jsvine/pdfplumber

pdfplumber is a Python library for extracting content from PDF files, particularly text and tables. Rather than treating a PDF as a flat stream of characters, it gives you access to the raw building blocks of each page: individual characters with their positions, lines, rectangles, images, and annotations. This level of detail makes it possible to extract data from PDFs that have complex layouts, like government documents, financial reports, or data tables, where simply copying the text would lose the structure. The main use cases are pulling plain text from a page, pulling tables out of a page and returning them as lists of rows and cells, and extracting form field values. Because pdfplumber knows where each character sits on the page, you can crop a region of interest and only extract content from that area, which is useful when a PDF mixes tables with body text that you want to ignore. Installation is a single pip command. The library can be used either through Python code or through a command-line tool that outputs information about every object in the PDF as CSV or JSON. A visual debugging feature lets you draw outlines around the objects pdfplumber detects, which helps when you are trying to understand why a table extraction is not picking up the right cells. The library works best on machine-generated PDFs rather than scanned documents. Scanned PDFs are images of text rather than actual text characters, so there is nothing to extract without first running optical character recognition on the image, which pdfplumber does not do. If the PDF was created by a word processor, a spreadsheet application, or a report generator, pdfplumber can typically read it well. It is built on top of an existing PDF parsing library called pdfminer.six and adds higher-level extraction tools on top. The full README is longer than what was shown.