refactor(pcre): hand back the match instead of copying it out

Preg::match4 and Preg::replace_callback gave callers a
PregMatchedGroups: an IndexMap rebuilt from the match with an owned
String per group, plus a second String for a named group's name key.
That is the copy PregMatches shed when it started wrapping
regex::Captures, reinstated one layer up -- and nearly every regex call
in the tree goes through Preg rather than the shim's preg_* directly, so
almost nothing saw the borrow.

PregMatchedGroups existed only to drop the null (unmatched) groups the
old PregMatches held as Option<String> values. PregMatches::get reports
a non-participating group as None on its own, so the two read alike and
the type collapses into it. Call sites still reach groups through
get(&CaptureKey::ByIndex(N)); what changes is that the value arrives as
a &str borrowed from the subject, which the signatures now carry as a
lifetime.

Three places needed the borrow reckoned with rather than a mechanical
rewrite: PhpFileCleaner::clean and Problem::get_messages read their
groups out before mutating what the match borrows, and
Git::get_authentication_failure names the lifetime of its url argument,
which the result borrows instead of self.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

11 files changed, 110 insertions, 81 deletions

diff --git a/crates/shirabe/src/util/composer_mirror.rs b/crates/shirabe/src/util/composer_mirror.rs
index 7344bfaf..5d9e8c31 100644
--- a/crates/shirabe/src/util/composer_mirror.rs
+++ b/crates/shirabe/src/util/composer_mirror.rs
@@ -61,14 +61,8 @@ impl ComposerMirror {
         ) {
             format!(
                 "gh-{}/{}",
-                gh_matches
-                    .get(&CaptureKey::ByIndex(1))
-                    .cloned()
-                    .unwrap_or_default(),
-                gh_matches
-                    .get(&CaptureKey::ByIndex(2))
-                    .cloned()
-                    .unwrap_or_default(),
+                gh_matches.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                gh_matches.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
             )
         } else if let Some(bb_matches) = Preg::match3(
             php_regex!(r"#^https://bitbucket\.org/([^/]+)/(.+?)(?:\.git)?/?$#"),
@@ -76,14 +70,8 @@ impl ComposerMirror {
         ) {
             format!(
                 "bb-{}/{}",
-                bb_matches
-                    .get(&CaptureKey::ByIndex(1))
-                    .cloned()
-                    .unwrap_or_default(),
-                bb_matches
-                    .get(&CaptureKey::ByIndex(2))
-                    .cloned()
-                    .unwrap_or_default(),
+                bb_matches.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                bb_matches.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
             )
         } else {
             Preg::replace(php_regex!(r"{[^a-z0-9_.-]}i"), "-", url.trim_matches('/'))
diff --git a/crates/shirabe/src/util/filesystem.rs b/crates/shirabe/src/util/filesystem.rs
index d59df55e..6302a6d2 100644
--- a/crates/shirabe/src/util/filesystem.rs
+++ b/crates/shirabe/src/util/filesystem.rs
@@ -741,8 +741,8 @@ impl Filesystem {
         ) {
             prefix = prefix_match
                 .get(&shirabe_pcre::CaptureKey::ByIndex(1))
-                .cloned()
-                .unwrap_or_default();
+                .unwrap_or_default()
+                .to_string();
             path = substr(&path, strlen(&prefix), None);
         }
 
@@ -765,11 +765,11 @@ impl Filesystem {
         // ensure c: is normalized to C:
         prefix = Preg::replace_callback(
             php_regex!("{(^|://)[a-z]:$}i"),
-            |m: &shirabe_pcre::PregMatchedGroups| -> String {
+            |m: &shirabe_pcre::PregMatches| -> String {
                 let s = m
                     .get(&shirabe_pcre::CaptureKey::ByIndex(0))
-                    .cloned()
-                    .unwrap_or_default();
+                    .unwrap_or_default()
+                    .to_string();
                 strtoupper(&s)
             },
             &prefix,
diff --git a/crates/shirabe/src/util/forgejo_url.rs b/crates/shirabe/src/util/forgejo_url.rs
index e5304f5d..7f9333e0 100644
--- a/crates/shirabe/src/util/forgejo_url.rs
+++ b/crates/shirabe/src/util/forgejo_url.rs
@@ -43,8 +43,8 @@ impl ForgejoUrl {
             .map(|i| {
                 matches
                     .get(&CaptureKey::ByIndex(i))
-                    .cloned()
                     .unwrap_or_default()
+                    .to_string()
             })
             .collect();
 
diff --git a/crates/shirabe/src/util/git.rs b/crates/shirabe/src/util/git.rs
index 2b012d4c..e7239987 100644
--- a/crates/shirabe/src/util/git.rs
+++ b/crates/shirabe/src/util/git.rs
@@ -14,7 +14,7 @@ use crate::util::ProcessExecutor;
 use crate::util::Url;
 use crate::util::{AuthHelper, StoreAuth};
 use indexmap::IndexMap;
-use shirabe_pcre::{CaptureKey, Preg, PregMatchedGroups};
+use shirabe_pcre::{CaptureKey, Preg, PregMatches};
 use shirabe_php_shim::{
     AnyThrowable, CmpOp, InvalidArgumentException, PHP_EOL, PhpMixed, RuntimeException, array_map,
     clearstatcache, explode, implode, in_array_loose, in_array_strict, is_dir, php_regex,
@@ -230,13 +230,16 @@ impl Git {
                 php_regex!(r"{^(?:composer|origin)\s+https?://(.+):(.+)@([^/]+)}im"),
                 &output,
             ) {
-                let m3 = m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default();
+                let m3 = m
+                    .get(&CaptureKey::ByIndex(3))
+                    .unwrap_or_default()
+                    .to_string();
                 if !self.io.has_authentication(&m3) {
                     self.io.borrow_mut().set_authentication(
                         m3,
-                        rawurldecode(&m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default()),
+                        rawurldecode(m.get(&CaptureKey::ByIndex(1)).unwrap_or_default()),
                         Some(rawurldecode(
-                            &m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
+                            m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
                         )),
                     );
                 }
@@ -262,8 +265,14 @@ impl Git {
                 _ => vec![],
             };
             for protocol in &protocols_list {
-                let m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
-                let m2 = m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default();
+                let m1 = m
+                    .get(&CaptureKey::ByIndex(1))
+                    .unwrap_or_default()
+                    .to_string();
+                let m2 = m
+                    .get(&CaptureKey::ByIndex(2))
+                    .unwrap_or_default()
+                    .to_string();
                 let proto_url = if protocol == "ssh" {
                     format!("git@{}:{}", m1, m2)
                 } else {
@@ -291,7 +300,10 @@ impl Git {
             }
 
             // failed to checkout, first check git accessibility
-            let m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
+            let m1 = m
+                .get(&CaptureKey::ByIndex(1))
+                .unwrap_or_default()
+                .to_string();
             if !self.io.has_authentication(&m1) && !self.io.is_interactive() {
                 self.throw_exception(
                     &format!(
@@ -357,8 +369,14 @@ impl Git {
                 )
             });
             if let Some(m) = github_matched {
-                let m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
-                let m2 = m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default();
+                let m1 = m
+                    .get(&CaptureKey::ByIndex(1))
+                    .unwrap_or_default()
+                    .to_string();
+                let m2 = m
+                    .get(&CaptureKey::ByIndex(2))
+                    .unwrap_or_default()
+                    .to_string();
                 if !self.io.has_authentication(&m1) {
                     let mut git_hub_util = GitHub::new(
                         self.io.clone(),
@@ -421,9 +439,14 @@ impl Git {
                     None,
                 )?;
 
-                let domain = m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default();
-                let mut repo_with_git_part =
-                    m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default();
+                let domain = m
+                    .get(&CaptureKey::ByIndex(2))
+                    .unwrap_or_default()
+                    .to_string();
+                let mut repo_with_git_part = m
+                    .get(&CaptureKey::ByIndex(3))
+                    .unwrap_or_default()
+                    .to_string();
                 if !repo_with_git_part.ends_with(".git") {
                     repo_with_git_part.push_str(".git");
                 }
@@ -565,9 +588,18 @@ impl Git {
                     url,
                 )
             }) {
-                let mut m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
-                let m2 = m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default();
-                let m3 = m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default();
+                let mut m1 = m
+                    .get(&CaptureKey::ByIndex(1))
+                    .unwrap_or_default()
+                    .to_string();
+                let m2 = m
+                    .get(&CaptureKey::ByIndex(2))
+                    .unwrap_or_default()
+                    .to_string();
+                let m3 = m
+                    .get(&CaptureKey::ByIndex(3))
+                    .unwrap_or_default()
+                    .to_string();
                 if m1 == "git" {
                     m1 = "https".to_string();
                 }
@@ -641,9 +673,18 @@ impl Git {
                 }
             } else if let Some(m) = self.get_authentication_failure(url) {
                 // private non-github/gitlab/bitbucket repo that failed to authenticate
-                let m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
-                let mut m2 = m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default();
-                let m3 = m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default();
+                let m1 = m
+                    .get(&CaptureKey::ByIndex(1))
+                    .unwrap_or_default()
+                    .to_string();
+                let mut m2 = m
+                    .get(&CaptureKey::ByIndex(2))
+                    .unwrap_or_default()
+                    .to_string();
+                let m3 = m
+                    .get(&CaptureKey::ByIndex(3))
+                    .unwrap_or_default()
+                    .to_string();
                 let mut auth_parts: Option<String> = None;
                 if m2.contains("@") {
                     let parts = explode("@", &m2);
@@ -1083,7 +1124,7 @@ impl Git {
         Ok(false)
     }
 
-    fn get_authentication_failure(&self, url: &str) -> Option<PregMatchedGroups> {
+    fn get_authentication_failure<'u>(&self, url: &'u str) -> Option<PregMatches<'u>> {
         let m = Preg::is_match3(php_regex!(r"{^(https?://)([^/]+)(.*)$}i"), url)?;
 
         let auth_failures = [
@@ -1172,8 +1213,8 @@ impl Git {
                     return Ok(Some(
                         matches
                             .get(&CaptureKey::ByIndex(1))
-                            .cloned()
-                            .unwrap_or_default(),
+                            .unwrap_or_default()
+                            .to_string(),
                     ));
                 }
             }
@@ -1295,7 +1336,7 @@ impl Git {
                 && let Some(matches) =
                     Preg::is_match3(php_regex!(r"/^git version (\d+(?:\.\d+)+)/m"), &output)
             {
-                *version = Some(matches.get(&CaptureKey::ByIndex(1)).cloned());
+                *version = Some(matches.get(&CaptureKey::ByIndex(1)).map(str::to_string));
             }
         }
         version.clone().unwrap_or(None)
diff --git a/crates/shirabe/src/util/github.rs b/crates/shirabe/src/util/github.rs
index 9c22a2b3..28a13738 100644
--- a/crates/shirabe/src/util/github.rs
+++ b/crates/shirabe/src/util/github.rs
@@ -326,7 +326,9 @@ impl GitHub {
                 continue;
             }
             if let Some(caps) = Preg::match3(php_regex!(r"{\burl=(?P<url>[^\s;]+)}"), header) {
-                return caps.get(&CaptureKey::ByName("url".to_string())).cloned();
+                return caps
+                    .get(&CaptureKey::ByName("url".to_string()))
+                    .map(str::to_string);
             }
         }
 
diff --git a/crates/shirabe/src/util/http/response.rs b/crates/shirabe/src/util/http/response.rs
index ea961768..7c5f5248 100644
--- a/crates/shirabe/src/util/http/response.rs
+++ b/crates/shirabe/src/util/http/response.rs
@@ -68,7 +68,7 @@ impl Response {
             if let Some(matches) = Preg::match3(&pattern, header)
                 && let Some(s) = matches.get(&shirabe_pcre::CaptureKey::ByIndex(1))
             {
-                value = Some(s.clone());
+                value = Some(s.to_string());
             }
         }
         value
diff --git a/crates/shirabe/src/util/http_downloader.rs b/crates/shirabe/src/util/http_downloader.rs
index b06d7004..c6e70117 100644
--- a/crates/shirabe/src/util/http_downloader.rs
+++ b/crates/shirabe/src/util/http_downloader.rs
@@ -246,14 +246,14 @@ impl HttpDownloader {
                 origin.clone(),
                 rawurldecode(
                     m.get(&CaptureKey::ByIndex(1))
-                        .cloned()
                         .unwrap_or_default()
+                        .to_string()
                         .as_str(),
                 ),
                 Some(rawurldecode(
                     m.get(&CaptureKey::ByIndex(2))
-                        .cloned()
                         .unwrap_or_default()
+                        .to_string()
                         .as_str(),
                 )),
             );
diff --git a/crates/shirabe/src/util/platform.rs b/crates/shirabe/src/util/platform.rs
index 629854b6..f2b8aea3 100644
--- a/crates/shirabe/src/util/platform.rs
+++ b/crates/shirabe/src/util/platform.rs
@@ -2,7 +2,7 @@
 
 use crate::util::ProcessExecutor;
 use crate::util::Silencer;
-use shirabe_pcre::{Preg, PregMatchedGroups};
+use shirabe_pcre::{Preg, PregMatches};
 use shirabe_php_shim::{
     PHP_ENV, PHP_SERVER, PhpMixed, PhpResource, RuntimeException, defined, file_exists,
     file_get_contents, fstat, function_exists, getcwd, getenv, ini_get, is_readable, mb_strlen,
@@ -98,15 +98,13 @@ impl Platform {
         // two forms are written as an explicit alternation: `$VAR` or `%VAR%`.
         Preg::replace_callback(
             php_regex!(r"#^(?:\$(?P<dvar>\w+)|%(?P<pvar>\w+)%)(?P<path>.*)#"),
-            |matches: &PregMatchedGroups| -> String {
+            |matches: &PregMatches| -> String {
                 let var = matches
                     .get(&CaptureKey::ByName("dvar".to_string()))
                     .or_else(|| matches.get(&CaptureKey::ByName("pvar".to_string())))
-                    .map(|s| s.as_str())
                     .unwrap_or("");
                 let path_part = matches
                     .get(&CaptureKey::ByName("path".to_string()))
-                    .map(|s| s.as_str())
                     .unwrap_or("");
                 // Treat HOME as an alias for USERPROFILE on Windows for legacy reasons
                 if Platform::is_windows() && var == "HOME" {
diff --git a/crates/shirabe/src/util/process_executor.rs b/crates/shirabe/src/util/process_executor.rs
index 48fa415f..2954ed7a 100644
--- a/crates/shirabe/src/util/process_executor.rs
+++ b/crates/shirabe/src/util/process_executor.rs
@@ -7,7 +7,7 @@ use crate::signal::SignalSubscription;
 use crate::util::GitHub;
 use crate::util::Platform;
 use indexmap::IndexMap;
-use shirabe_pcre::{CaptureKey, Preg, PregMatchedGroups};
+use shirabe_pcre::{CaptureKey, Preg, PregMatches};
 use shirabe_php_shim::Catch as _;
 use shirabe_php_shim::{
     LogicException, PHP_EOL, PhpMixed, RuntimeException, array_intersect, array_map,
@@ -219,7 +219,10 @@ impl ProcessExecutor {
             if Platform::is_windows()
                 && let Some(m) = Preg::is_match3(php_regex!(r"{^([^:/\\]++) }"), &command_str)
             {
-                let m1 = m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default();
+                let m1 = m
+                    .get(&CaptureKey::ByIndex(1))
+                    .unwrap_or_default()
+                    .to_string();
                 command_str = substr_replace(
                     &command_str,
                     &Self::escape(&Self::get_executable(&m1)),
@@ -831,23 +834,20 @@ impl ProcessExecutor {
         };
         let safe_command = Preg::replace_callback(
             php_regex!(r"{://(?P<user>[^:/\s]+):(?P<password>[^@\s/]+)@}i"),
-            |m: &PregMatchedGroups| -> String {
+            |m: &PregMatches| -> String {
                 let user_key = CaptureKey::ByName("user".to_string());
                 // if the username looks like a long (12char+) hex string, or a modern github token (e.g. ghp_xxx, github_pat_xxx) we obfuscate that
                 if Preg::is_match(
                     GitHub::GITHUB_TOKEN_REGEX,
-                    m.get(&user_key).cloned().unwrap_or_default().as_str(),
+                    m.get(&user_key).unwrap_or_default(),
                 ) {
                     return "://***:***@".to_string();
                 }
-                if Preg::is_match(
-                    r"{^[a-f0-9]{12,}$}",
-                    m.get(&user_key).cloned().unwrap_or_default().as_str(),
-                ) {
+                if Preg::is_match(r"{^[a-f0-9]{12,}$}", m.get(&user_key).unwrap_or_default()) {
                     return "://***:***@".to_string();
                 }
 
-                format!("://{}:***@", m.get(&user_key).cloned().unwrap_or_default())
+                format!("://{}:***@", m.get(&user_key).unwrap_or_default())
             },
             &command_string,
         );
diff --git a/crates/shirabe/src/util/svn.rs b/crates/shirabe/src/util/svn.rs
index 8018efa5..72ca2321 100644
--- a/crates/shirabe/src/util/svn.rs
+++ b/crates/shirabe/src/util/svn.rs
@@ -410,8 +410,8 @@ impl Svn {
                 *cached = Some(
                     matches
                         .get(&CaptureKey::ByIndex(1))
-                        .cloned()
-                        .unwrap_or_default(),
+                        .unwrap_or_default()
+                        .to_string(),
                 );
             }
         }
diff --git a/crates/shirabe/src/util/url.rs b/crates/shirabe/src/util/url.rs
index 0a3722eb..8b751b2a 100644
--- a/crates/shirabe/src/util/url.rs
+++ b/crates/shirabe/src/util/url.rs
@@ -22,9 +22,9 @@ impl Url {
             ) {
                 url = format!(
                     "https://api.github.com/repos/{}/{}/{}ball/{}",
-                    m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(3)).unwrap_or_default(),
                     r#ref
                 );
             } else if let Some(m) = Preg::match3(
@@ -35,9 +35,9 @@ impl Url {
             ) {
                 url = format!(
                     "https://api.github.com/repos/{}/{}/{}ball/{}",
-                    m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(3)).unwrap_or_default(),
                     r#ref
                 );
             } else if let Some(m) = Preg::match3(
@@ -48,9 +48,9 @@ impl Url {
             ) {
                 url = format!(
                     "https://api.github.com/repos/{}/{}/{}ball/{}",
-                    m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(3)).cloned().unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(3)).unwrap_or_default(),
                     r#ref
                 );
             }
@@ -63,10 +63,10 @@ impl Url {
             ) {
                 url = format!(
                     "https://bitbucket.org/{}/{}/get/{}.{}",
-                    m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
                     r#ref,
-                    m.get(&CaptureKey::ByIndex(4)).cloned().unwrap_or_default()
+                    m.get(&CaptureKey::ByIndex(4)).unwrap_or_default()
                 );
             }
         } else if host == "gitlab.com" || host == "www.gitlab.com" {
@@ -78,8 +78,8 @@ impl Url {
             ) {
                 url = format!(
                     "https://gitlab.com/api/v4/projects/{}/repository/archive.{}?sha={}",
-                    m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(),
-                    m.get(&CaptureKey::ByIndex(2)).cloned().unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(1)).unwrap_or_default(),
+                    m.get(&CaptureKey::ByIndex(2)).unwrap_or_default(),
                     r#ref
                 );
             }
@@ -165,12 +165,12 @@ impl Url {
             |m| {
                 let user = m
                     .get(&CaptureKey::ByName("user".to_string()))
-                    .cloned()
-                    .unwrap_or_default();
+                    .unwrap_or_default()
+                    .to_string();
                 let prefix = m
                     .get(&CaptureKey::ByName("prefix".to_string()))
-                    .cloned()
-                    .unwrap_or_default();
+                    .unwrap_or_default()
+                    .to_string();
                 // if the username looks like a long (12char+) hex string, or a modern github token (e.g. ghp_xxx, github_pat_xxx) we obfuscate that
                 if Preg::is_match(GitHub::GITHUB_TOKEN_REGEX, &user) {
                     format!("{}***:***@", prefix)