extract text from pdf documents


pdftxt extracts text from PDF documents, hence the name. this software was built to extract coherent indexable text and tries it's best to assemble lines and paragraphs, dehyphenate divised text and polish the bonnet.

pdftxt requires an unixoid operating system with pdf3json installed.

npm install pdftxt
var pdftxt = require("pdftxt");
pdftext("file.pdf", function(errdata) {
    if (err) return console.error(err);

Warning: This has changed again since the last version.

The data object is quite simple. It's an array with one item per page.

Every page item has it's number, width and height as properties. The proprty blocks contains an array of text block objects.

Every block object contains a bbox bounding box array with its left, bottom, width and height positions on the page, and an array of lines within that block.

All lines are trimmed and made of pure unicode with decoded entities.

        "page": num,
        "width": width,
        "height": height,
        "blocks": [{
            "bbox": [left, bottom, width, height],
            "lines": [
            // ... 

Public Domain