Robust Multi-Format Document Conversion Engine

I need a self-contained conversion engine that takes common office files and produces clean, faithful output in other formats without losing a single detail. The converter must reliably handle the following flows: • PDF → Word • PDF → XML • Word → HTML • Word → XML For every run, the resulting file must keep all original elements intact—overall layout and styling, embedded images and graphics, live hyperlinks, tables, italics, and any special characters. An end user should be able to open the converted document and see no visual or structural difference from the source, aside from the new file type. I am open to whichever stack or library you believe best meets these goals (e.g., Python with PDFPlumber + python-docx, Java with Apache POI or Aspose, C# with iText, or even a headless LibreOffice/Pandoc workflow). The key requirement is accuracy and speed under batch processing. Deliverables • Source code with clear build/run instructions • Command-line tool or callable API that receives an input path, output path, and target format • Brief read-me describing any third-party dependencies and their licences • A small test suite proving conversions of at least five sample files per route, highlighting preservation of all critical aspects Acceptance criteria 1. Pixel-level comparison of original vs. converted screenshots shows no misalignment. 2. Automated diff confirms all hyperlinks, image counts, and table structures are present. 3. All sample files pass without manual correction. If you already have a similar solution, let me know; otherwise outline your proposed approach, main libraries, and estimated timeline so we can move forward quickly.

Project ID: 40476194