Added POSIX modifiers support for BCP47 conversion

1. Renamed function `parse_locale_code` to `posix_to_bcp47`.
2. Changed return type from `Option<String>` to `String`.
3. Changed locale cropping to trim at the first occurrence of either dot (`.`) or at sign (`@`), removing any modifiers.
4. Added tests for modifier cropping.
5. Added doc comment for `posix_to_bcp47`.

src/unix.rs

@@ -33,20 +33,41 @@ fn _get(env: &impl EnvAccess) -> Option<String> {
         .filter(|val| !val.is_empty())
         .or_else(|| env.get(LANG))?;
 
-    parse_locale_code(&code)
+    Some(posix_to_bcp47(&code))
 }
 
-fn parse_locale_code(code: &str) -> Option<String> {
-    // Some locales are returned with the char encoding too, such as `en_US.UTF-8`
-    code.split_once('.')
-        .map(|(s, _)| s)
-        .or(Some(code))
-        .map(|s| s.replace('_', "-"))
+/// Converts a POSIX locale string to a BCP47 locale string.
+///
+/// This function processes the input `code` by removing any character encoding
+/// (the part after the first `.`) and any modifiers (the part after the first `@`).
+/// It replaces underscores (`_`) with hyphens (`-`) to conform to BCP47 formatting.
+///
+/// # Examples
+///
+/// ```do-not-compile
+/// let bcp47 = posix_to_bcp47("en_US.UTF-8");
+/// assert_eq!(bcp47, "en-US");
+///
+/// let bcp47 = posix_to_bcp47("fr_FR@dict");
+/// assert_eq!(bcp47, "fr-FR");
+///
+/// let bcp47 = posix_to_bcp47("de_DE.UTF-8@euro");
+/// assert_eq!(bcp47, "de-DE");
+/// ```
+///
+/// # TODO
+///
+/// Implement POSIX to BCP47 modifier conversion.
+fn posix_to_bcp47(code: &str) -> String {
+    code.chars()
+        .take_while(|&c| c != '.' && c != '@')
+        .collect::<String>()
+        .replace('_', "-")
 }
 
 #[cfg(test)]
 mod tests {
-    use super::{parse_locale_code, EnvAccess, _get, LANG, LC_ALL, LC_CTYPE};
+    use super::{posix_to_bcp47, EnvAccess, _get, LANG, LC_ALL, LC_CTYPE};
     use std::{
         collections::HashMap,
         ffi::{OsStr, OsString},
@@ -64,21 +85,15 @@ mod tests {
 
     #[test]
     fn parse_identifier() {
-        let identifier = "fr_FR.UTF-8";
-        assert_eq!(parse_locale_code(identifier).as_deref(), Some(PARSE_LOCALE));
+        assert_eq!(posix_to_bcp47("fr_FR.UTF-8"), "fr-FR");
+        assert_eq!(posix_to_bcp47("fr_FR@euro"), "fr-FR");
+        assert_eq!(posix_to_bcp47("fr_FR.UTF-8@euro"), "fr-FR");
     }
 
     #[test]
     fn parse_non_suffixed_identifier() {
-        assert_eq!(
-            parse_locale_code(PARSE_LOCALE).as_deref(),
-            Some(PARSE_LOCALE)
-        );
-
-        assert_eq!(
-            parse_locale_code(LANG_PARSE_LOCALE).as_deref(),
-            Some(PARSE_LOCALE)
-        );
+        assert_eq!(posix_to_bcp47(PARSE_LOCALE), PARSE_LOCALE);
+        assert_eq!(posix_to_bcp47(LANG_PARSE_LOCALE), PARSE_LOCALE);
     }
 
     #[test]
@@ -111,14 +126,14 @@ mod tests {
 
         let set_code = _get(&env).unwrap();
         assert_eq!(set_code, PARSE_LOCALE);
-        assert_eq!(parse_locale_code(&set_code).as_deref(), Some(PARSE_LOCALE));
+        assert_eq!(posix_to_bcp47(&set_code), PARSE_LOCALE);
 
         // Ensure the 2nd will be skipped when empty as well.
         env.insert(LC_CTYPE.into(), String::new());
         env.insert(LANG.into(), PARSE_LOCALE.to_owned());
 
         let set_code = _get(&env).unwrap();
         assert_eq!(set_code, PARSE_LOCALE);
-        assert_eq!(parse_locale_code(&set_code).as_deref(), Some(PARSE_LOCALE));
+        assert_eq!(posix_to_bcp47(&set_code), PARSE_LOCALE);
     }
 }