mirror of
https://github.com/Mintplex-Labs/langchain-python.git
synced 2026-07-21 00:35:23 -04:00
123-fake-st
9bd6e9df36
- **Description:** Update 5 pdf document loaders in `langchain.document_loaders.pdf`, to store a url in the metadata (instead of a temporary, local file path) if the user provides a web path to a pdf: `PyPDFium2Loader`, `PDFMinerLoader`, `PDFMinerPDFasHTMLLoader`, `PyMuPDFLoader`, and `PDFPlumberLoader` were updated. - The updates follow the approach used to update `PyPDFLoader` for the same behavior in #12092 - The `PyMuPDFLoader` changes required additional work in updating `langchain.document_loaders.parsers.pdf.PyMuPDFParser` to be able to process either an `io.BufferedReader` (from local pdf) or `io.BytesIO` (from online pdf) - The `PDFMinerPDFasHTMLLoader` change used a simpler approach since the metadata is assigned by the loader and not the parser - **Issue:** Fixes #7034 - **Dependencies:** None ```python # PyPDFium2Loader example: # old behavior >>> from langchain.document_loaders import PyPDFium2Loader >>> loader = PyPDFium2Loader('https://arxiv.org/pdf/1706.03762.pdf') >>> docs = loader.load() >>> docs[0].metadata {'source': '/var/folders/7z/d5dt407n673drh1f5cm8spj40000gn/T/tmpm5oqa92f/tmp.pdf', 'page': 0} # new behavior >>> from langchain.document_loaders import PyPDFium2Loader >>> loader = PyPDFium2Loader('https://arxiv.org/pdf/1706.03762.pdf') >>> docs = loader.load() >>> docs[0].metadata {'source': 'https://arxiv.org/pdf/1706.03762.pdf', 'page': 0} ```