Small CLI that ingests full JEE papers in PDF or Word (DOCX) and outputs a clean CSV: each row contains the full question text, each option in its own column, and a separate correct answer column.

docling for PDF parsing/layout-aware extraction, TinyDB for document metadata and extracted table/image records, Chroma (local persistent) for vector embeddings and semantic search, transformers-based ...

可将PDF转换成docx文件的Python库。该项目通过PyMuPDF库提取PDF文件中的数据，然后采用python-docx库解析内容的布局、段落、图片、表格等，最后自动生成docx文件。 pdf2docx功能 - 解析和创建页面布局 ...

Standard RAG pipelines treat documents as flat strings of text. They use "fixed-size chunking" (cutting a document every 500 ...

To complete the above system, the author’s main research work includes: 1) Office document automation based on python-docx. 2) Use the Django framework to develop the website.