From 7061051e19dc0ccbbdb694f030aedb05802287e4 Mon Sep 17 00:00:00 2001
From: nsfisis <nsfisis@gmail.com>
Date: Tue, 18 Aug 2026 01:57:02 +0900
Subject: refactor(preg): anchor A patterns at the call site

The shim carried the PCRE A (anchored) modifier alongside every compiled
pattern so preg_match2 could honour it by searching the sub-slice at the
offset. Only two call sites ever passed such a pattern, and each can cut
that slice itself, so the flag is gone from the cache, ResolvedPattern,
PregPattern and the php_regex! macro, and a pattern still carrying A is
now rejected rather than silently searched unanchored.

StringInput::tokenize and PhpFileCleaner::match search from their cursor
with a `^`-prefixed pattern instead. The rule is written down in
docs/dev/regex-porting.md.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---
 crates/shirabe-class-map-generator/src/php_file_cleaner.rs | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

(limited to 'crates/shirabe-class-map-generator')

diff --git a/crates/shirabe-class-map-generator/src/php_file_cleaner.rs b/crates/shirabe-class-map-generator/src/php_file_cleaner.rs
index 78cee261..cfd8f77f 100644
--- a/crates/shirabe-class-map-generator/src/php_file_cleaner.rs
+++ b/crates/shirabe-class-map-generator/src/php_file_cleaner.rs
@@ -1,7 +1,7 @@
 //! ref: composer/vendor/composer/class-map-generator/src/PhpFileCleaner.php
 
 use indexmap::IndexMap;
-use shirabe_php_shim::{PregMatches, preg_match2};
+use shirabe_php_shim::{PregMatches, preg_match, preg_match2};
 use std::sync::Mutex;
 
 #[derive(Debug, Clone)]
@@ -52,7 +52,7 @@ impl PhpFileCleaner {
         }
 
         let keys: String = type_config.keys().collect();
-        let rest_pattern = format!("{{[^?\"'</{}]+}}A", keys);
+        let rest_pattern = format!("{{^[^?\"'</{}]+}}", keys);
 
         *REST_PATTERN.lock().unwrap() = Some(rest_pattern);
         *TYPE_CONFIG.lock().unwrap() = Some(type_config);
@@ -103,7 +103,7 @@ impl PhpFileCleaner {
                     // no backreferences, so the three quote states (none, `'`, `"`) are expanded
                     // into separate alternatives, each capturing the identifier in its own group.
                     if let Some(r#match) = self.r#match(
-                        r#"{<<<[ \t]*(?:"([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*)"|'([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*)'|([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*))(?:\r\n|\n|\r)}A"#,
+                        r#"{^<<<[ \t]*(?:"([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*)"|'([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*)'|([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*))(?:\r\n|\n|\r)}"#,
                     ) {
                         let matched_len = r#match
                             .get(0)
@@ -282,7 +282,12 @@ impl PhpFileCleaner {
         self.index + 1 < self.len && self.contents.as_bytes()[self.index + 1] as char == char
     }
 
+    // Regex pattern compatibility:
+    // PHP runs `$regex` anchored (`A`) at `$this->index`, so it must match starting exactly there.
+    // The `regex` crate anchors only at the head of the haystack, so the search runs over the part
+    // of the contents that begins at the index and the patterns carry a leading `^` instead of the
+    // `A` modifier.
     fn r#match(&self, regex: &str) -> Option<PregMatches<'_>> {
-        preg_match2(regex, &self.contents, self.index)
+        preg_match(regex, &self.contents[self.index..])
     }
 }
-- 
cgit v1.3.1-4-g156e