Enhance Python PDF-to-CSV Extractor

I have an existing Python script whose sole purpose is data extraction: it turns PDF documents into CSV files. Right now the results are incomplete and the code is a little fragile. I need it tightened up so every piece of data in each PDF is captured and written to a clean, well-structured CSV.

What I already have
• A working—but imperfect—Python script
• Sample PDFs that show the range of layouts the tool must handle
• A sample CSV that illustrates the column order I expect

What needs to improve
• Reliable parsing across multiple pages and varied table structures
• Accurate capture of every field, not just the obvious text blocks
• Clear, readable code with comments so future tweaks are simple
• A straightforward command-line call such as
python pdf2csv.py input.pdf output.csv

Useful libraries are entirely up to you—pdfplumber, PyPDF2, tabula-py, Camelot, pandas, or a combo—so long as the final script runs on standard Python 3 and requires only pip-installable packages.

Deliverables
1. Updated script (single .py file or a small module)
2. Requirements.txt listing any external dependencies
3. One example CSV generated from my sample PDFs to prove full data coverage
4. Brief README with run instructions

Acceptance Criteria
• Running the script on my test PDFs produces a CSV that matches the source data exactly, column for column and row for row.
• No hard-coded file paths; everything is parameterised.
• Code executes without warnings or errors on Python 3.10 under Windows and Linux.

Please keep the focus on robust extraction—the project’s primary goal—so I can drop new PDFs in and get accurate CSVs every time.