fix: Handle char expansion in islower, isupper

joshuamegnauth54 · joshuamegnauth54 · commit 113ba92266d3 · 2026-04-12T15:24:09.000-04:00
Closes: #7526 `py_islower` and `py_isupper` need to handle expansions for letter casing. Comparing chars directly can miss edge cases in certain languages. Unfortunately, like the last PR, this allocates to handle potential expansions. I also had to add `icu_casemap` as a dependency. RustPython is already using parts of icu4x so this doesn't add many transitive dependencies.
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -222,13 +222,16 @@ strum = "0.28"
 strum_macros = "0.28"
 syn = "2"
 thiserror = "2.0"
+icu_casemap = "2"
+icu_locale = "2"
 icu_properties = "2"
 icu_normalizer = "2"
 unicode-casing = "0.1.1"
 unic-ucd-age = "0.9.0"
 unicode_names2 = "2.0.0"
 widestring = "1.2.0"
 windows-sys = "0.61.2"
+writeable = "0.6.1"
 wasm-bindgen = "0.2.106"
 
 # Lints
diff --git a/crates/vm/Cargo.toml b/crates/vm/Cargo.toml
@@ -86,7 +86,10 @@ timsort = "0.1.2"
 # TODO: use unic for this; needed for title case:
 # https://github.com/RustPython/RustPython/pull/832#discussion_r275428939
 unicode-casing = { workspace = true }
+icu_casemap = { workspace = true }
+icu_locale = { workspace = true }
 icu_properties = { workspace = true }
+writeable = { workspace = true }
 
 [target.'cfg(unix)'.dependencies]
 rustix = { workspace = true }
diff --git a/crates/vm/src/anystr.rs b/crates/vm/src/anystr.rs
@@ -4,7 +4,10 @@ use crate::{
     convert::TryFromBorrowedObject,
     function::OptionalOption,
 };
+use icu_casemap::CaseMapper;
+use icu_locale::LanguageIdentifier;
 use num_traits::{cast::ToPrimitive, sign::Signed};
+use writeable::Writeable;
 
 use core::ops::Range;
 
@@ -408,12 +411,18 @@ pub trait AnyStr {
     //  unicode_islower_impl
     fn py_islower(&self) -> bool {
         let mut lower = false;
-        for c in self.elements() {
-            if c.is_uppercase() {
+        let mut lowercased = String::with_capacity(self.bytes_len());
+        let cm = CaseMapper::new();
+        for chunk in self.as_bytes().utf8_chunks().map(|c| c.valid()) {
+            let writer = cm.lowercase(chunk, &LanguageIdentifier::UNKNOWN);
+            lowercased.clear();
+            writer
+                .write_to(&mut lowercased)
+                .expect("Writing to a buffer is infallible");
+            if chunk != lowercased {
                 return false;
-            } else if !lower && c.is_lowercase() {
-                lower = true
             }
+            lower = true;
         }
         lower
     }
@@ -423,12 +432,18 @@ pub trait AnyStr {
     //  unicode_isupper_impl
     fn py_isupper(&self) -> bool {
         let mut upper = false;
-        for c in self.elements() {
-            if c.is_lowercase() {
+        let mut uppercased = String::with_capacity(self.bytes_len());
+        let cm = CaseMapper::new();
+        for chunk in self.as_bytes().utf8_chunks().map(|c| c.valid()) {
+            let writer = cm.uppercase(chunk, &LanguageIdentifier::UNKNOWN);
+            uppercased.clear();
+            writer
+                .write_to(&mut uppercased)
+                .expect("Writing to a buffer is infallible");
+            if chunk != uppercased {
                 return false;
-            } else if !upper && c.is_uppercase() {
-                upper = true
             }
+            upper = true;
         }
         upper
     }
diff --git a/extra_tests/snippets/builtin_str.py b/extra_tests/snippets/builtin_str.py
@@ -69,6 +69,8 @@
 assert not a.isdecimal()
 assert not a.isnumeric()
 assert a.istitle()
+assert "\u1c89".istitle()
+# assert "Ǳ".title() == "ǲ"
 assert a.isalpha()
 
 s = "1 2 3"
@@ -220,6 +222,8 @@
 assert "abc\t12345\txyz".expandtabs() == "abc     12345   xyz"
 assert "-".join(["1", "2", "3"]) == "1-2-3"
 assert "HALLO".isupper()
+assert "\u0295".islower()
+assert "\u1c89".isupper()
 assert "hello, my name is".partition("my ") == ("hello, ", "my ", "name is")
 assert "hello".partition("is") == ("hello", "", "")
 assert "hello, my name is".rpartition("is") == ("hello, my name ", "is", "")
@@ -236,6 +240,8 @@
 assert not "123".isidentifier()
 
 assert "Σίσυφος".swapcase() == "σΊΣΥΦΟΣ"
+assert "\u0295".swapcase() == "\u0295"
+assert "\u1c89".swapcase() == "\u1c8a"
 
 # String Formatting
 assert "{} {}".format(1, 2) == "1 2"
