Support Unicode in field/variant renaming

Author: ginnyTheCat

serde_derive/src/internals/case.rs

@@ -57,54 +57,47 @@ impl RenameRule {
     pub fn apply_to_variant(self, variant: &str) -> String {
         match self {
             None | PascalCase => variant.to_owned(),
-            LowerCase => variant.to_ascii_lowercase(),
-            UpperCase => variant.to_ascii_uppercase(),
-            CamelCase => variant[..1].to_ascii_lowercase() + &variant[1..],
-            SnakeCase => {
-                let mut snake = String::new();
-                for (i, ch) in variant.char_indices() {
-                    if i > 0 && ch.is_uppercase() {
-                        snake.push('_');
-                    }
-                    snake.push(ch.to_ascii_lowercase());
-                }
-                snake
+            LowerCase => variant.to_lowercase(),
+            UpperCase => variant.to_uppercase(),
+            CamelCase => {
+                let mut chars = variant.chars();
+                let Some(first) = chars.next() else {
+                    return String::new();
+                };
+
+                let mut camel = String::with_capacity(variant.len());
+                camel.extend(first.to_uppercase());
+                camel.push_str(chars.as_str());
+                camel
             }
-            ScreamingSnakeCase => SnakeCase.apply_to_variant(variant).to_ascii_uppercase(),
-            KebabCase => SnakeCase.apply_to_variant(variant).replace('_', "-"),
-            ScreamingKebabCase => ScreamingSnakeCase
-                .apply_to_variant(variant)
-                .replace('_', "-"),
+            SnakeCase => separate_pascal_case(variant, false, '_'),
+            ScreamingSnakeCase => separate_pascal_case(variant, true, '_'),
+            KebabCase => separate_pascal_case(variant, false, '-'),
+            ScreamingKebabCase => separate_pascal_case(variant, true, '-'),
         }
     }
 
     /// Apply a renaming rule to a struct field, returning the version expected in the source.
     pub fn apply_to_field(self, field: &str) -> String {
         match self {
             None | LowerCase | SnakeCase => field.to_owned(),
-            UpperCase => field.to_ascii_uppercase(),
-            PascalCase => {
-                let mut pascal = String::new();
-                let mut capitalize = true;
-                for ch in field.chars() {
-                    if ch == '_' {
-                        capitalize = true;
-                    } else if capitalize {
-                        pascal.push(ch.to_ascii_uppercase());
-                        capitalize = false;
-                    } else {
-                        pascal.push(ch);
+            UpperCase => field.to_uppercase(),
+            PascalCase => snake_case_to_camel_case(field, true),
+            CamelCase => snake_case_to_camel_case(field, false),
+            ScreamingSnakeCase => field.to_uppercase(),
+            KebabCase => field.replace('_', "-"),
+            ScreamingKebabCase => {
+                let kebab = field.to_uppercase();
+
+                let mut kebab_vec = Vec::from(kebab);
+                for b in &mut kebab_vec {
+                    if *b == b'_' {
+                        *b = b'-';
                     }
                 }
-                pascal
-            }
-            CamelCase => {
-                let pascal = PascalCase.apply_to_field(field);
-                pascal[..1].to_ascii_lowercase() + &pascal[1..]
+                // we only replaced ASCII in place, it's still valid UTF-8
+                String::try_from(kebab_vec).unwrap()
             }
-            ScreamingSnakeCase => field.to_ascii_uppercase(),
-            KebabCase => field.replace('_', "-"),
-            ScreamingKebabCase => ScreamingSnakeCase.apply_to_field(field).replace('_', "-"),
         }
     }
 
@@ -117,6 +110,37 @@ impl RenameRule {
     }
 }
 
+fn separate_pascal_case(pascal: &str, screaming: bool, line: char) -> String {
+    let mut separated = String::with_capacity(pascal.len());
+    for (i, ch) in pascal.char_indices() {
+        if i > 0 && ch.is_uppercase() {
+            separated.push(line);
+        }
+        if screaming {
+            separated.extend(ch.to_uppercase());
+        } else {
+            separated.extend(ch.to_lowercase());
+        }
+    }
+    separated
+}
+
+fn snake_case_to_camel_case(snake: &str, pascal: bool) -> String {
+    let mut camel = String::new();
+    let mut capitalize = pascal;
+    for ch in snake.chars() {
+        if ch == '_' {
+            capitalize = true;
+        } else if capitalize {
+            camel.extend(ch.to_uppercase());
+            capitalize = false;
+        } else {
+            camel.push(ch);
+        }
+    }
+    camel
+}
+
 pub struct ParseError<'a> {
     unknown: &'a str,
 }

test_suite/tests/test_macros.rs

@@ -658,8 +658,7 @@ fn test_rename_all() {
     enum E {
         #[serde(rename_all = "camelCase")]
         Serialize {
-            serialize: bool,
-            serialize_seq: bool,
+            serialize: bool, etwas_ändern: bool
         },
         #[serde(rename_all = "kebab-case")]
         SerializeSeq {
@@ -676,21 +675,21 @@ fn test_rename_all() {
     #[derive(Serialize, Deserialize, Debug, PartialEq)]
     #[serde(rename_all = "PascalCase")]
     struct S {
-        serialize: bool,
+        ändern: bool,
         serialize_seq: bool,
     }
 
     #[derive(Serialize, Deserialize, Debug, PartialEq)]
     #[serde(rename_all = "SCREAMING-KEBAB-CASE")]
     struct ScreamingKebab {
-        serialize: bool,
+        grüße: bool,
         serialize_seq: bool,
     }
 
     assert_tokens(
         &E::Serialize {
             serialize: true,
-            serialize_seq: true,
+            etwas_ändern: true,
         },
         &[
             Token::StructVariant {
@@ -700,7 +699,7 @@ fn test_rename_all() {
             },
             Token::Str("serialize"),
             Token::Bool(true),
-            Token::Str("serializeSeq"),
+            Token::Str("etwasÄndern"),
             Token::Bool(true),
             Token::StructVariantEnd,
         ],
@@ -746,12 +745,12 @@ fn test_rename_all() {
 
     assert_tokens(
         &S {
-            serialize: true,
+            ändern: true,
             serialize_seq: true,
         },
         &[
             Token::Struct { name: "S", len: 2 },
-            Token::Str("Serialize"),
+            Token::Str("Ändern"),
             Token::Bool(true),
             Token::Str("SerializeSeq"),
             Token::Bool(true),
@@ -761,15 +760,15 @@ fn test_rename_all() {
 
     assert_tokens(
         &ScreamingKebab {
-            serialize: true,
+            grüße: true,
             serialize_seq: true,
         },
         &[
             Token::Struct {
                 name: "ScreamingKebab",
                 len: 2,
             },
-            Token::Str("SERIALIZE"),
+            Token::Str("GRÜSSE"),
             Token::Bool(true),
             Token::Str("SERIALIZE-SEQ"),
             Token::Bool(true),