fix #124714 str.to_lowercase sigma handling

This commit is contained in:
Marcondiro 2024-05-05 23:01:09 +02:00
parent e3029d220f
commit bbdf97254a
No known key found for this signature in database
GPG Key ID: A622B776F2B49522
2 changed files with 9 additions and 4 deletions

View File

@ -375,14 +375,16 @@ pub fn to_lowercase(&self) -> String {
// Safety: We have written only valid ASCII to our vec // Safety: We have written only valid ASCII to our vec
let mut s = unsafe { String::from_utf8_unchecked(out) }; let mut s = unsafe { String::from_utf8_unchecked(out) };
for (i, c) in rest[..].char_indices() { for (i, c) in rest.char_indices() {
if c == 'Σ' { if c == 'Σ' {
// Σ maps to σ, except at the end of a word where it maps to ς. // Σ maps to σ, except at the end of a word where it maps to ς.
// This is the only conditional (contextual) but language-independent mapping // This is the only conditional (contextual) but language-independent mapping
// in `SpecialCasing.txt`, // in `SpecialCasing.txt`,
// so hard-code it rather than have a generic "condition" mechanism. // so hard-code it rather than have a generic "condition" mechanism.
// See https://github.com/rust-lang/rust/issues/26035 // See https://github.com/rust-lang/rust/issues/26035
map_uppercase_sigma(rest, i, &mut s) let out_len = self.len() - rest.len();
let sigma_lowercase = map_uppercase_sigma(&self, i + out_len);
s.push(sigma_lowercase);
} else { } else {
match conversions::to_lower(c) { match conversions::to_lower(c) {
[a, '\0', _] => s.push(a), [a, '\0', _] => s.push(a),
@ -400,13 +402,13 @@ pub fn to_lowercase(&self) -> String {
} }
return s; return s;
fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) { fn map_uppercase_sigma(from: &str, i: usize) -> char {
// See https://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992 // See https://www.unicode.org/versions/Unicode7.0.0/ch03.pdf#G33992
// for the definition of `Final_Sigma`. // for the definition of `Final_Sigma`.
debug_assert!('Σ'.len_utf8() == 2); debug_assert!('Σ'.len_utf8() == 2);
let is_word_final = case_ignorable_then_cased(from[..i].chars().rev()) let is_word_final = case_ignorable_then_cased(from[..i].chars().rev())
&& !case_ignorable_then_cased(from[i + 2..].chars()); && !case_ignorable_then_cased(from[i + 2..].chars());
to.push_str(if is_word_final { "ς" } else { "σ" }); if is_word_final { 'ς' } else { 'σ' }
} }
fn case_ignorable_then_cased<I: Iterator<Item = char>>(iter: I) -> bool { fn case_ignorable_then_cased<I: Iterator<Item = char>>(iter: I) -> bool {

View File

@ -1848,6 +1848,9 @@ fn to_lowercase() {
assert_eq!("ΑΣ'Α".to_lowercase(), "ασ'α"); assert_eq!("ΑΣ'Α".to_lowercase(), "ασ'α");
assert_eq!("ΑΣ''Α".to_lowercase(), "ασ''α"); assert_eq!("ΑΣ''Α".to_lowercase(), "ασ''α");
// https://github.com/rust-lang/rust/issues/124714
assert_eq!("abcdefghijklmnopΣ".to_lowercase(), "abcdefghijklmnopς");
// a really long string that has it's lowercase form // a really long string that has it's lowercase form
// even longer. this tests that implementations don't assume // even longer. this tests that implementations don't assume
// an incorrect upper bound on allocations // an incorrect upper bound on allocations