…ument (#2617)

* First function is working: is_alnum.

* is_alpha

* is_decimal

* is_lower

* is_digit

* is_numeric

* is_printable

* is_space

* is_upper

* is_title

* is_ascii; done with string predicates

* capitalize

* lower

* upper

* upper

* title

* T -> T operations on bytestrings should return bytestrings.

* repeat (the first that needs a broadcastable argument)

* reverse (because it's easy)

* replace_slice

* replace_substring

* Also test 'max_replacements' in replace_substring.

* replace_substring_regex: done with string transforms

* center

* lpad and rpad

* trim

* trim_whitespace

* ltrim

* rtrim

* rtrim_whitespace

* ltrim_whitespace

* slice

* feat: add `split_whitespace`

* test: add test for `split_whitespace`

* test: correct test

* feat: add `split_pattern`

* refactor: rename `_get_action`

* feat: add `ak_split_pattern_regex`

* test: update tests for new features

* Fixed UnmaskedArray._drop_none.

* fix: adjust for numexpr 2.8.5, which hid getContext's frame_depth argument (#2617)

* extract_regex.

* join (almost entirely from https://gist.github.com/agoose77/28e5bb0250678e454356a85861a16368)

* use dispatch correctly

* fix: drop unused arg

* join_element_wise

* Revert "use dispatch correctly"

This reverts commit 559073b.

* fix: broadcast `num_repeats`

* feat: add `count_substring[_pattern]`

* docs: fixup docstring

* feat: add `ends_with`

* feat: add `starts_with`

* docs: fix link

* feat: add `find_substring`

* docs: fix typo

* feat: add `find_substring_regex`

* docs: fix link

* feat: add `match_like`

* test: improve test

* feat: add `match_substring`, `match_substring_regex`

* feat: add `is_in` and `index_in`

* fix: operate at leaf depth

* refactor: add internal `pyarrow.compute` helper

* refactor: use pyarrow import helper

* refactor: add `module` and `name` arguments to `high_level_function`

* fix: pass `module` to str `high_level_function`

* docs: homogenize docstrings

* docs: add see also

* docs: include `ak.str` in toctree

* chore: update pre-commit hooks (#2619)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* refactor: cleanup error handling

* Rename ak_*.py modules -> akstr_*.py.

---------

Co-authored-by: Angus Hollands <goosey15@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

…rame (#2735)

* fix: backport numba old to new style error capturing to avoid warnings

* backporting #2617

* backporting #2670 and tests that depend on Parquet list separator

* style: pre-commit fixes

* backporting #2617 for the _v2 subdirectory

* RDataFrame handling is broken in _v2 subdirectory; don't bother fixing because it's correct in the real Awkward 2.x

* The next awkward 1.x backport will be version 1.10.5.

* Windows can't build ('x86' vs 'x86_64'). Maybe use the old VM?

* No, change that back.

* Remove all v2 tests.

* I was changing the wrong workflow file. Let's try new Windows.

* What I really need to test is the wheel-building.

* Try again.

* On second thought, don't upload to PyPI.

* Remove 32-bit Windows from build-test.yml and modernize wheels.yml.

---------

Co-authored-by: Jim Pivarski <jpivarski@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>