Add basic search to MediaWiki static site generator #4
Conversation
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
This commit implements a MediaWiki-style search feature for the static site generator: Backend (Rust): - Added SearchEntry struct to represent indexed pages - Implemented extract_search_data() to extract text content and headings from wikitext AST - Modified generate_wiki() to build and write search-index.json during site generation - Added serde dependency for JSON serialization Frontend (JavaScript): - Created WikiSearch class with intelligent ranking algorithm - Implemented prefix and substring matching for queries - Added result highlighting with context snippets - Lazy-loading of search index on first interaction - Debounced search input for performance UI: - Added search bar to navigation with dropdown results - Styled with Tailwind CSS for consistency - Shows highlighted matches in titles and content snippets - Displays relevant headings for each result Features: - Multi-word search support - Relevance scoring (title > headings > content) - Result snippets showing query context - Yellow highlighting of matching terms - Up to 20 results per query
- Remove redundant if/else with identical branches in extract_search_data - Apply rustfmt formatting throughout codebase
Performance improvements:
- Changed from linear full-text search to inverted index lookup
- Search index now maps words to page indices instead of storing full content
- Title words weighted 3x higher for better ranking
- Multi-word queries require all words to be present
Benefits:
- O(k) lookup time where k = number of query words (vs O(n) scanning all pages)
- Smaller memory footprint during search
- Faster query responses
- Better relevance ranking
Index structure:
- pages: array of {title, url, headings}
- words: map of word -> [page indices with occurrence counts]
Stats:
- 1,208 pages indexed
- 5,585 unique words
- 442KB index file
Performance optimization: - Generate individual .txt files for each page (alongside .html) - Text files contain extracted searchable content (~376KB total) - JavaScript loads text on-demand for top 5 results only - Snippets show highlighted context around query matches - Text files cached in-memory after first load Benefits: - Initial search remains fast (426KB index) - Average text file only 0.3KB (~300 bytes) - Only ~1.5KB additional data loaded per search (5 × 0.3KB) - Better UX with context snippets for top results - Headings still shown for results 6-20 File structure: /wiki/Lua/Client/Window.html (36KB HTML) /wiki/Lua/Client/Window.txt (465B text content) /wiki/Lua/Client/Window.json (debug AST) Stats: - 1,208 .txt files generated - Search index: 426KB (unchanged from pure metadata) - Text files: 376KB total (loaded on-demand)
Index structure optimizations:
- Pages now stored as simple title array (not objects)
- URLs derived client-side from titles (title.replace(' ', '_'))
- Removed headings from index (loaded from .txt if needed)
- Use (page_id, weight) tuples instead of duplicate entries
Weighting system:
- Title words: weight 5 (highest priority)
- Heading words: weight 3 (medium priority)
- Content words: weight 1 (base priority)
- Uses max() when word appears in multiple contexts
Results:
- Index size: 426KB → 298KB (30% reduction, 128KB saved)
- 1,208 pages indexed
- 5,585 unique words
- Smarter relevance ranking with explicit weights
Example index format:
{
"pages": ["Lua/Client/Window", "Feature Overview", ...],
"words": {
"window": [[173, 5], [174, 5], [9, 1], ...],
...
}
}
Where [173, 5] means page 173 with weight 5 (title match)
With average text files of ~300 bytes, loading all 20 results only costs ~6KB total. This provides better UX by showing context snippets for all results, not just the top 5. Benefits: - All results show highlighted context snippets - Still efficient (~6KB for 20 results) - Text files cached after first load - Parallel loading keeps it fast
philpax
deleted the
claude/add-search-feature-014pEUUDjASbdvUPmQVpjrTS
branch
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
This commit implements a MediaWiki-style search feature for the static site generator.
Features: