/************** License stuff **************/ I have used Apache License v2.0, so the same license probably apply. /************** About the code **************/ There is two parts in this project. 1/ The first one is the back end processing. It's done with Java, takes a PDF and do two things: - Extract all the words with their position and save it in a .txt file with a JSON format. - Convert the background (PDF without the text) in images with ImageMagick (because it's faster than with PDFBox). It's almost the original source code. This project used to crawl through directories, find pdf and convert them and save the publications converted in a MySQL database. I have removed all the code related to this. I have removed all the unuse classes (things I tried). If you are interested in, just mail me (it's free). I'm saving everything in a file, but it would be better to store it in a proper database. Forget about saving it in a MySQL database with one word per entry, your database won't scale. 2/ The second part is the interface. I have rewrote the core of the interface to display the images and let the user make selection. Compare to the original version, the tree of HTML element is better balanced, the user experience is slightly better (I have spotted some bugs that I have corrected) and the sources are way more clean. There is just the core, so it's easier if you want to adapt it. I have remove all the things like export to the clipboard, add comments etc. If you are interested in the old sources, again mail me (it's free). /************** To come **************/ I plan to fork this project and try something more scalable. The idea is to to convert PDF files on the fly. For that I plan to use something like Node.js, and convert one page at a time. One of the reason I have rewrote the interface is that it's going to be easier to do it now. If you are interested/need it, contact me, it could be interesting. /************** Contact **************/ Michel Tu orphee@gmail.com http://www.neumino.com