'eng' claimed in original title that "urllib.parse.parse_qsl cannot parse large data".  On original PR, said problem with 6-7 millions bytes.

Claim should be backed up by a generated example that fails with original code and succeeds with new code.  Claims of 'faster' also needs some examples.

Original PRs must nearly all propose merging a branch created from main into main.  Performance enhancements are often not backported.