From 91692846909ed191addb7ec1c34aad11392ab88b Mon Sep 17 00:00:00 2001 From: nsfisis Date: Sat, 18 Jul 2026 15:03:55 +0900 Subject: perf(regex): eliminate per-call clone overhead in preg_* dispatch regex::Regex::clone() does not share the underlying meta engine's search-cache pool, so every fresh clone pays a ~10us warmup cost on its first use. Two changes together eliminate this across nearly all preg_* call sites: - A php_regex! macro resolves PHP-style patterns to a per-call-site &'static regex::Regex (via regex-macro's LazyLock), applied at the majority of call sites throughout the codebase. - Call sites still passing dynamic pattern strings go through PATTERN_CACHE, which now stores Arc<(Regex, bool)> and hands out Arc::clone()s instead of cloning the Regex itself. PregPattern::resolve() returns a ResolvedPattern enum (Arc or 'static reference) rather than an owned Regex, so neither path ever clones the Regex proper. Co-Authored-By: Claude Sonnet 5 --- crates/shirabe/src/json/json_file.rs | 18 ++++++---- crates/shirabe/src/json/json_manipulator.rs | 54 +++++++++++++++++------------ 2 files changed, 44 insertions(+), 28 deletions(-) (limited to 'crates/shirabe/src/json') diff --git a/crates/shirabe/src/json/json_file.rs b/crates/shirabe/src/json/json_file.rs index 075e1f64..0ce55e55 100644 --- a/crates/shirabe/src/json/json_file.rs +++ b/crates/shirabe/src/json/json_file.rs @@ -14,8 +14,8 @@ use shirabe_external_packages::seld::json_lint::{ParsingException, ParsingExcept use shirabe_php_shim::{ InvalidArgumentException, JSON_PRETTY_PRINT, JSON_UNESCAPED_SLASHES, JSON_UNESCAPED_UNICODE, PhpMixed, RuntimeException, UnexpectedValueException, dirname, file_exists, file_get_contents, - file_put_contents, is_dir, is_file, json_decode, json_encode_ex, mkdir, realpath, str_contains, - str_ends_with, str_repeat, strlen, strpos, usleep, + file_put_contents, is_dir, is_file, json_decode, json_encode_ex, mkdir, php_regex, realpath, + str_contains, str_ends_with, str_repeat, strlen, strpos, usleep, }; #[derive(Debug, Clone)] @@ -113,7 +113,7 @@ impl JsonFile { http_downloader: Option>>, io: Option>>, ) -> anyhow::Result { - if http_downloader.is_none() && Preg::is_match(r"{^https?://}i", &path) { + if http_downloader.is_none() && Preg::is_match(php_regex!(r"{^https?://}i"), &path) { return Err(InvalidArgumentException { message: "http urls require a HttpDownloader instance to be passed".to_string(), code: 0, @@ -466,7 +466,7 @@ impl JsonFile { // Pretty printing and not using default indentation let indent_owned = options.indent.clone(); return Preg::replace_callback( - r"#^ {4,}#m", + php_regex!(r"#^ {4,}#m"), move |m: &indexmap::IndexMap< shirabe_external_packages::composer::pcre::CaptureKey, String, @@ -510,7 +510,9 @@ impl JsonFile { { let mut count: usize = 0; let replaced = Preg::replace5( - r#"{\r?\n<<<<<<< [^\r\n]+\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n(?:\|{7} [^\r\n]+\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n)?=======\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n>>>>>>> [^\r\n]+(\r?\n)}"#, + php_regex!( + r#"{\r?\n<<<<<<< [^\r\n]+\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n(?:\|{7} [^\r\n]+\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n)?=======\r?\n\s+"content-hash": *"[0-9a-f]+", *\r?\n>>>>>>> [^\r\n]+(\r?\n)}"# + ), " \"content-hash\": \"VCS merge conflict detected. Please run `composer update --lock`.\",$1", json, -1, @@ -574,7 +576,11 @@ impl JsonFile { pub fn detect_indenting(json: Option<&str>) -> String { let mut m: IndexMap = IndexMap::new(); - if Preg::is_match3(r##"#^([ \t]+)"#m"##, json.unwrap_or(""), Some(&mut m)) { + if Preg::is_match3( + php_regex!(r##"#^([ \t]+)"#m"##), + json.unwrap_or(""), + Some(&mut m), + ) { return m.get(&CaptureKey::ByIndex(1)).cloned().unwrap_or_default(); } diff --git a/crates/shirabe/src/json/json_manipulator.rs b/crates/shirabe/src/json/json_manipulator.rs index b11178ff..e501940c 100644 --- a/crates/shirabe/src/json/json_manipulator.rs +++ b/crates/shirabe/src/json/json_manipulator.rs @@ -8,8 +8,8 @@ use shirabe_external_packages::composer::pcre::{CaptureKey, Preg}; use shirabe_php_shim::{ InvalidArgumentException, LogicException, PhpMixed, addcslashes, array_key_exists, array_keys, array_reverse, empty, explode, implode, in_array, is_array, is_int, is_numeric, json_decode, - php_truthy, preg_quote, rtrim, str_contains, str_repeat, str_replace, strlen, strnatcmp, - strpos, substr, trim, uksort, + php_regex, php_truthy, preg_quote, rtrim, str_contains, str_repeat, str_replace, strlen, + strnatcmp, strpos, substr, trim, uksort, }; #[derive(Debug)] @@ -35,7 +35,7 @@ impl JsonManipulator { if contents.is_empty() { contents = "{}".to_string(); } - if !Preg::is_match3("#^\\{(.*)\\}$#s", &contents, None) { + if !Preg::is_match3(php_regex!("#^\\{(.*)\\}$#s"), &contents, None) { return Err(InvalidArgumentException { message: "The json file must be an object ({})".to_string(), code: 0, @@ -115,7 +115,7 @@ impl JsonManipulator { } else { let mut groups: IndexMap = IndexMap::new(); if Preg::is_match3( - "#^\\s*\\{\\s*\\S+.*?(\\s*\\}\\s*)$#s", + php_regex!("#^\\s*\\{\\s*\\S+.*?(\\s*\\}\\s*)$#s"), &links, Some(&mut groups), ) { @@ -125,7 +125,7 @@ impl JsonManipulator { .unwrap_or_default(); // link missing but non empty links links = Preg::replace( - &format!("{{{}$}}", preg_quote(&groups_1, None)), + format!("{{{}$}}", preg_quote(&groups_1, None)), // addcslashes is used to double up backslashes/$ since preg_replace resolves them as back references otherwise, see #1588 &addcslashes( &format!( @@ -175,7 +175,7 @@ impl JsonManipulator { let replacements = ["0-$0", "1-$0", "2-$0", "3-$0", "4-$0"]; let mut result = requirement.to_string(); for (p, r) in patterns.iter().zip(replacements.iter()) { - result = Preg::replace(p, r, &result); + result = Preg::replace(*p, r, &result); } result } else { @@ -738,7 +738,9 @@ impl JsonManipulator { } else { let mut leading_match: IndexMap = IndexMap::new(); if Preg::is_match_named( - "#^\\{(?P\\s*?)(?P\\S+.*?)?(?P\\s*)\\}$#s", + php_regex!( + "#^\\{(?P\\s*?)(?P\\S+.*?)?(?P\\s*)\\}$#s" + ), &children, &mut leading_match, ) { @@ -762,7 +764,7 @@ impl JsonManipulator { // child missing but non empty children if append { children = Preg::replace( - &format!("#{}}}$#", whitespace), + format!("#{}}}$#", whitespace), &addcslashes( &format!( ",{}{}{}{}: {}{}}}", @@ -780,7 +782,7 @@ impl JsonManipulator { } else { whitespace = leading_space.clone(); children = Preg::replace( - &format!("#^{{{}#", whitespace), + format!("#^{{{}#", whitespace), &addcslashes( &format!( "{{{}{}: {},{}{}{}", @@ -894,7 +896,7 @@ impl JsonManipulator { // try and find a match for the subkey let key_regex = str_replace("/", "\\\\?/", &preg_quote(&name_owned, None)); let mut children_clean: Option = None; - if Preg::is_match3(&format!("{{\"{}\"\\s*:}}i", key_regex), &children, None) { + if Preg::is_match3(format!("{{\"{}\"\\s*:}}i", key_regex), &children, None) { // find best match for the value of "name". The PHP pattern `"name"\s*:\s*(?&json)` is // not anchored, so it can match the key at several nesting levels; collect every such // occurrence and keep the longest, reproducing PHP's behaviour. @@ -908,7 +910,7 @@ impl JsonManipulator { } let mut count_out: usize = 0; let cleaned = Preg::replace5( - &format!("{{,\\s*{}}}i", preg_quote(&best_match, None)), + format!("{{,\\s*{}}}i", preg_quote(&best_match, None)), "", &children, -1, @@ -916,7 +918,7 @@ impl JsonManipulator { ); if 1 != count_out { let cleaned2 = Preg::replace5( - &format!("{{{}\\s*,?\\s*}}i", preg_quote(&best_match, None)), + format!("{{{}\\s*,?\\s*}}i", preg_quote(&best_match, None)), "", &cleaned, -1, @@ -942,7 +944,7 @@ impl JsonManipulator { // no child data left, $name was the only key in let mut empty_match: IndexMap = IndexMap::new(); if Preg::is_match_named( - "#^\\{\\s*?(?P\\S+.*?)?(?P\\s*)\\}$#s", + php_regex!("#^\\{\\s*?(?P\\S+.*?)?(?P\\s*)\\}$#s"), &children_clean, &mut empty_match, ) && empty_match.get("content").is_none() @@ -1039,7 +1041,9 @@ impl JsonManipulator { let mut leading_match: IndexMap = IndexMap::new(); if Preg::is_match_named( - "#^\\[(?P\\s*?)(?P\\S+.*?)?(?P\\s*)\\]$#s", + php_regex!( + "#^\\[(?P\\s*?)(?P\\S+.*?)?(?P\\s*)\\]$#s" + ), &children, &mut leading_match, ) { @@ -1067,7 +1071,7 @@ impl JsonManipulator { // child missing but non empty children if append { children = Preg::replace( - &format!("#{}\\]$#", whitespace), + format!("#{}\\]$#", whitespace), &addcslashes( &format!( ",{}{}{}]", @@ -1082,7 +1086,7 @@ impl JsonManipulator { } else { whitespace = leading_whitespace.clone(); children = Preg::replace( - &format!("#^\\[{}#", whitespace), + format!("#^\\[{}#", whitespace), &addcslashes( &format!( "[{}{},{}", @@ -1330,13 +1334,17 @@ impl JsonManipulator { // append at the end of the file and keep whitespace let mut tail_match: IndexMap = IndexMap::new(); - if Preg::is_match3("#[^{\\s](\\s*)\\}$#", &self.contents, Some(&mut tail_match)) { + if Preg::is_match3( + php_regex!("#[^{\\s](\\s*)\\}$#"), + &self.contents, + Some(&mut tail_match), + ) { let tail_match_1 = tail_match .get(&CaptureKey::ByIndex(1)) .cloned() .unwrap_or_default(); self.contents = Preg::replace( - &format!("#{}\\}}$#", tail_match_1), + format!("#{}\\}}$#", tail_match_1), &addcslashes( &format!( ",{}{}{}: {}{}}}", @@ -1356,7 +1364,7 @@ impl JsonManipulator { // append at the end of the file self.contents = Preg::replace( - "#\\}$#", + php_regex!("#\\}$#"), &addcslashes( &format!( "{}{}: {}{}}}", @@ -1406,15 +1414,17 @@ impl JsonManipulator { // check that we are not leaving a dangling comma on the previous line if the last line was removed let mut start = self.contents[..m.key_pos].to_string(); let end = self.contents[e..].to_string(); - if Preg::is_match3("#,\\s*$#", &start, None) && Preg::is_match3("#^\\}$#", &end, None) { + if Preg::is_match3(php_regex!("#,\\s*$#"), &start, None) + && Preg::is_match3(php_regex!("#^\\}$#"), &end, None) + { start = rtrim( - &Preg::replace("#,(\\s*)$#", "$1", &start), + &Preg::replace(php_regex!("#,(\\s*)$#"), "$1", &start), Some(&self.indent), ); } self.contents = format!("{}{}", start, end); - if Preg::is_match3("#^\\{\\s*\\}\\s*$#", &self.contents, None) { + if Preg::is_match3(php_regex!("#^\\{\\s*\\}\\s*$#"), &self.contents, None) { self.contents = "{\n}".to_string(); } -- cgit v1.3.1