Every implementation in documents.py and rebuild.py takes the session it
writes to, so a set cannot reach one: the facade narrows once and passes the
database on, rather than checking and carrying a union. Tests calling an
implementation directly go through `writing()`.
haiku.rag.config exported two configuration instances: the lazy _config
behind get_config/set_config, and Config, loaded at import time. Nothing
linked them, and eleven signatures captured Config as a default argument,
so set_config could not reach the factories, the client, the store or the
MCP server. reranking/base.py went further and snapshotted the configured
reranker name into a class attribute at import.
Config is removed. Internal defaults are config: AppConfig | None = None,
resolved through get_config() per call. RerankerBase._model is None and
CohereReranker takes its model name as an argument, like every other
reranker.
The suite patched attributes on Config while production read the instance
get_config() returns, a different object, so those patches were no-ops
waiting to happen. They now go through get_config().
`resolve_refs`, `get_items_in_range`, `get_caption_picture_refs` and
`get_all_items_grouped` have no callers left: the grouped equivalents serve every
path that used them. `get_all_items_grouped` had none even before this branch.
Tests whose subject was a removed method go with it. Tests that only used one to
fetch a fixture now use the grouped call, so what they assert is unchanged.
Context expansion is now automatic and structure-aware. For structured
documents, expands within the section containing the match. For sections
that exceed the budget or are too small, expands item-by-item outward
skipping noise labels. Unstructured documents use budget-based outward
expansion. Results sorted by relevance score.
- Add doc_item_refs and headings assertions to test_expand_context_docling_merges_metadata
- Add page count and markdown content assertions to test_convert_pdf_with_ocr_engine
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Strengthen test_search_returns_search_result with chunk_id, document_id, and
non-empty labels assertions
- Remove test_chunks_include_document_title (subsumed by strengthened test) and
its cassette
- Rename test_chunks_include_document_info to
test_search_chunk_includes_document_provenance; add document_title is None
assertion for untitled documents
- Move test_search_result_get_primary_label from test_context_enhancement.py to
test_search.py; rename, remove unnecessary @pytest.mark.vcr() decorator
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
PictureItem.export_to_markdown() defaults to ImageRefMode.EMBEDDED,
which embeds base64 image data. When _extract_item_text() called this
method during context expansion, base64 data would leak into results.
Now explicitly handles PictureItem with ImageRefMode.PLACEHOLDER to
prevent base64 while still including VLM descriptions and captions.
