Unicode 18.0.0 support, releases today

[jquast]

We expect Unicode 18.0 to release today.

• Match new 18.0.0 UAX #20 rules on grapheme clustering InCB/Indic_Conjunct_Break
• parse names from unicode.org data instead of current python
• wcstwidth() correction table updated with latest ucs-detect results (contour and rio are better-conforming)

[codspeed]

Merging this PR will degrade performance by 32.31%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

❌ 21 regressed benchmarks
✅ 87 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
❌	test_iter_graphemes_combining	76.4 µs	285.9 µs	-73.29%
❌	test_iter_graphemes_emoji	42.5 µs	112.3 µs	-62.16%
❌	test_iter_graphemes_ascii	46.2 µs	107.4 µs	-56.96%
❌	test_clip_cursor_cr_control_codes_ignore	341.5 µs	604.2 µs	-43.48%
❌	test_clip_cursor_cr_overwrite	1.8 ms	2.8 ms	-37.35%
❌	test_clip_long_cjk_overtype	7.5 ms	11.6 ms	-35.43%
❌	test_clip_long_cjk_past_window	326.7 µs	477.6 µs	-31.6%
❌	test_clip_cursor_csi_backward	571.7 µs	833.8 µs	-31.43%
❌	test_clip_long_cjk_control_codes_ignore	329.8 µs	479.9 µs	-31.29%
❌	test_clip_complex_sgr_no_overtype	180.4 µs	244.1 µs	-26.09%
❌	test_clip_complex_sgr	186.8 µs	251.8 µs	-25.81%
❌	test_clip_dense_ansi_no_overtype	419.3 µs	536.9 µs	-21.9%
❌	test_clip_dense_ansi_control_codes_ignore	419 µs	536.4 µs	-21.88%
❌	test_clip_japanese	97.6 µs	124.3 µs	-21.49%
❌	test_clip_dense_ansi_overtype	547.3 µs	657.6 µs	-16.77%
❌	test_clip_with_ansi_no_propagate	101.1 µs	121.3 µs	-16.72%
❌	test_wcstwidth_term_program[ghostty]	862.2 µs	1,032.9 µs	-16.53%
❌	test_width_term_program[ghostty]	915.8 µs	1,086.8 µs	-15.73%
❌	test_clip_dense_ansi_past_window	656 µs	774 µs	-15.25%
❌	test_wrap_japanese	5.8 ms	6.6 ms	-11.18%
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing jq/unicode-18 (36f165b) with jq/libwcwidth-rc (1042e5f)}

[jquast]

The remaining test failure is because the 3.14 step also runs "tox -e fetch" but without the new --draft argument that was used, it can be ignored.

Unicode 18 releases today, so this could be merged and released as soon as end of day.

[jquast]

I have reverted to the python-only version of iter_graphemes(), as I can't say for sure if or when standard python's unicodedata.iter_graphemes will be fixed for the InCB/Indic_Conjunct_Break rules changed in unicode 18, so this is expected to degrade performance for any grapheme-related functions.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.98%. Comparing base (1042e5f) to head (36f165b).

Additional details and impacted files

@@                 Coverage Diff                  @@
##           jq/libwcwidth-rc     #238      +/-   ##
====================================================
- Coverage            100.00%   98.98%   -1.02%     
====================================================
  Files                    27       27              
  Lines                  1999     1975      -24     
  Branches                467      460       -7     
====================================================
- Hits                   1999     1955      -44     
- Misses                    0       11      +11     
- Partials                  0        9       +9     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.