Merged
Conversation
…gChunker, RecursiveCharacterChunker
- Add HybridRetriever: BM25 + vector search merged via Reciprocal Rank
Fusion (k=60), over-fetching top_n*2 from each retriever
- Add HybridNL2SQL flow: HybridRetriever → SQLGenerator → SQLExecutor
- Clean up BaselineNL2SQL: remove retriever= injection and isinstance branch
- Export HybridRetriever, HybridNL2SQL from public API
- Add 8 unit tests (99 total)
- Add DocumentLoaderPort protocol to core/ports.py - Add MarkdownLoader: .md file/directory → list[TextDocument], stdlib only - Add PlainTextLoader: .txt file/directory → list[TextDocument], stdlib only - Add DirectoryLoader: extension-based dispatch across a directory tree - Add PDFLoader (integrations): pymupdf optional, one doc per page - Export new loaders from public API - Add 8 unit tests (107 total)
…iveCharacterChunker LangChain-style convenience: split(list) → list[IndexedChunk] wraps the existing chunk() per-item method.
Consistent with CatalogChunker and RecursiveCharacterChunker so all chunkers share the same split(docs) interface.
…) internals
- from_chunks(): LangChain-style factory that takes pre-split IndexedChunks
- from_sources(): now delegates to from_chunks() internally; renames
document_chunker → splitter parameter
- add(): signature changed from list[TextDocument] to list[IndexedChunk];
removes RuntimeError guard and _index_builder reference
Aligns HybridRetriever with the updated VectorRetriever.from_sources() API.
IndexBuilder is superseded by the from_chunks() explicit pipeline. Delete index_builder.py and remove all imports/exports from retrieval/__init__.py and the top-level __init__.py.
- Replace IndexBuilder-based tests (8, 9, 10, 14) with from_chunks() equivalents - Update test 15: remove _index_builder assertion - Update test 16: add() now requires pre-split IndexedChunks - Add 3 new tests: from_chunks_empty, from_chunks_mixed, from_chunks_add_incremental - Add test for CatalogChunker.split() batch method
- Section 0/1: remove IndexBuilder mention, add .split() description and PDFLoader - Section 2: update 'advanced control' to reference from_chunks() pattern - Section 3C: rewrite explicit pipeline example using split() + from_chunks() - Section 3D (new): DirectoryLoader → HybridNL2SQL direct pattern - Section 3E (new): PDFLoader usage with DirectoryLoader - Section 4: add constraint note for add() requiring pre-split IndexedChunks
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.
#️⃣ Issue Number
📝 요약(Summary)
💬 To Reviewers (선택)
PR Checklist
reference) How to Code Review