fix(partition): recognize addresses in Alabama and Wyoming (#4581)
## Summary
`US_CITY_STATE_ZIP_PATTERN` wrapped both state alternations in `{` and
`}`:
```
(?:{Alabama|Alaska|...|Wyoming}|{AL|AK|...|WY})
```
In a Python regex these braces are literal characters, so the first and
last entry of each list only matched with a brace attached (`{Alabama`,
`Wyoming}`, `{AL`, `WY}`). The other states were not affected. Removing
the four braces fixes it; the alternatives and their order stay the
same.
User impact: addresses in Alabama and Wyoming are now `Address`
elements. Before, `partition_text` returned `Title` and `partition_html`
/ `partition_docx` returned `Text`:
```
# input lines: "Birmingham, AL 35203", "Cheyenne, Wyoming 82001", "Boston, MA 02108"
text ['Title', 'Title', 'Address'] -> ['Address', 'Address', 'Address']
html ['Text', 'Text', 'Address'] -> ['Address', 'Address', 'Address']
docx ['Text', 'Text', 'Address'] -> ['Address', 'Address', 'Address']
```
No file in `test_unstructured_ingest/expected-structured-output`
contains an Alabama or Wyoming address, so the ingest fixtures do not
change.
## Validation
Four new `test_is_us_city_state_zip` cases (AL, Alabama, WY, Wyoming)
and one `partition_text` test fail on main and pass here.
```
$ pytest test_unstructured/partition/test_text_type.py test_unstructured/partition/test_text.py test_unstructured/partition/html test_unstructured/partition/test_docx.py -q
845 passed
```
## Checklist
- [x] The PR title is a user-facing release-note entry with a
conventional commit prefix.
- [x] Apply a category label (`enhancement`, `bug`, or an appropriate
existing label), or ask a maintainer to apply it.
- [ ] For breaking changes, use `!` in the title and explain impact and
migration above.
Signed-off-by: Dmitry Sergeev <sergeevdms@gmail.com>
unstructured
8f553b77 - fix(partition): recognize addresses in Alabama and Wyoming (#4581)
