shipfastlabs/parsel

Parsel is an open-source PHP library for reading and extracting content from documents including PDFs, Word files, spreadsheets, presentations, and images. Everything runs locally on your own server, so your documents are never sent to a third-party service. The library is available as a Composer package, which is the standard way to install PHP dependencies. The API is designed to feel straightforward. You point it at a file or provide raw file bytes, choose what you want back, and call a method. The simplest use is getting the full text of a document as a plain string. For more detailed work, a structured parse returns a document object containing each page, and within each page a list of text items with their position coordinates, which is useful when you need to know where on the page something appears rather than just what it says. For documents that are scanned images rather than text-based files, Parsel can apply OCR, which stands for optical character recognition: a process that looks at the image and identifies the characters in it. OCR is off by default because it takes more time. You can turn it on per request and specify a language, a path to language data files, or the number of parallel workers. The underlying OCR engine is Tesseract, accessed through a companion tool called liteparse. Page selection lets you limit parsing to specific pages, a range, or a maximum count rather than always processing the entire document. For very large documents, a streaming mode lets you process pages one at a time without loading everything into memory. Other supported features include rendering page screenshots to image files, saving output as plain text or JSON, setting rendering resolution, providing a password for encrypted PDFs, and adjusting the timeout for slow documents. The library requires PHP 8.4 or newer and depends on LibreOffice for Office format conversion and ImageMagick for image handling, both of which are standard open-source tools available on Linux, macOS, and Windows.