Handling Complex Datasets In Project Delivery

Data resellers weighing a PDF-only source: my House disclosure prototype, honestly

Plenty of useful public data lives only inside PDFs, which is why vendors tend to skip it. Here is a small prototype against the House Clerk's disclosure portal, with its real status.

The portal searches by last name and filing year. My script opens a session and picks up the anti-forgery token, searches, downloads the PDF, and reads the periodic transaction table with pdfplumber.

Fields returned: owner, asset, type, date, notification_date, amount, cap_gains, description.

Status: a 382-line command-line prototype, tested on three sample PDFs. There is no volume to quote and I am not calling it a feed.

What it does show is the shape. A form, a token and a table inside a PDF is a common portal pattern, and newsletters and alert services use this kind of data.

Which PDF-only sources do clients ask you about that you currently turn down?

11 comments

Handling Complex Datasets In Project Delivery