3 ms·

I think the comment was about bulk extraction of data. For instance, it is pretty difficult to do bibliographic studies if all the articles you have are provide

I think the comment was about bulk extraction of data. For instance, it is pretty difficult to do bibliographic studies if all the articles you have are provided only as PDFs (even if it is text pdf). For starters, I do not think PDFs have a notion resembling a "paragraph of text", because every symbol is placed separately with its own unique coordinates. Extracting tables of numbers from PDFs is also a pain.

maxxxxx 8y ago
Exactly. I have written software for legal document search. Most of what they get is in PDF and it’s a major PITA to get data out of them. Forget about tables. Just try to extract text without some garbled characters and you will lose your mind.