sudacode/SubMiner
1
0
Fork 0
mirror of https://github.com/ksyasuda/SubMiner.git synced 2026-02-27 18:22:41 -08:00
Code Issues Packages Projects Releases Wiki Activity

chore(task-85): update launcher workflow and backlog tracking

Browse Source
This commit is contained in:
Kyle Yasuda
2026-02-20 03:51:38 -08:00
parent 06892b4838
commit 28d2da1e64
15 changed files with 373 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
backlog/tasks/task-85 - Refactor-large-files-for-maintainability-and-readability.md
View File

@@ -4,7 +4,7 @@ title: Refactor large files for maintainability and readability
status: In Progress
assignee: []
created_date: '2026-02-19 09:46'
updated_date: '2026-02-19 10:01'
updated_date: '2026-02-20 11:42'
labels:
- architecture
- refactor
@@ -31,14 +31,24 @@ Several core files are oversized and high-coupling (`src/main.ts`, `src/anki-int
<!-- SECTION:PLAN:BEGIN -->
1. Add file-budget guardrails and baseline report.
2. Split `src/main.ts` into runtime domain modules.
3. Split `src/anki-integration.ts` into focused collaborators.
4. Split `src/config/service.ts` by load/migrate/validate/warn phases.
2. Split `src/main.ts` into runtime domain modules (ongoing; major wiring extracted).
3. Split `src/anki-integration.ts` into focused collaborators (started; constructor composition decomposed).
4. Split `src/config/service.ts` by load/migrate/validate/warn phases (started; load/apply flow decomposition in place).
5. Split immersion tracker service by state, persistence, sync responsibilities.
6. Clarify generated launcher artifact workflow and docs.
7. Run full build/test gate and publish maintainability report.
<!-- SECTION:PLAN:END -->
## Progress Notes
<!-- SECTION:PROGRESS:BEGIN -->
- 2026-02-20: Large `src/main.ts` composition slices extracted into runtime handler modules (startup, CLI, tray/window/bootstrap, shortcuts, OSD/secondary-sub, numeric/overlay shortcut lifecycle, and related seams) with focused parity tests.
- 2026-02-20: `src/anki-integration.ts` constructor decomposed into targeted private factory/config methods (`normalizeConfig`, `createKnownWordCache`, `createPollingRunner`, `createCardCreationService`, `createFieldGroupingService`) to reduce orchestration complexity.
- 2026-02-20: `src/config/service.ts` load/apply duplication reduced by introducing `applyResolvedConfig`, `resolveExistingConfigPath`, and `parseConfigContent`.
- Validation checkpoint: `bun run build` and focused suites (`dist/config/config.test.js`, `dist/anki-integration.test.js`) passing.
- Current strategy: prioritize high-ROI extractions (behavioral seams and churn-heavy hotspots) over further low-impact micro-shuffles.
<!-- SECTION:PROGRESS:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 `src/main.ts` reduced to orchestration-focused module with extracted runtime domains

46
backlog/tasks/task-88 - Remove-MeCab-fallback-tokenizer-and-simplify-Yomitan-token-flow.md Normal file
View File

@@ -0,0 +1,46 @@
---
id: TASK-88
title: Remove MeCab fallback tokenizer and simplify Yomitan token flow
status: To Do
assignee: []
created_date: '2026-02-20 00:00'
labels:
- tokenizer
- refactor
dependencies: []
priority: medium
---
## Description
<!-- SECTION:DESCRIPTION:BEGIN -->
Remove the MeCab fallback tokenization path and associated merge-selection complexity in subtitle tokenization. Treat Yomitan parser output as the single source of token boundaries/grouping, and keep only minimal normalization needed for downstream known-word, JLPT, and frequency annotation.
<!-- SECTION:DESCRIPTION:END -->
## Action Steps
<!-- SECTION:PLAN:BEGIN -->
1. Remove MeCab fallback execution from `tokenizeSubtitle` and delete dead fallback-specific branches.
2. Remove merge/candidate-selection code that is only needed to reconcile MeCab-vs-Yomitan tokenization strategies.
3. Keep Yomitan parsing pipeline with minimal structural token normalization only.
4. Update MeCab usage so it is no longer required for tokenization fallback (retain only explicitly needed behavior, if any).
5. Update docs/config notes to reflect Yomitan-only tokenization flow.
6. Add regression tests for Yomitan-only success/failure paths and token annotation continuity.
<!-- SECTION:PLAN:END -->
## Acceptance Criteria
<!-- AC:BEGIN -->
- [ ] #1 Subtitle tokenization no longer falls back to MeCab when Yomitan parsing fails.
- [ ] #2 Token grouping logic is simplified to rely on Yomitan structure; redundant custom merge-selection logic removed.
- [ ] #3 Known-word, JLPT, frequency, and N+1 annotations still work on Yomitan-derived tokens.
- [ ] #4 If Yomitan parsing fails, behavior is explicit and tested (for example `tokens: null` without MeCab fallback path).
- [ ] #5 Documentation reflects that tokenization flow is Yomitan-first and Yomitan-only.
<!-- AC:END -->
## Definition of Done
<!-- DOD:BEGIN -->
- [ ] #1 `src/core/services/tokenizer.ts` no longer contains MeCab fallback tokenization branch.
- [ ] #2 Tests cover Yomitan-only pipeline and failure behavior regressions.
- [ ] #3 Any removed MeCab-only merge helpers are deleted with no unused exports/imports.
- [ ] #4 Build and relevant tokenizer/subtitle tests pass.
<!-- DOD:END -->