repository·main·Indexed 27 days ago
https://github.com/py-pdf/pypdfA free and open-source pure-python PDF library for manipulation tasks such as splitting, merging, cropping, transforming, and extracting text or metadata. The library includes the PdfReader class for parsing and text extraction, and the PdfWriter class for generating PDF files.
pypdf is a free and open-source pure-python PDF library. It provides tools for common PDF manipulation tasks including:
pypdf is a library designed to simplify interactions with PDF documents. Its core capabilities include:
Note on OCR: pypdf does not perform Optical Character Recognition (OCR). It can only extract text that is already present in a text layer. If you cannot select or copy text in a PDF, it is likely an image, and you will need an OCR tool like tesseract to extract it.
pypdf uses the pypdf.PdfReader class to parse PDF files. The parsing process follows these steps:
PdfReader.resolved_objects. These are populated via cache_indirect_object.FlateDecode or LZWDecode) only when an object is explicitly requested via PdfReader.get_object (which internally uses PdfReader._get_object_from_stream).PDF/A is an ISO-standardized version of PDF designed for long-term archiving. It requires embedding all necessary fonts, images, and metadata (like XMP) within the document to ensure consistent reproduction without external dependencies.
Important Note: Currently, pypdf does not make any guarantees regarding PDF/A compliance. If your workflow requires strict adherence to PDF/A standards, you should be aware that pypdf may not automatically satisfy all requirements (such as embedded fonts, device-independent color spaces, or specific MarkInfo objects) during document manipulation.
A PDF file is composed of four main sections that define its content and how to navigate it:
%PDF-1.7). By default, pypdf generates files with a PDF 1.3 header but strives to support the 1.7 specification./Root catalog and /Size of the xref table) and the startxref pointer, which indicates the byte location of the xref table. The file ends with the %%EOF marker.It is important to distinguish between text extraction and Optical Character Recognition (OCR):
o, O, 0, ö) because it reads the underlying encoding.PdfDocCommon is an abstract base class used by pypdf. It provides common functionality shared between PdfReader and PdfWriter.
When interacting with the API, if a method or property is documented as accepting a PdfDocCommon type, you can provide either an instance of pypdf.PdfReader or pypdf.PdfWriter.
When choosing a PDF library, consider the following comparisons with pypdf:
pdfminer.six and focused on data extraction. It can convert PDFs to images and draw on them, but it cannot export or modify PDF files.pypdf library is the successor to PyPDF2. PyPDF2 was originally a fork of the original pyPdf, and after several years, the fork was merged back into the current pypdf project (which uses all lowercase naming).Use pypdf if you need a pure-Python library to perform the following tasks:
reader.add_form_topname("group_name") before appending or merging. This prefixes existing field names with the group name (e.g., field1 becomes group_name.field1).The beginbfchar directive in a CMap maps a specific byte to a specific Unicode character.
Example:
1 beginbfchar
<1B> <FB00>
endbfcharIn this case, the byte 1B (hexadecimal for 27) maps to the Unicode character FB00 (the ligature ff).