agronholm / anyio / 38068867565

92%

push

github

Use an incremental encoder in TextSendStream so the BOM is written once (#1405)

codecs.getencoder() is stateless, so every send() with a BOM-emitting encoding (utf-16, utf-32, utf-8-sig) prepended a new byte order mark and the peer decoded a stray U+FEFF in front of every message after the first. TextReceiveStream already uses getincrementaldecoder() for the same reason.

4 of 4 new or added lines in 1 file covered. (100.0%)

6839 of 7479 relevant lines covered (91.44%)

11.67 hits per line