Feature implementation from commits 2f2cfa7..ad5b512

[yashuatla]

This PR contains changes from a range of commits from the original repository.

Commit Range: 2f2cfa7..ad5b512
Files Changed: 344 (145 programming files)
Programming Ratio: 42.2%

Commits included:

• remove unused columns and rename document_attribute_columns (remove unused columns and rename document_attribute_columns  microsoft/graphrag#1672)
• Fix recursive report generation (Fix recursive report generation microsoft/graphrag#1669)
• Add generate_text_embeddings to FGR (Add generate_text_embeddings to FGR microsoft/graphrag#1667)
• Require explicit azure auth settings when using AOI. (Require explicit azure auth settings when using AOI. microsoft/graphrag#1665)
• NLP graph extraction (NLP graph extraction microsoft/graphrag#1652)
• Add vector store id reference to embeddings config. (Add vector store id reference to embeddings config. microsoft/graphrag#1662)
• Multi-index querying for API layer (Multi-index querying for API layer microsoft/graphrag#1644)
• Update auto_prompt_tuning.md (Update auto_prompt_tuning.md microsoft/graphrag#1659)
• Add smoke tests for drift (Add smoke tests for drift microsoft/graphrag#1658)
• Refactor config (Refactor config microsoft/graphrag#1593)
  ... and 5 more commits

[codeowlai]

🐛 Correctness Issue

Removed critical validation check.

The validation preventing simultaneous resume and update operations was removed, which could lead to data corruption or unexpected behavior.

Current Code (Diff):

-    is_update_run = bool(config.update_index_output)
+    is_update_run = bool(config.update_index_output)
+
+    if is_resume_run and is_update_run:
+        msg = "Cannot resume and update a run at the same time."
+        raise ValueError(msg)

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	is_update_run = bool(config.update_index_output)
	is_update_run = bool(config.update_index_output)
	if is_resume_run and is_update_run:
	msg = "Cannot resume and update a run at the same time."
	raise ValueError(msg)

[codeowlai]

🐛 Correctness Issue

Misplaced else clause nullifies vector_store_args.

The else clause after the for loop will always execute, setting vector_store_args to None and causing the application to crash when get_embedding_store tries to use it

Current Code (Diff):

- |4 spaces|else:
- |8 spaces|vector_store_args = None

🔄 Dependencies Affected

graphrag/api/query.py

Function: get_embedding_store

Issue: Function expects config_args to be a dictionary, but will receive None

Suggestion: Remove the else clause that sets vector_store_args to None

---

[codeowlai]

🐛 Correctness Issue

Type mismatch in match-case statement.

The match statement is using SearchType.LOCAL but the method parameter is of type SearchMethod, which would cause a runtime error when executed.

Current Code (Diff):

-         case SearchType.LOCAL:
+         case SearchMethod.LOCAL:

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	case SearchMethod.LOCAL:
	case SearchMethod.LOCAL:

[codeowlai]

🐛 Correctness Issue

Breaking API Change: Removed Embedding Filtering.

The code removes the ability to skip specific embeddings when using TextEmbeddingTarget.all, which will break existing code that relies on this filtering behavior.

Current Code (Diff):

- return all_embeddings
+ return all_embeddings.difference(settings.embeddings.skip)

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	return all_embeddings
	return all_embeddings.difference(settings.embeddings.skip)

[codeowlai]

🐛 Correctness Issue

Undefined variable 'all_embeddings' will cause NameError.

The function references 'all_embeddings' which is not defined or imported, causing a runtime exception when the function is called with validate=True.

Current Code (Diff):

    if validate and embedding_name not in all_embeddings:

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	if validate and embedding_name not in all_embeddings:
	if validate and embedding_name not in all_embeddings:

🔄 Dependencies Affected

graphrag/index/operations/embed_text/embed_text.py

Function: _get_collection_name

Issue: Will raise NameError when calling create_collection_name() due to undefined all_embeddings variable

Suggestion: Ensure all_embeddings is imported in the file where create_collection_name is defined

---

graphrag/utils/api.py

Function: get_embedding_store

Issue: Will raise NameError when calling create_collection_name() due to undefined all_embeddings variable

Suggestion: Ensure all_embeddings is imported in the file where create_collection_name is defined

---

[codeowlai]

🐛 Correctness Issue

Breaking API Change: Class Rename.

Renaming StorageType to OutputType will break all existing code that imports or uses the StorageType class, causing compilation failures and runtime exceptions.

Proposed Code:

class OutputType(str, Enum):
    """The output type for the pipeline."""

[codeowlai]

🐛 Correctness Issue

Breaking API Change: Default Config Behavior Removed.

The function now raises FileNotFoundError when no config file is found instead of creating a default configuration, which will break existing code that relies on the old behavior.

Current Code (Diff):

-    if not config_path:
-        msg = f"Config file not found in root directory: {root_dir}"
-        raise FileNotFoundError(msg)
+    if not config_path:
+        # For backward compatibility, create default config instead of raising error
+        return create_graphrag_config(root_dir=str(root_dir))

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	if not config_path:
	if not config_path:
	# For backward compatibility, create default config instead of raising error
	return create_graphrag_config(root_dir=str(root_dir))

[codeowlai]

🐛 Correctness Issue

Breaking Change: Import Path Modification.

Moving ChunkStrategyType to a different module breaks imports in dependent files, causing runtime exceptions.

🔄 Dependencies Affected

graphrag/index/operations/chunk_text/chunk_text.py

Function: import statements

Issue: Import statement will fail because ChunkStrategyType is no longer defined in chunking_config.py

Suggestion: Update import statement to import ChunkStrategyType from graphrag.config.enums

Current Code (Diff):

- from graphrag.config.models.chunking_config import ChunkingConfig, ChunkStrategyType
+ from graphrag.config.models.chunking_config import ChunkingConfig
+ from graphrag.config.enums import ChunkStrategyType

---

[codeowlai]

🐛 Correctness Issue

Breaking API change in method signature.

The resolved_strategy method signature has changed from accepting an encoding_model string to requiring a LanguageModelConfig object, which will break existing callers.

Current Code (Diff):

- |4 spaces|     def resolved_strategy(
- |8 spaces|         self, root_dir: str, model_config: LanguageModelConfig
- |4 spaces|     ) -> dict:
+ |4 spaces|     def resolved_strategy(
+ |8 spaces|         self, root_dir: str, model_config: LanguageModelConfig = None, encoding_model: str | None = None
+ |4 spaces|     ) -> dict:

Proposed Code:

    def resolved_strategy(
        self, root_dir: str, model_config: LanguageModelConfig = None, encoding_model: str | None = None
    ) -> dict:

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	def resolved_strategy(
	self, root_dir: str, model_config: LanguageModelConfig
	) -> dict:
	def resolved_strategy(
	self, root_dir: str, model_config: LanguageModelConfig = None, encoding_model: str | None = None
	) -> dict:

🔄 Dependencies Affected

Unknown

Function: Any caller of ClaimExtractionConfig.resolved_strategy

Issue: Existing callers pass an encoding_model string parameter which is no longer accepted

Suggestion: Update callers to create and pass a LanguageModelConfig object instead of an encoding_model string

---

[codeowlai]

🐛 Correctness Issue

Breaking API Change: New Required Parameter.

The resolved_strategy method signature has been changed to require a new parameter which will break all existing callers of this method.

Proposed Code:

    def resolved_strategy(
        self, root_dir: str, model_config: LanguageModelConfig
    ) -> dict:

[codeowlai]

🐛 Correctness Issue

Method reference used without calling it.

The strip method is being referenced but not called, causing the empty string check to always fail which will lead to runtime errors with empty URIs.

Current Code (Diff):

-                 if not store.db_uri or store.db_uri.strip == "":
+                 if not store.db_uri or store.db_uri.strip() == "":

📝 Committable suggestion

‼️ IMPORTANT
Trust, but verify! 🕵️ Please review this suggestion with the care of a code archaeologist - check that it perfectly replaces the highlighted code, preserves all lines, maintains proper indentation, and won't break anything in production. Your future self will thank you! 🚀

Suggested change

	if not store.db_uri or store.db_uri.strip == "":
	if not store.db_uri or store.db_uri.strip() == "":