Problem Statement: We are having different version of a PDF (Different Content, Different table Structure, etc), so we need to Highlight all the differences among the text and the tables. Firstly I converted the Pdf's data into Markdown using "PyMuPDF4LLM" and fetching the section wise data, and then used PDFPlumber for extracting the tables for each PDF. Once we have the data, we are using sentence transformers model to find the similar content and matching them and then using Difflib Library to find the difference among them. Once we have those differences, we are using Pymupdf for highlighting back onto the PDF. Once the text Highlighting part is done, we move on to tables. We use PDFPlumber to fetch the tables, we compare each cell, and find out the differences and Highlight them back on the PDF. Issues: Not able to find the correct occurrence of change of a word in a sentence. While Highlighting the tables, Difficult to Highlight the cell when there is change in table structure. Continue reading...
