add_slots failed for classes with a __weakref__ slot.

Initial Commit