North Carolina Court Portal Scraper

Automated reCAPTCHA V2 with the 2Captcha API, extracting the sitekey at runtime, fetching a solved token, and injecting it into the hidden field to unlock results with no manual step.

Executed custom JavaScript in the browser context to force Kendo UI's hidden pagination controls to their maximum page size, since the underlying <select> elements aren't directly automatable.

Reconstructed REST API endpoints from case IDs parsed out of the HTML, then queried attorney and PDF-metadata APIs separately to compile one unified record per case.

Engineered valid, authenticated PDF download URLs by assembling document type, fragment, and location IDs pulled from a prior API response, then streamed large files in chunks to avoid memory blowups.

Built a dynamic CSV writer that scans the full dataset first to size out per-case attorney columns (Attorney Name 1, Attorney Address 1, Attorney Name 2, …) to a strict flat-file spec.