Inline Char16Trie traversal methods

[Xuanwo]

This PR isolates the Char16Trie traversal inline annotations from #8113.

The change is intentionally small: it only adds #[inline] to the hot Char16TrieIterator traversal entry points and helper paths in components/collections/src/char16trie/trie.rs. It does not change public APIs and does not include the ResultCache, zero-copy complex segmentation, or benchmark implementation changes from the larger performance branch.

Why separate:

• Char16TrieIterator is shared by published segmenter code and the in-progress neo segmenter.
• The inline behavior is an independent optimization from the segmenter cache and slicing changes.
• Keeping this PR scoped makes the performance impact easier to review and attribute.

Benchmark setup:

• Baseline: latest origin/main at 4f4c9f1fa068b6e4e0331b385104a7b2872489c5.
• This branch: 665a7db7fa Inline Char16Trie traversal methods.
• Benchmark code: PR Add segmenter comparison benchmarks #8141, Add segmenter comparison benchmarks, head 044cc38edec9ae46d03db66e9f5016bcd85e5835.
• Method: the Add segmenter comparison benchmarks #8141 benchmark patch was temporarily applied to both the baseline worktree and this branch's worktree. The benchmark file is intentionally not included in this PR diff.
• Reproduction outline: apply Add segmenter comparison benchmarks #8141's benchmark patch to origin/main, run the command below; then apply the same patch to this branch and rerun the same command.
• Command: cargo bench -p icu_segmenter --bench bench --features unstable -- dictionary --sample-size 10 --measurement-time 1 --warm-up-time 0.5.
• Values below are criterion median point estimates for the dictionary filter. This was a short local run, so small single-digit changes and short English/Latin1 paths should be treated as noisy; the table is intended to show the current published and neo dictionary/Char16Trie-path direction under the latest benchmark code.

Representative dictionary / Char16Trie-related results:

Case	Before	After	Delta
Segmenter_word_segment_str/published_dictionary_thai	2.355 us	2.347 us	-0.3%
Segmenter_word_segment_str/published_dictionary_japanese	1.661 us	1.663 us	+0.1%
Segmenter_word_segment_str/published_dictionary_han	1.530 us	1.530 us	-0.0%
Segmenter_word_segment_str/neo_dictionary_thai	1.452 us	1.373 us	-5.5%
Segmenter_word_segment_utf8/neo_dictionary_thai	1.720 us	1.727 us	+0.5%
Segmenter_word_segment_utf16/neo_dictionary_thai	1.271 us	1.202 us	-5.4%
Segmenter_line_segment_str/published_dictionary_thai	2.461 us	2.427 us	-1.4%
Segmenter_line_segment_str/neo_dictionary_thai	1.768 us	1.681 us	-5.0%
Segmenter_line_segment_utf8/published_dictionary_thai	2.545 us	2.684 us	+5.5%
Segmenter_line_segment_utf8/neo_dictionary_thai	2.391 us	2.130 us	-10.9%

Dictionary benchmark delta:

Case	Before	After	Delta
Segmenter_line_segment_latin1/neo_dictionary_css_english	3.010 us	4.279 us	+42.1%
Segmenter_line_segment_latin1/neo_dictionary_english	2.105 us	4.481 us	+112.9%
Segmenter_line_segment_latin1/published_dictionary_css_english	2.730 us	3.825 us	+40.1%
Segmenter_line_segment_latin1/published_dictionary_english	4.048 us	4.101 us	+1.3%
Segmenter_line_segment_str/neo_dictionary_css_english	6.446 us	5.552 us	-13.9%
Segmenter_line_segment_str/neo_dictionary_english	3.902 us	3.763 us	-3.6%
Segmenter_line_segment_str/neo_dictionary_han	309.7 ns	251.5 ns	-18.8%
Segmenter_line_segment_str/neo_dictionary_japanese	400.0 ns	344.5 ns	-13.9%
Segmenter_line_segment_str/neo_dictionary_thai	1.768 us	1.681 us	-5.0%
Segmenter_line_segment_str/neo_dictionary_thai_han	675.3 ns	712.1 ns	+5.5%
Segmenter_line_segment_str/neo_dictionary_thai_japanese	793.6 ns	748.7 ns	-5.7%
Segmenter_line_segment_str/published_dictionary_css_english	7.274 us	7.726 us	+6.2%
Segmenter_line_segment_str/published_dictionary_english	6.482 us	6.230 us	-3.9%
Segmenter_line_segment_str/published_dictionary_han	285.3 ns	288.9 ns	+1.3%
Segmenter_line_segment_str/published_dictionary_japanese	381.9 ns	375.9 ns	-1.6%
Segmenter_line_segment_str/published_dictionary_thai	2.461 us	2.427 us	-1.4%
Segmenter_line_segment_str/published_dictionary_thai_han	838.3 ns	821.0 ns	-2.1%
Segmenter_line_segment_str/published_dictionary_thai_japanese	961.9 ns	954.0 ns	-0.8%
Segmenter_line_segment_utf16/neo_dictionary_css_english	7.386 us	5.719 us	-22.6%
Segmenter_line_segment_utf16/neo_dictionary_english	4.931 us	4.774 us	-3.2%
Segmenter_line_segment_utf16/neo_dictionary_han	346.1 ns	221.8 ns	-35.9%
Segmenter_line_segment_utf16/neo_dictionary_japanese	378.4 ns	286.6 ns	-24.3%
Segmenter_line_segment_utf16/neo_dictionary_thai	1.528 us	1.518 us	-0.7%
Segmenter_line_segment_utf16/neo_dictionary_thai_han	621.8 ns	517.1 ns	-16.8%
Segmenter_line_segment_utf16/neo_dictionary_thai_japanese	740.8 ns	603.5 ns	-18.5%
Segmenter_line_segment_utf16/published_dictionary_css_english	6.348 us	6.004 us	-5.4%
Segmenter_line_segment_utf16/published_dictionary_english	6.063 us	5.969 us	-1.5%
Segmenter_line_segment_utf16/published_dictionary_han	293.0 ns	285.0 ns	-2.7%
Segmenter_line_segment_utf16/published_dictionary_japanese	384.7 ns	387.6 ns	+0.8%
Segmenter_line_segment_utf16/published_dictionary_thai	2.154 us	2.420 us	+12.3%
Segmenter_line_segment_utf16/published_dictionary_thai_han	691.1 ns	691.6 ns	+0.1%
Segmenter_line_segment_utf16/published_dictionary_thai_japanese	831.6 ns	851.2 ns	+2.4%
Segmenter_line_segment_utf8/neo_dictionary_css_english	11.801 us	7.075 us	-40.1%
Segmenter_line_segment_utf8/neo_dictionary_english	7.939 us	5.882 us	-25.9%
Segmenter_line_segment_utf8/neo_dictionary_han	458.5 ns	345.9 ns	-24.6%
Segmenter_line_segment_utf8/neo_dictionary_japanese	595.4 ns	446.6 ns	-25.0%
Segmenter_line_segment_utf8/neo_dictionary_thai	2.391 us	2.130 us	-10.9%
Segmenter_line_segment_utf8/neo_dictionary_thai_han	924.4 ns	842.1 ns	-8.9%
Segmenter_line_segment_utf8/neo_dictionary_thai_japanese	1.369 us	1.087 us	-20.6%
Segmenter_line_segment_utf8/published_dictionary_css_english	8.242 us	7.735 us	-6.2%
Segmenter_line_segment_utf8/published_dictionary_english	6.856 us	8.144 us	+18.8%
Segmenter_line_segment_utf8/published_dictionary_han	314.3 ns	339.4 ns	+8.0%
Segmenter_line_segment_utf8/published_dictionary_japanese	414.5 ns	435.9 ns	+5.2%
Segmenter_line_segment_utf8/published_dictionary_thai	2.545 us	2.684 us	+5.5%
Segmenter_line_segment_utf8/published_dictionary_thai_han	899.9 ns	895.1 ns	-0.5%
Segmenter_line_segment_utf8/published_dictionary_thai_japanese	1.007 us	1.037 us	+3.0%
Segmenter_word_segment_str/neo_dictionary_english	4.043 us	4.010 us	-0.8%
Segmenter_word_segment_str/neo_dictionary_han	980.5 ns	770.5 ns	-21.4%
Segmenter_word_segment_str/neo_dictionary_japanese	1.863 us	906.4 ns	-51.3%
Segmenter_word_segment_str/neo_dictionary_thai	1.452 us	1.373 us	-5.5%
Segmenter_word_segment_str/neo_dictionary_thai_han	623.2 ns	579.5 ns	-7.0%
Segmenter_word_segment_str/neo_dictionary_thai_japanese	1.259 us	864.7 ns	-31.3%
Segmenter_word_segment_str/published_dictionary_english	4.412 us	4.392 us	-0.5%
Segmenter_word_segment_str/published_dictionary_han	1.530 us	1.530 us	-0.0%
Segmenter_word_segment_str/published_dictionary_japanese	1.661 us	1.663 us	+0.1%
Segmenter_word_segment_str/published_dictionary_thai	2.355 us	2.347 us	-0.3%
Segmenter_word_segment_str/published_dictionary_thai_han	1.042 us	1.020 us	-2.1%
Segmenter_word_segment_str/published_dictionary_thai_japanese	1.518 us	1.483 us	-2.3%
Segmenter_word_segment_utf16/neo_dictionary_english	4.930 us	5.135 us	+4.1%
Segmenter_word_segment_utf16/neo_dictionary_han	776.6 ns	699.4 ns	-9.9%
Segmenter_word_segment_utf16/neo_dictionary_japanese	878.8 ns	818.8 ns	-6.8%
Segmenter_word_segment_utf16/neo_dictionary_thai	1.271 us	1.202 us	-5.4%
Segmenter_word_segment_utf16/neo_dictionary_thai_han	609.4 ns	510.8 ns	-16.2%
Segmenter_word_segment_utf16/neo_dictionary_thai_japanese	1.113 us	767.8 ns	-31.0%
Segmenter_word_segment_utf16/published_dictionary_english	4.344 us	3.782 us	-12.9%
Segmenter_word_segment_utf16/published_dictionary_han	2.601 us	1.410 us	-45.8%
Segmenter_word_segment_utf16/published_dictionary_japanese	2.151 us	1.554 us	-27.8%
Segmenter_word_segment_utf16/published_dictionary_thai	2.159 us	2.147 us	-0.6%
Segmenter_word_segment_utf16/published_dictionary_thai_han	993.2 ns	946.1 ns	-4.7%
Segmenter_word_segment_utf16/published_dictionary_thai_japanese	1.702 us	1.376 us	-19.1%
Segmenter_word_segment_utf8/neo_dictionary_english	6.351 us	6.594 us	+3.8%
Segmenter_word_segment_utf8/neo_dictionary_han	933.5 ns	909.3 ns	-2.6%
Segmenter_word_segment_utf8/neo_dictionary_japanese	1.091 us	1.075 us	-1.4%
Segmenter_word_segment_utf8/neo_dictionary_thai	1.720 us	1.727 us	+0.5%
Segmenter_word_segment_utf8/neo_dictionary_thai_han	795.5 ns	748.7 ns	-5.9%
Segmenter_word_segment_utf8/neo_dictionary_thai_japanese	1.177 us	1.130 us	-4.0%
Segmenter_word_segment_utf8/published_dictionary_english	5.042 us	4.975 us	-1.3%
Segmenter_word_segment_utf8/published_dictionary_han	1.665 us	1.624 us	-2.5%
Segmenter_word_segment_utf8/published_dictionary_japanese	1.759 us	1.756 us	-0.1%
Segmenter_word_segment_utf8/published_dictionary_thai	2.594 us	2.556 us	-1.5%
Segmenter_word_segment_utf8/published_dictionary_thai_han	1.120 us	1.155 us	+3.2%
Segmenter_word_segment_utf8/published_dictionary_thai_japanese	1.616 us	1.585 us	-1.9%

Validation:

• cargo fmt -p icu_collections -p icu_segmenter
• git diff --check
• cargo test -p icu_collections --features serde,databake
• cargo test -p icu_segmenter --features unstable

Changelog

icu_collections: Improve Char16Trie traversal performance with inline annotations.