Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 713 Bytes

README.md

File metadata and controls

13 lines (10 loc) · 713 Bytes

webcrawler

MVP todo:

  • simple program to GET a html resource & find links (relative & absolute)
  • add the feature 'simple text analysis' (tally words used)
  • spin up a DB/Elasticsearch, connect to it, insert links & searchable data
  • make webcrawler 'oneshot' per page (it kinda already is, but keep it simple by keeping it this way)
  • perhaps another 'scheduler' service to fetch latest 'unprocessed' URL from DB, and fire 'oneshot' crawler on it (processing it).
    • exponential backoff on content diff (be the good guy!)
    • also log failure count on URL's, stop after multiple fails (consider it broken, regardless of actual error)