aboutsummaryrefslogtreecommitdiffhomepage
path: root/crates/shirabe-symfony-string/src/byte_string.rs
diff options
context:
space:
mode:
authornsfisis <nsfisis@gmail.com>2026-08-09 11:13:06 +0900
committernsfisis <nsfisis@gmail.com>2026-08-09 11:13:06 +0900
commit880ba0fd1d05faf98588cc326b3d9fbe625ebf2b (patch)
tree055341e80b6297c59cfb267863e0a21419aa6ac3 /crates/shirabe-symfony-string/src/byte_string.rs
parent66c3eba15ba6302d43de057a9063f7feee8c6fb3 (diff)
downloadphp-shirabe-880ba0fd1d05faf98588cc326b3d9fbe625ebf2b.tar.gz
php-shirabe-880ba0fd1d05faf98588cc326b3d9fbe625ebf2b.tar.zst
php-shirabe-880ba0fd1d05faf98588cc326b3d9fbe625ebf2b.zip
refactor(symfony-string): extract symfony/string into the shirabe-symfony-string crate
Move `Symfony\Component\String` out of shirabe-external-packages and into its own crate, so the path is `shirabe_symfony_string::ByteString` instead of `shirabe_external_packages::symfony::string::ByteString`. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Diffstat (limited to 'crates/shirabe-symfony-string/src/byte_string.rs')
-rw-r--r--crates/shirabe-symfony-string/src/byte_string.rs45
1 files changed, 45 insertions, 0 deletions
diff --git a/crates/shirabe-symfony-string/src/byte_string.rs b/crates/shirabe-symfony-string/src/byte_string.rs
new file mode 100644
index 00000000..cf597bbf
--- /dev/null
+++ b/crates/shirabe-symfony-string/src/byte_string.rs
@@ -0,0 +1,45 @@
+//! ref: composer/vendor/symfony/string/ByteString.php
+
+use crate::code_point_string::CodePointString;
+
+#[derive(Debug, Clone)]
+pub struct ByteString {
+ pub(crate) string: String,
+}
+
+impl ByteString {
+ pub fn new(string: &str) -> Self {
+ Self {
+ string: string.to_string(),
+ }
+ }
+
+ /// `from_encoding` is `""` for PHP's `null`.
+ pub fn to_code_point_string(&self, from_encoding: &str) -> CodePointString {
+ // The source `string` is always valid UTF-8 (Rust `String`/`&str` guarantee), so
+ // `preg_match('//u', ...)` always holds.
+ if matches!(from_encoding, "" | "utf8" | "utf-8" | "UTF8" | "UTF-8") {
+ return CodePointString {
+ string: self.string.clone(),
+ };
+ }
+
+ let valid_encoding = shirabe_php_shim::mb_detect_encoding(
+ &self.string,
+ Some(vec![from_encoding.to_string()]),
+ true,
+ )
+ .is_some();
+ // PHP throws InvalidArgumentException. Callers detect `from_encoding` from this very
+ // string, so a mismatch is a programming error.
+ assert!(valid_encoding, "Invalid \"{}\" string.", from_encoding);
+
+ CodePointString {
+ string: shirabe_php_shim::mb_convert_encoding(
+ self.string.clone().into_bytes(),
+ "UTF-8",
+ from_encoding,
+ ),
+ }
+ }
+}