You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Not 100% this is the root cause but it isn't helping...
Our URL and email detection is very rudimentary With comment-depth.txt, we see the whole file as a valid email username, so iterate to the very end and then see no @ at the end and fail. So on nearly every byte we are then looking ahead through the entire file, checking every byte. Twice.
This also means we checking every token type on every byte. We aren't getting any savings from one capturing 10 or 20 bytes.
Capping email usernames and url userinfo's at 1024 characters makes typos take 3s on comment-depth.txtwith a debug build.
In some test data for rinja, they check some parsing corner cases.
Unfortunately for us, also hit a performance corner case.
The entire file was a valid email username but without an `@`.
This mean for every byte, we checked that every byte after it was a
valid username but then backtracked at the end, repeating this until the
whole file was read.
Fixescrate-ci#1088
comment-depth.txt
filter-recursion.txt
My guess would be that some sort of exponential backtracking is happening. I guess the program would ultimately finish.
The text was updated successfully, but these errors were encountered: