fix: improve text search normalization #149
Conversation
camshaft
force-pushed
the
camshaft/text-find-fixes
branch
from
c5b5fad to
b9beea2
Compare
camshaft
force-pushed
the
camshaft/text-find-fixes
branch
from
b9beea2 to
c7539d3
Compare
| @@ -32,58 +52,214 @@ fn fast_find(needle: &str, haystack: &str) -> Option<Range<usize>> { | |||
|
|
|||
| /// TODO we should probaly deprecate this - it's better to enforce strict matching | |||
There was a problem hiding this comment.
still believe this?
There was a problem hiding this comment.
yep. that's why i've added the ability to warn on it 😄. still need to wire up the warnings though.
There was a problem hiding this comment.
oh I read the TODO as suggesting not allowing fuzzy finds at all, so it wouldn't be a warning, it would just not match
There was a problem hiding this comment.
Yeah that's what I want to get to. But removing the fuzzy matching right now would be a breaking change so I'd like to give some buffer for codebases to fix the matches and then we can deprecate some time in the future.
There was a problem hiding this comment.
ah I get it, so the warning would be something like "this is a fuzzy match, which is deprecated and will no longer be supported"
There was a problem hiding this comment.
Yeah that's exactly what I had in mind
Description of changes:
This change fixes a bug in the
text::findmodule that doesn't correctly map characters to the original positions. The new version is more accurate. It also includes a fuzz test that has some checks around the normalized string having the specified properties.By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.