A method to grab the comments from YouTube videos
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
JustAnotherArchivist 1df5702c02 Fix deep recursion issues on videos with many comments 2 months ago
LICENSE Initial commit 2 months ago
README.md Fix deep recursion issues on videos with many comments 2 months ago
comments Initial commit 2 months ago
comments.py Fix deep recursion issues on videos with many comments 2 months ago

README.md

A method to grab the comments from YouTube videos

  • Requires qwarc and realpath from GNU coreutils 8.23+.
  • Execute as comments VIDEOID where VIDEOID is the 11-character video ID from YouTube.
  • You can pass multiple video IDs at once as well: comments VIDEOID1 VIDEOID2 .... They get executed sequentially.
  • Comments are grabbed in all available sort orders (i.e. “top” and “new”, though “top” is retrieved twice since YT returns two continuation tokens for it), including replies and nesting.
  • The “top” sort order sometimes doesn’t return all comments but might be missing some. The reason for this is unclear.
  • On videos with many comments, the “top” sort order retrieval will fail rapidly due to the continuation token being too long and causing an HTTP 413 Request Entity Too Large error. (The “new” sort order should succeed.)
  • Everything’s written to a few files in the current directory called youtube-comments-VIDEOID-DATE*.
    • After the retrieval finished cleanly and you’re satisfied with the results, you can delete the .db and .log files. The former is essentially useless, and the latter is contained in the -meta.warc.gz file.