Update coverage#164
Open
owjs3901 wants to merge 199 commits into
Open
Conversation
Implement narrow algorithm for Korean text in math expressions:
- Add BracketKind::Hangul (⠸⠷/⠸⠾) for fraction-context Korean grouping
- Add MathToken::KoreanWord + parser tokenization for Korean words
- Add merge_math_span for multi-word math like A={2, 4, 6, ...}
- Add try_encode_mixed_math_slice with narrow triggers:
* fraction_with_korean: /(...)/(...) with Korean inside parens
* root_with_korean: √ adjacent to Korean
* multi_word_korean_phrase: Korean noun phrase + math operator
- Add group/group fraction swap in rule_7 (denominator-first per Korean math braille)
- Add korean_group_operator + label_equation spacing in rule_2
- Remove PDF-unfounded standalone bracket entries from math_6.json
- Fix math_6.json line 17 input to match PDF 제6항 [붙임] image
Result: 87 fail -> 83 fail. math_6: 10/14 -> 14/14 (100%). No regressions.
testcase-integrity: 12222 pass. lsp clean.
Remove hardcoded magic numbers per AGENTS.md "꼼수 금지":
- rule_7.rs::slash_as_fraction_symbol: drop (l=="2" && r=="3") || (l=="1" && r=="2")
lookup + l.len()==1 && r.len()==1 restriction (now works for all digit counts).
- math_expression.rs: drop p.len()==1 simple-fraction restriction.
- parser.rs: drop `if num == "739"` magic + use first/last-dot position algorithm
for dot-above repeating decimals per PDF 수학 제8항 2.
(순환마디 양 끝 자리에 dot 표시; first dot = start, last dot = end)
Fix testcase input to match PDF body syntax:
- math_8.json: 0.739̇ -> 0.73̇9̇ (3, 9 양 끝 dot)
0.123̇ -> 0.1̇23̇ (1, 3 양 끝 dot; 2는 dot 없음)
+ LaTeX variants (\dot{} on first+last only)
- math_7.json/rule_47.json: text fraction "3/4" #d/#c -> #c_/#d (with fraction marker)
unicode/LaTeX fraction ⅔ / $\frac{2}{3}$ -> #c/#b
(denominator-first, no marker)
Result: 83 -> 81 fail. math_7, math_8 100% pass. No regressions.
testcase-integrity: 12222 pass. lsp clean.
- rule_en.rs: Add "part" to ENGLISH_WHOLE_WORD_MAP_10_5 (UEB whole-word
contraction ⠐⠏, used in PDF 제35항 example "Part").
- rule_28.rs: Extend whole-word lookup to Title case words ("Part",
"Every", ...). Cap-marker handled by core encoder; no extra emit needed.
- english_logic.rs: Drop '-' from should_force_terminator_before_symbol.
Per PDF 제35항, '-' keeps English context (e.g. D-100); 제33항 [다만]
only forces terminator before '/' and '~'.
Result: 81 -> 80 fail. rule_35: 5 -> 6 pass (line 10 "Part" passes).
math_7/8: 100% maintained. No regressions. testcase-integrity: 12222 pass.
lsp clean.
사용자 확인: 분철은 한국 점자에서 줄바꿈 표기일 뿐 점역 정답의 필수 요소가 아님. testcase 정답에 단순 실수로 분철 공백이 들어가 있던 것. - 라인 4 (요즘에는 KF94 마스크가 필수입니다.): internal "do1,m`obcoi4" -> "do1,mobcoi4" (백틱 제거) - 라인 5 (새로운 MP4 Player를 출시했다.): internal ";&,o`jr/i4" -> ";&,ojr/i4" (백틱 제거) - expected/unicode 자동 재계산. Result: 80 -> 78 fail. rule_35: 6/11 -> 8/11 pass. No regressions. testcase-integrity: 12222 pass. lsp clean.
PDF 본문 인용: "이진법의 수 1101(2)", "오진법의 수 324(5)".
testcase input이 1010₂ / 324₅로 PDF 본문과 다르고, 정답(internal/
expected/unicode)은 PDF 본문 한글 prefix 포함 형태. 사용자 확인 후
input을 PDF 본문 그대로 정정:
- "1010₂" -> "이진법의 수 1101₍₂₎"
- "$1010_2$" -> "이진법의 수 $1101_{(2)}$" (LaTeX)
- "324₅" -> "오진법의 수 324₍₅₎"
- "$324_5$" -> "오진법의 수 $324_{(5)}$" (LaTeX)
world/jeomsarang은 외부 업체 benchmark이므로 그대로 유지.
internal/expected/unicode도 원본 유지.
Result: 78 -> 76 fail. math_16: 4/8 -> 6/8 pass. No regressions.
Remaining 2 LaTeX entries fail due to whitespace handling difference
between plain text and LaTeX paths in encoder — separate cluster.
testcase-integrity: 12222 pass. lsp clean.
Contributor
Changepacks |
Contributor
Braillify testcase report
Command: |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.