PDF-Dicer

Split PDF files into many based on barcode separators.

This is useful if scanning a large number of documents in a batch (e.g. via an automated office scanner) which then need to be split up again.

PDF-Dicer takes a single PDF file made up of multiple scanned documents. Each sub-document has a starting and ending barcode.

PDF-Dicer takes this file, splits on each barcode set, validates the barcodes and outputs back into individual files.

Installing

This module requires ImageMagick, GhostScript and Poppler.

You can install them as follows:

• Ubuntu Linux - sudo apt-get install imagemagick ghostscript poppler-utils pdftk
• OSX (Yosemite) - brew install imagemagick ghostscript poppler
  • Install PDFTK from website.

Example

var pdfDicer = require('pdf-dicer');
 
var dicer = new pdfDicer();
 
dicer
    .on('split', (data, buffer) => {
      fs.writeFile('output.pdf', buffer);
    })
    .split('input.pdf', function(err, output) {
        if (err) console.log(`Something went wrong: ${err}`);
    });

API

dicer (class)

The main class of this module.

The constructor takes an optional settings object which is used to populate the initial setup.

var dicer = new pdfDicer({driver: 'quagga'});

dicer.settings (object)

An object of the instance settings. These can be set either on construction, via a call to set() or directly.

The following settings are supported:

Setting	Type	Default	Profile	Description
areas	Array	{top:'0%',right:'0%',left:'0%',bottom:'0%'}	Quagga	The areas of the input pages that Quagga should scan
imageFormat	String	png (Quagga), tif (Bardecode)	All	The intermediate image format to use before processing the barcode
magickOptions	Object	Various (Quagga), {} (Bardecode)	All	Additional options to pass to ImageMagick when converting the PDF to images
bardecode	Object	See below	Bardecode	Options specific to Bardecode
bardecode.bin	String	/opt/bardecoder/bin/bardecode	Bardecode	Path to the bardecode binary
bardecode.checkEvaluation	Boolean	true	Bardecode	Check that the barcode doesn't end in ??? and raise a warning if it does
bardecode.serial	String	""	Bardecode	Your Bardecode serial number
filter	Function	(page) => true	All	Optional filter to discard pages before calculating ranges
quagga	Object	See below	Quagga	Options specific to Quagga
quagga.locate	Boolean	false	Quagga	Indicates if Quagga should try to detect the barcode or we should use areas
quagga.decoder	Object	{readers:['code_128_reader'],multiple: false}	Quagga	Options passed to the Quagga decoder
temp	Object	See below	All	Options passed to Temp when generating a temporary directory
tempClean	Boolean	true	All	Automatically erase the temporary directory when done
temp.prefix	String	pdfdicer-	All	The prefix used when generating a temporary directory
threads	Object	See below	All	Options used for async threading
threads.pages	Number	1	All	The number of threads allowed to run simultaneously when processing pages
threads.areas	Number	1	Quagga	The number of threads allowed to run simultaneously when processing page areas

dicer.set(setting, value)

Convenience function to quickly set a setting. Dotted notation is allowed for setting.

dicer.profile(profile)

Convenience function to configure the module with optimal settings for the supported barcode readers.

Supported profiles are:

• quagga
• bardecode

dicer.split(inputPath, callback)

Process the inputPath (usually a PDF) and split it into multiple PDF files.

Hook into the output of this function by trapping events.

Events

The following events are fired by this module:

Event	Arguments	Description
stage	(stageName)	Fired for each stage of operation. ENUM: 'init', 'readPDF', 'readPages', 'extracted', 'filtering', 'loadRange', 'preSplit'
tempDir	(path)	Fired when a temp directory has been allocated
pageConverted	(page, pageOffset)	Fired for each page that is converted
pagesConverted	(pages)	Fired when all pages have been converted
pageAnalyze	(page)	Fired before an individual page is analyzed
barcodeFiltered	(page)	Fired if a page is filtered out
barcodePassed	(page)	Fired if a page passes filtering and is not filtered out
pageAnalyzed	(page)	Fired after a page has been analyzed
pagesAnalyzed	(pages)	Fired when all pages have been analyzed
split	(range, buffer)	Fired when a range has been detected and a buffer is ready