Import Vocabulary

From the extracted content, identify vocabulary entries and create a markdown file in the same directory as the source document.

Filename pattern: {original_name}.vocab.md

Example: صفات.docx → صفات.vocab.md

Markdown format:

# Vocabulary: صفات (Adjectives)

Source: صفات.docx
Extracted: 2024-01-15
Category: Adjectives

## Words

| # | Arabic | English | Notes | Import |
|---|--------|---------|-------|--------|
| 1 | جديد / جديدة | new | masc/fem | ✓ |
| 2 | قديم / قديمة | old | masc/fem | ✓ |
| 3 | ثقيل / ثقيلة | heavy | masc/fem | ✓ |
| 4 | خفيف / خفيفة | light | masc/fem | ✓ |

## Summary

- Total words found: 12
- Ready to import: 12
- Skipped: 0

Parsing rules:

• Strip {dir="rtl"} markers from Arabic text
• Ignore image references like ![...](media/...)
• Skip exercise blanks (____), instructions, non-vocabulary content
• Handle masculine/feminine pairs: "جديد- جديدة" → "جديد / جديدة"
• Preserve Arabic diacritics (tashkeel)

Step 3: Ask for Confirmation

Present the generated markdown to the user and ask:

I've extracted 12 vocabulary items from "صفات.docx"

Category detected: Adjectives (صفات)
Target deck: [New deck "Adjectives (صفات)" / Existing deck "..."]

Preview:
1. جديد / جديدة - new
2. قديم / قديمة - old
3. ثقيل / ثقيلة - heavy
...

Would you like to:
- Import all words
- Review the full list in صفات.vocab.md first
- Skip certain words (specify which)

Step 4: Check/Create Deck

Check existing decks:

curl -s http://localhost:3001/api/decks

Determine deck:

1. Check filename for category hints:
   • "صفات" = Adjectives
   • "أسماء" = Nouns
   • "أفعال" = Verbs
   • "definite article" = Grammar
2. Match against existing deck names
3. If no match: create new deck or use "Import YYYY-MM-DD"

Create new deck (if needed):

curl -X POST "http://localhost:3000/api/decks" \
  -H "Content-Type: application/json" \
  -d '{"name": "Adjectives (صفات)", "description": "Imported from صفات.docx"}'

Step 5: Import Cards via API

curl -X POST "http://localhost:3000/api/decks/{deck_id}/cards" \
  -H "Content-Type: application/json" \
  -d '[
    {"front": "جديد / جديدة", "back": "new", "notes": "masc/fem"},
    {"front": "قديم / قديمة", "back": "old", "notes": "masc/fem"}
  ]'

Step 6: Archive to Processed Folder

After successful import, move both files to a processed/ subfolder:

# Create processed folder if it doesn't exist
mkdir -p "path/to/processed"

# Move the original docx
mv "path/to/document.docx" "path/to/processed/"

# Move the generated vocab markdown
mv "path/to/document.vocab.md" "path/to/processed/"

Folder structure after processing:

example-arabic-docs/
├── processed/
│   ├── صفات.docx
│   ├── صفات.vocab.md
│   ├── الأسماء.docx
│   └── الأسماء.vocab.md
└── new-document.docx  (not yet processed)

Step 7: Report Results

✓ Imported 12 cards to deck "Adjectives (صفات)" (ID: 5)

Files archived:
  → processed/صفات.docx
  → processed/صفات.vocab.md

View deck: http://localhost:3000/deck/5

Error Handling

• If dev server isn't running: prompt user to run npm run dev
• If API call fails: keep files in original location, report error
• If no vocabulary found: create empty .vocab.md noting "No vocabulary extracted"

Notes

• The .vocab.md files serve as a permanent record of what was extracted
• User can edit .vocab.md before confirming import if needed
• The processed/ folder makes it easy to see what's been imported
• WebSearch can be used to look up unclear translations