mirror of
https://github.com/deepset-ai/haystack.git
synced 2025-06-26 22:00:13 +00:00
f025501792
* test(extractors): Add unit test for LLMMetadataExtractor with no content
Adds a new unit test `test_run_with_document_content_none` to `TestLLMMetadataExtractor`.
This test verifies that `LLMMetadataExtractor` correctly handles documents where `document.content` is None or an empty string.
It ensures that:
- Such documents are added to the `failed_documents` list.
- The correct error message ("Document has no content, skipping LLM call.") is present in their metadata.
- No actual LLM call is attempted for these documents.
This test provides coverage for the fix that prevents an AttributeError when processing documents with no content.
* chore: update comment to reflect new behavior in _run_on_thread method
* docs: Add release note for LLMMetadataExtractor no content fix
* Update releasenotes/notes/fix-llm-metadata-extractor-no-content-910067ea72094f18.yaml
* Update fix-llm-metadata-extractor-no-content-910067ea72094f18.yaml
---------
Co-authored-by: David S. Batista <dsbatista@gmail.com>
9 lines
344 B
YAML
9 lines
344 B
YAML
---
|
|
fixes:
|
|
- |
|
|
Fixed a bug in the `LLMMetadataExtractor` that occurred when
|
|
processing `Document` objects with `None` or empty string content. The
|
|
component now gracefully handles these cases by marking such documents as
|
|
failed and providing an appropriate error message in their metadata, without
|
|
attempting an LLM call.
|