|
|
@@ -308,7 +308,7 @@ class UnstructuredIngestionProvider(IngestionProvider):
|
|
|
# TODO - Remove code duplication between Unstructured & R2R
|
|
|
logger.info(f"Parser overrides: {parser_overrides}")
|
|
|
logger.info(f"R2R fallback parsers is: {document.document_type.value in parser_overrides or document.document_type.value in self.EXTRA_PARSERS.keys()}")
|
|
|
- if document.document_type.value in parser_overrides or document.document_type.value in self.EXTRA_PARSERS.keys():
|
|
|
+ if document.document_type.value in parser_overrides:
|
|
|
'''
|
|
|
logger.info(
|
|
|
f"Using parser_override for {document.document_type} with input value {parser_overrides[document.document_type.value]}"
|
|
|
@@ -333,7 +333,6 @@ class UnstructuredIngestionProvider(IngestionProvider):
|
|
|
f"Using OCR parser_override for {document.document_type}"
|
|
|
)
|
|
|
elements.append(element)
|
|
|
- '''
|
|
|
|
|
|
async for element in self.parse_fallback(
|
|
|
file_content,
|
|
|
@@ -344,6 +343,7 @@ class UnstructuredIngestionProvider(IngestionProvider):
|
|
|
f"Using parser_override for {document.document_type}"
|
|
|
)
|
|
|
elements.append(element)
|
|
|
+ '''
|
|
|
|
|
|
elif document.document_type in self.R2R_FALLBACK_PARSERS.keys():
|
|
|
logger.info(
|
