On 6/25/22 16:22, Ben Beasley wrote:
It appears that requests switched purely because the Apache Software
Foundation doesn’t allow LGPL dependencies[1]. Is there another
rationale that the various Python upstreams that use chardet would
find convincing? Both projects appear to be actively maintained. I
know that charset-normalizer is supposed to be much faster than
chardet, but I haven’t seen any benchmarks.
The readme [0] of charset_normalizer provides a comparison of features
and also info about performance.
[0]
https://github.com/ousret/charset_normalizer#readme