zed-industries/zed
1
1
Fork 0
mirror of https://github.com/zed-industries/zed.git synced 2024-12-26 20:32:22 +03:00
Code Issues Packages Projects Releases Wiki Activity

Fix fuzzy string match invariant check (#22032)

Browse Source 
Version in #21983 only handled out of range issues rather than utf-8
boundary issues (thanks to @s3bba for pointing this out)

Release Notes:

- N/A
This commit is contained in:
Michael Sloan 2024-12-15 01:15:22 -07:00 committed by GitHub
parent 7e6233d70f
commit f64fcedabb
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 14 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
crates/fuzzy/src/strings.rs
View File

@ -61,24 +61,24 @@ impl StringMatch {
let mut positions = self.positions.iter().peekable();
iter::from_fn(move || {
if let Some(start) = positions.next().copied() {
if start >= self.string.len() {
let Some(char_len) = self.char_len_at_index(start) else {
log::error!(
"Invariant violation: Index {start} out of range in string {:?}",
"Invariant violation: Index {start} out of range or not on a utf-8 boundary in string {:?}",
self.string
);
return None;
}
let mut end = start + self.char_len_at_index(start);
};
let mut end = start + char_len;
while let Some(next_start) = positions.peek() {
if end == **next_start {
if end >= self.string.len() {
let Some(char_len) = self.char_len_at_index(end) else {
log::error!(
"Invariant violation: Index {end} out of range in string {:?}",
"Invariant violation: Index {end} out of range or not on a utf-8 boundary in string {:?}",
self.string
);
return None;
}
end += self.char_len_at_index(end);
};
end += char_len;
positions.next();
} else {
break;
@ -91,8 +91,12 @@ impl StringMatch {
})
}
fn char_len_at_index(&self, ix: usize) -> usize {
self.string[ix..].chars().next().unwrap().len_utf8()
/// Gets the byte length of the utf-8 character at a byte offset. If the index is out of range
/// or not on a utf-8 boundary then None is returned.
fn char_len_at_index(&self, ix: usize) -> Option<usize> {
self.string
.get(ix..)
.and_then(|slice| slice.chars().next().map(|char| char.len_utf8()))
}
}