After a bit of testing, I found out that my block runs into an infinite loop if you're searching for a single character, and that character is at the end. I fixed it by allowing the substring block to report an empty string if the start index is greater than the length of the string.
And just for good measure, here's a script of these blocks with fixed
And just cause, I decided to time the blocks as well.
(I used @dardoro's timing script)
I just realized something after 4+ years of snap that makes my block so much faster
That only works in simple cases
And I don't imagine -1 is a defined use case (it might be but I'd personally avoid using it)
Oh, that’s really strange.
compilation split execution time by 2!
this new one?
Yeah, this one: you rock !
Thanks!
Cool, but can you change overlap to a drop-down please
What's the point of making it a dropdown when it can just be a switch? If there were more options, then I'd understand having it be a menu.
I mean, it just looks sorta clunky with the parentheses, question Mark, long label, and Boolean.
Um the block still uses the old one
Sorry, this is the new version !
It's a very big string: 489000 char
on my computer, your block (non-overlapping) run in 650ms
on my computer, bh block (overlapping) run in 2790ms
Do you wanna try to beat bh speed with the overtapping version ?
Ok I will try to make a faster version than
