ref(pattern): Replace regex based matching

.gitignore

@@ -1,6 +1,10 @@
 # Rust
 **/*.rs.bk
 target
+**/fuzz/corpus
+**/fuzz/artifacts
+**/fuzz/coverage
+**/fuzz/*.html
 *.pending-snap
 
 # Python

CHANGELOG.md

@@ -18,6 +18,7 @@
 **Internal:**
 
 - Add the dynamic sampling rate to standalone spans as a measurement so that it can be stored, queried, and used for extrapolation. ([#4063](https://github.com/getsentry/relay/pull/4063))
+- Use custom wildcard matching instead of regular expressions. ([#4073](https://github.com/getsentry/relay/pull/4073))
 - Allowlist the SentryUptimeBot user-agent. ([#4068](https://github.com/getsentry/relay/pull/4068))
 
 ## 24.9.0

Cargo.toml

@@ -1,5 +1,5 @@
 [workspace]
-members = ["relay", "relay-*", "tools/*"]
+members = ["relay", "relay-*", "relay-*/fuzz/", "tools/*"]
 default-members = ["relay"]
 resolver = "2"
 

relay-pattern/Cargo.toml

@@ -18,7 +18,6 @@ serde = ["dep:serde"]
 
 [dependencies]
 memchr = { workspace = true }
-regex-lite = { workspace = true }
 serde = { workspace = true, optional = true}
 
 [dev-dependencies]

relay-pattern/benches/benches.rs

@@ -4,10 +4,18 @@ use criterion::{criterion_group, criterion_main, BenchmarkGroup, Criterion};
 use relay_pattern::Pattern;
 
 fn bench(group: &mut BenchmarkGroup<'_, WallTime>, haystack: &str, needle: &str) {
-    group.bench_function("pattern", |b| {
+    group.bench_function("case_sensitive", |b| {
         let pattern = Pattern::new(needle).unwrap();
         b.iter(|| assert!(pattern.is_match(haystack)))
     });
+
+    group.bench_function("case_insensitive", |b| {
+        let pattern = Pattern::builder(needle)
+            .case_insensitive(true)
+            .build()
+            .unwrap();
+        b.iter(|| assert!(pattern.is_match(haystack)))
+    });
 }
 
 fn literal_match(c: &mut Criterion) {
@@ -64,12 +72,25 @@ fn wildcard_match(c: &mut Criterion) {
     group.finish();
 }
 
+fn complex_match(c: &mut Criterion) {
+    let mut group = c.benchmark_group("complex_match");
+
+    const HAYSTACK: &str = "foobarwithacrazylongprefixandanditactuallymatches";
+    const NEEDLE: &str =
+        "*b??{foo,bar,baz,with}*cr{x,y,z,?}zyl[a-z]ng{suffix,pr?*ix}*it?**?uallymatches";
+
+    bench(&mut group, HAYSTACK, NEEDLE);
+
+    group.finish();
+}
+
 criterion_group!(
     benches,
     literal_match,
     prefix_match,
     suffix_match,
     contains_match,
     wildcard_match,
+    complex_match,
 );
 criterion_main!(benches);

relay-pattern/fuzz/Cargo.toml

@@ -0,0 +1,26 @@
+[package]
+name = "relay-pattern-fuzz"
+version = "0.0.0"
+publish = false
+edition = "2021"
+
+[package.metadata]
+cargo-fuzz = true
+
+[dependencies]
+libfuzzer-sys = "0.4"
+relay-pattern = { path = ".." }
+
+[[bin]]
+name = "is_match_case_sensitive"
+path = "fuzz_targets/is_match_case_sensitive.rs"
+test = false
+doc = false
+bench = false
+
+[[bin]]
+name = "is_match_case_insensitive"
+path = "fuzz_targets/is_match_case_insensitive.rs"
+test = false
+doc = false
+bench = false

relay-pattern/fuzz/fuzz_targets/is_match_case_insensitive.rs

@@ -0,0 +1,13 @@
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+fuzz_target!(|input: (&str, &str)| {
+    let (pattern, haystack) = input;
+    if let Ok(pattern) = relay_pattern::Pattern::builder(pattern)
+        .case_insensitive(true)
+        .build()
+    {
+        std::hint::black_box(pattern.is_match(haystack));
+    }
+});

relay-pattern/fuzz/fuzz_targets/is_match_case_sensitive.rs

@@ -0,0 +1,10 @@
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+fuzz_target!(|input: (&str, &str)| {
+    let (pattern, haystack) = input;
+    if let Ok(pattern) = relay_pattern::Pattern::new(pattern) {
+        std::hint::black_box(pattern.is_match(haystack));
+    }
+});