PDF Translation Workflow

Phase 1: Extract PDF

cd scripts
uv run python extract_pdf.py ../data/pdfs/<filename>.pdf

Outputs (in data/markdown/):

Phase 2: Configure Chapters

1. •Generate template: uv run python split_chapters.py --init
2. •Edit chapters.json:

{
  "source": "data/markdown/<name>_pages.md",
  "output_dir": "docs/src/content/docs",
  "clean_patterns": ["\\(Order #\\d+\\)"],
  "chapters": {
    "section-slug": {
      "title": "章節標題",
      "order": 1,
      "files": {
        "filename": {
          "title": "頁面標題",
          "description": "SEO 描述",
          "pages": [1, 10],
          "order": 0
        }
      }
    }
  }
}

1. •Split: uv run python split_chapters.py

Phase 3: Translation

For each markdown file in docs/src/content/docs/:

1. •Identify terms - Extract English game terms
2. •Check glossary - Lookup in glossary.json
3. •Translate - Apply consistent terminology
4. •Preserve structure - Keep frontmatter, headings, lists intact

Translation Rules

• •Maintain original meaning, avoid over-localization
• •Game mechanics terms: use established translations
• •Proper nouns: keep original or use accepted translations
• •Format: preserve markdown syntax exactly

Phase 4: Quality Check

Troubleshooting