lenny (01:03:30)
the program already has a set of rules to workaround problematic lines, and they are made of regex as well

mirrorbird (01:02:37)
lenny what language/environment are you using?

mirrorbird (01:02:12)
(did you know there's such a thing as a regex denial-of-service attack?)

lenny (01:02:08)
i'm dealing with some subtitles, which have a few lines with incomplete suspensive points

mirrorbird (01:00:44)
and by 1 KB i mean 1 MB

mirrorbird (01:00:27)
as i said: if i knew the string was always gonna be short (like <1 KB) i would just use split. it's 2025, memory is cheap.

mirrorbird (01:00:00)
if you can just read your string left-to-right and set/unset a boolean flag when you see something that isn't A-Z, it may be cheaper and faster.

lenny (00:59:30)
since "regular" search won't be useful to match exactly those two points only

mirrorbird (00:59:26)
lenny: i would consider carefully the need for regex. if you are dealing with 'human book text' (where a word may end on a comma, period, etc.) then maybe best. if it's just a load of words with spaces, god no.

lenny (00:58:37)
it is a really trivial task when you think of it, but not when it comes to use regex to solve it

Dawnfall

Dafunk (Dejan Subotin)

lenny

Time Left: 7:50