From 608d637e4d3423db6beb68f7ac32ce6c7768e1a5 Mon Sep 17 00:00:00 2001
From: camchenry <1514176+camchenry@users.noreply.github.com>
Date: Thu, 19 Sep 2024 22:40:01 +0000
Subject: [PATCH] perf(linter): use `aho-corasick` instead of `regex` for
 string matching in `jsx-a11y/img-redundant-alt` (#5892)

hypothesis: profiling shows that Regex creation takes a decent amount of time. the `regex` crate uses `aho-corasick` internally for string matching, which is all we need in some cases. in theory, we could save time by using the lib directly and not needing the full regex syntax.
---
 Cargo.lock                                    |  1 +
 Cargo.toml                                    |  1 +
 crates/oxc_linter/Cargo.toml                  |  1 +
 .../src/rules/jsx_a11y/img_redundant_alt.rs   | 30 ++++++++++++-------
 4 files changed, 23 insertions(+), 10 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 47e6742c90dac..78fbf6d025dc8 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1627,6 +1627,7 @@ dependencies = [
 name = "oxc_linter"
 version = "0.9.6"
 dependencies = [
+ "aho-corasick",
  "bitflags 2.6.0",
  "convert_case",
  "cow-utils",
diff --git a/Cargo.toml b/Cargo.toml
index 592331a330472..787004f30f207 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -108,6 +108,7 @@ napi = "3.0.0-alpha.8"
 napi-build = "2.1.3"
 napi-derive = "3.0.0-alpha.7"
 
+aho-corasick = "1.1.3"
 allocator-api2 = "0.2.18"
 assert-unchecked = "0.1.2"
 base64 = "0.22.1"
diff --git a/crates/oxc_linter/Cargo.toml b/crates/oxc_linter/Cargo.toml
index bd53c767a4409..985d3b00fe7fb 100644
--- a/crates/oxc_linter/Cargo.toml
+++ b/crates/oxc_linter/Cargo.toml
@@ -34,6 +34,7 @@ oxc_semantic = { workspace = true }
 oxc_span = { workspace = true, features = ["schemars", "serialize"] }
 oxc_syntax = { workspace = true }
 
+aho-corasick = { workspace = true }
 bitflags = { workspace = true }
 convert_case = { workspace = true }
 cow-utils = { workspace = true }
diff --git a/crates/oxc_linter/src/rules/jsx_a11y/img_redundant_alt.rs b/crates/oxc_linter/src/rules/jsx_a11y/img_redundant_alt.rs
index d867ae65ba78a..5fa68623a2c95 100644
--- a/crates/oxc_linter/src/rules/jsx_a11y/img_redundant_alt.rs
+++ b/crates/oxc_linter/src/rules/jsx_a11y/img_redundant_alt.rs
@@ -1,3 +1,4 @@
+use aho_corasick::AhoCorasick;
 use oxc_ast::{
     ast::{JSXAttributeItem, JSXAttributeName, JSXAttributeValue, JSXExpression},
     AstKind,
@@ -5,7 +6,6 @@ use oxc_ast::{
 use oxc_diagnostics::OxcDiagnostic;
 use oxc_macros::declare_oxc_lint;
 use oxc_span::{CompactStr, Span};
-use regex::{Regex, RegexBuilder};
 use serde_json::Value;
 
 use crate::{
@@ -28,7 +28,7 @@ pub struct ImgRedundantAlt(Box<ImgRedundantAltConfig>);
 #[derive(Debug, Clone)]
 pub struct ImgRedundantAltConfig {
     types_to_validate: Vec<CompactStr>,
-    redundant_words: Regex,
+    redundant_words: AhoCorasick,
 }
 
 impl std::ops::Deref for ImgRedundantAlt {
@@ -45,21 +45,25 @@ impl Default for ImgRedundantAltConfig {
     fn default() -> Self {
         Self {
             types_to_validate: vec![CompactStr::new("img")],
-            redundant_words: Self::union(&REDUNDANT_WORDS).unwrap(),
+            redundant_words: AhoCorasick::builder()
+                .ascii_case_insensitive(true)
+                .build(REDUNDANT_WORDS)
+                .expect("Could not build AhoCorasick"),
         }
     }
 }
 impl ImgRedundantAltConfig {
-    fn new(types_to_validate: Vec<&str>, redundant_words: &[&str]) -> Result<Self, regex::Error> {
+    fn new(
+        types_to_validate: Vec<&str>,
+        redundant_words: &[&str],
+    ) -> Result<Self, aho_corasick::BuildError> {
         Ok(Self {
             types_to_validate: types_to_validate.into_iter().map(Into::into).collect(),
-            redundant_words: Self::union(redundant_words)?,
+            redundant_words: AhoCorasick::builder()
+                .ascii_case_insensitive(true)
+                .build(redundant_words)?,
         })
     }
-
-    fn union(strs: &[&str]) -> Result<Regex, regex::Error> {
-        RegexBuilder::new(&format!(r"(?i)\b({})\b", strs.join("|"))).case_insensitive(true).build()
-    }
 }
 
 declare_oxc_lint!(
@@ -191,7 +195,13 @@ impl Rule for ImgRedundantAlt {
 impl ImgRedundantAlt {
     #[inline]
     fn is_redundant_alt_text(&self, alt_text: &str) -> bool {
-        self.redundant_words.is_match(alt_text)
+        for mat in self.redundant_words.find_iter(alt_text) {
+            // check if followed by space or is whole text
+            if mat.end() == alt_text.len() || alt_text.as_bytes()[mat.end()] == b' ' {
+                return true;
+            }
+        }
+        false
     }
 }