github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-08-26 21:11:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix revue

Browse Source
This commit is contained in:
Philippe Prados 2025-03-07 16:25:33 +01:00
parent b76e9bd579
commit 496a933f68
1 changed files with 14 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
libs/community/langchain_community/document_loaders/parsers/pdf.py
View File

@ -1470,6 +1470,17 @@ class PDFPlumberParser(BaseBlobParser):
return extract_from_images_with_rapidocr(images) return extract_from_images_with_rapidocr(images)
_map_extract_tables: Dict[Literal["markdown", "html", None], str] = {
"markdown": "",
"html": "But, use html syntax for convert all tables. ",
}
_map_extract_images = {
RapidOCRBlobParser: "",
TesseractBlobParser: "",
LLMImageBlobParser: "If you come across a picture, "
"diagram or other illustration, "
"describe it. ",
}
class ZeroxPDFParser(BaseBlobParser): class ZeroxPDFParser(BaseBlobParser):
"""Parse a blob from a PDF using `py-zerox` library. """Parse a blob from a PDF using `py-zerox` library.
@ -1520,25 +1531,8 @@ class ZeroxPDFParser(BaseBlobParser):
print(docs[0].page_content[:100]) print(docs[0].page_content[:100])
print(docs[0].metadata) print(docs[0].metadata)
""" """
warnings.filterwarnings(
"ignore",
module=r"^pyzerox.models.modellitellm$",
message=r"\s*Custom system prompt was provided which.*",
)
_warn_images_to_text = False _warn_images_to_text = False
_warn_creator = False _warn_creator = False
_map_extract_tables: Dict[Literal["markdown", "html", None], str] = {
"markdown": "",
"html": "But, use html syntax for convert all tables. ",
}
_map_extract_images = {
RapidOCRBlobParser: "",
TesseractBlobParser: "",
LLMImageBlobParser: "If you come across a picture, "
"diagram or other illustration, "
"describe it. ",
}
_prompt = ( _prompt = (
"Convert the following PDF page to markdown. " "Convert the following PDF page to markdown. "
"{prompt_tables}" "{prompt_tables}"
@ -1680,10 +1674,10 @@ class ZeroxPDFParser(BaseBlobParser):
zerox_prompt = self.custom_system_prompt zerox_prompt = self.custom_system_prompt
if not zerox_prompt and self.images_parser or self.extract_tables: if not zerox_prompt and self.images_parser or self.extract_tables:
prompt_tables = ZeroxPDFParser._map_extract_tables[self.extract_tables] prompt_tables = _map_extract_tables[self.extract_tables]
clazz = self.images_parser.__class__ clazz = self.images_parser.__class__
if clazz in ZeroxPDFParser._map_extract_images: if clazz in _map_extract_images:
prompt_images = ZeroxPDFParser._map_extract_images[clazz] prompt_images = _map_extract_images[clazz]
else: else:
if not ZeroxPDFParser._warn_creator: if not ZeroxPDFParser._warn_creator:
ZeroxPDFParser._warn_creator = True ZeroxPDFParser._warn_creator = True