globset: add opt-in Arbitrary trait implementations

This feature is mandatory when using `Glob` in fuzz testing.

Closes #2720

.github/workflows/ci.yml

@@ -230,3 +230,28 @@ jobs:
         env:
           RUSTDOCFLAGS: -D warnings
         run: cargo doc --no-deps --document-private-items --workspace
+
+  fuzz_testing:
+    name: Compile Fuzz Test Targets
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Install required packages (Ubuntu)
+        run: |
+          sudo apt-get update
+          sudo apt-get install g++ --yes
+
+      - name: Install Rust
+        uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: stable
+
+      - name: Install Fuzzer
+        run: cargo install cargo-fuzz
+        working-directory: fuzz
+
+      - name: Verify fuzz targets build
+        run: cargo check
+        working-directory: fuzz

Cargo.lock

@@ -17,6 +17,15 @@ version = "1.0.98"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e16d2d3311acee920a9eb8d33b8cbc1787ce4a264e85f964c2404b969bdcd487"
 
+[[package]]
+name = "arbitrary"
+version = "1.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "dde20b3d026af13f561bdd0f15edf01fc734f0dafcedbaf42bba506a9517f223"
+dependencies = [
+ "derive_arbitrary",
+]
+
 [[package]]
 name = "bitflags"
 version = "2.9.1"
@@ -85,6 +94,17 @@ version = "0.8.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
 
+[[package]]
+name = "derive_arbitrary"
+version = "1.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "30542c1ad912e0e3d22a1935c290e12e8a29d704a420177a31faad4a601a0800"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
 [[package]]
 name = "encoding_rs"
 version = "0.8.35"
@@ -126,6 +146,7 @@ name = "globset"
 version = "0.4.16"
 dependencies = [
  "aho-corasick",
+ "arbitrary",
  "bstr",
  "glob",
  "log",

Cargo.toml

@@ -20,6 +20,7 @@ exclude = [
   "/pkg/brew",
   "/benchsuite/",
   "/scripts/",
+  "/crates/fuzz",
 ]
 build = "build.rs"
 autotests = false

ci/ubuntu-install-packages

@@ -11,4 +11,4 @@ if ! command -V sudo; then
 fi
 sudo apt-get update
 sudo apt-get install -y --no-install-recommends \
-  zsh xz-utils liblz4-tool musl-tools brotli zstd
+  zsh xz-utils liblz4-tool musl-tools brotli zstd g++

crates/globset/Cargo.toml

@@ -21,6 +21,7 @@ bench = false
 
 [dependencies]
 aho-corasick = "1.1.1"
+arbitrary = { version = "1.3.2", optional = true, features = ["derive"] }
 bstr = { version = "1.6.2", default-features = false, features = ["std"] }
 log = { version = "0.4.20", optional = true }
 serde = { version = "1.0.188", optional = true }
@@ -41,6 +42,7 @@ serde_json = "1.0.107"
 
 [features]
 default = ["log"]
+arbitrary = ["dep:arbitrary"]
 # DEPRECATED. It is a no-op. SIMD is done automatically through runtime
 # dispatch.
 simd-accel = []

crates/globset/src/glob.rs

@@ -72,6 +72,7 @@ impl MatchStrategy {
 /// It cannot be used directly to match file paths, but it can be converted
 /// to a regular expression string or a matcher.
 #[derive(Clone, Debug, Eq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 pub struct Glob {
     glob: String,
     re: String,
@@ -194,6 +195,7 @@ pub struct GlobBuilder<'a> {
 }
 
 #[derive(Clone, Copy, Debug, Eq, Hash, PartialEq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 struct GlobOptions {
     /// Whether to match case insensitively.
     case_insensitive: bool,
@@ -220,6 +222,7 @@ impl GlobOptions {
 }
 
 #[derive(Clone, Debug, Default, Eq, PartialEq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 struct Tokens(Vec<Token>);
 
 impl std::ops::Deref for Tokens {
@@ -236,6 +239,7 @@ impl std::ops::DerefMut for Tokens {
 }
 
 #[derive(Clone, Debug, Eq, PartialEq)]
+#[cfg_attr(feature = "arbitrary", derive(arbitrary::Arbitrary))]
 enum Token {
     Literal(char),
     Any,

crates/globset/src/lib.rs

@@ -94,6 +94,19 @@ Standard Unix-style glob syntax is supported:
 
 A `GlobBuilder` can be used to prevent wildcards from matching path separators,
 or to enable case insensitive matching.
+
+# Crate Features
+
+This crate includes optional features that can be enabled if necessary.
+These features are not required but may be useful depending on the use case.
+
+The following features are available:
+
+* **arbitrary** -
+  Enabling this feature introduces a public dependency on the
+  [`arbitrary`](https://crates.io/crates/arbitrary)
+  crate. Namely, it implements the `Arbitrary` trait from that crate for the
+  [`Glob`] type. This feature is disabled by default.
 */
 
 #![deny(missing_docs)]

fuzz/.gitignore

@@ -0,0 +1,4 @@
+target
+corpus
+artifacts
+coverage

fuzz/Cargo.lock

@@ -0,0 +1,188 @@
+# This file is automatically @generated by Cargo.
+# It is not intended for manual editing.
+version = 4
+
+[[package]]
+name = "aho-corasick"
+version = "1.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b2969dcb958b36655471fc61f7e416fa76033bdd4bfed0678d8fee1e2d07a1f0"
+dependencies = [
+ "memchr",
+]
+
+[[package]]
+name = "arbitrary"
+version = "1.3.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7d5a26814d8dcb93b0e5a0ff3c6d80a8843bafb21b39e8e18a6f05471870e110"
+dependencies = [
+ "derive_arbitrary",
+]
+
+[[package]]
+name = "bstr"
+version = "1.9.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c48f0051a4b4c5e0b6d365cd04af53aeaa209e3cc15ec2cdb69e73cc87fbd0dc"
+dependencies = [
+ "memchr",
+ "serde",
+]
+
+[[package]]
+name = "cc"
+version = "1.0.83"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f1174fb0b6ec23863f8b971027804a42614e347eafb0a95bf0b12cdae21fc4d0"
+dependencies = [
+ "jobserver",
+ "libc",
+]
+
+[[package]]
+name = "derive_arbitrary"
+version = "1.3.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "67e77553c4162a157adbf834ebae5b415acbecbeafc7a74b0e886657506a7611"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "fuzz"
+version = "0.0.1"
+dependencies = [
+ "globset",
+ "libfuzzer-sys",
+]
+
+[[package]]
+name = "globset"
+version = "0.4.16"
+dependencies = [
+ "aho-corasick",
+ "arbitrary",
+ "bstr",
+ "log",
+ "regex-automata",
+ "regex-syntax",
+]
+
+[[package]]
+name = "jobserver"
+version = "0.1.27"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8c37f63953c4c63420ed5fd3d6d398c719489b9f872b9fa683262f8edd363c7d"
+dependencies = [
+ "libc",
+]
+
+[[package]]
+name = "libc"
+version = "0.2.152"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "13e3bf6590cbc649f4d1a3eefc9d5d6eb746f5200ffb04e5e142700b8faa56e7"
+
+[[package]]
+name = "libfuzzer-sys"
+version = "0.4.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a96cfd5557eb82f2b83fed4955246c988d331975a002961b07c81584d107e7f7"
+dependencies = [
+ "arbitrary",
+ "cc",
+ "once_cell",
+]
+
+[[package]]
+name = "log"
+version = "0.4.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b5e6163cb8c49088c2c36f57875e58ccd8c87c7427f7fbd50ea6710b2f3f2e8f"
+
+[[package]]
+name = "memchr"
+version = "2.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "523dc4f511e55ab87b694dc30d0f820d60906ef06413f93d4d7a1385599cc149"
+
+[[package]]
+name = "once_cell"
+version = "1.19.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3fdb12b2476b595f9358c5161aa467c2438859caa136dec86c26fdd2efe17b92"
+
+[[package]]
+name = "proc-macro2"
+version = "1.0.78"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e2422ad645d89c99f8f3e6b88a9fdeca7fabeac836b1002371c4367c8f984aae"
+dependencies = [
+ "unicode-ident",
+]
+
+[[package]]
+name = "quote"
+version = "1.0.35"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "291ec9ab5efd934aaf503a6466c5d5251535d108ee747472c3977cc5acc868ef"
+dependencies = [
+ "proc-macro2",
+]
+
+[[package]]
+name = "regex-automata"
+version = "0.4.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3b7fa1134405e2ec9353fd416b17f8dacd46c473d7d3fd1cf202706a14eb792a"
+dependencies = [
+ "aho-corasick",
+ "memchr",
+ "regex-syntax",
+]
+
+[[package]]
+name = "regex-syntax"
+version = "0.8.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c08c74e62047bb2de4ff487b251e4a92e24f48745648451635cec7d591162d9f"
+
+[[package]]
+name = "serde"
+version = "1.0.195"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "63261df402c67811e9ac6def069e4786148c4563f4b50fd4bf30aa370d626b02"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.195"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "46fe8f8603d81ba86327b23a2e9cdf49e1255fb94a4c5f297f6ee0547178ea2c"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "syn"
+version = "2.0.48"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0f3531638e407dfc0814761abb7c00a5b54992b849452a0646b7f65c9f770f3f"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "unicode-ident",
+]
+
+[[package]]
+name = "unicode-ident"
+version = "1.0.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3354b9ac3fae1ff6755cb6db53683adb661634f67557942dea4facebec0fee4b"

fuzz/Cargo.toml

@@ -0,0 +1,25 @@
+[package]
+name = "fuzz"
+version = "0.0.1"
+publish = false
+edition = "2021"
+
+[package.metadata]
+cargo-fuzz = true
+
+[dependencies]
+libfuzzer-sys = "0.4"
+globset = { path = "../crates/globset", features = ["arbitrary"] }
+
+# Prevent this from interfering with workspaces
+[workspace]
+members = ["."]
+
+[profile.release]
+debug = 1
+
+[[bin]]
+name = "fuzz_glob"
+path = "fuzz_targets/fuzz_glob.rs"
+test = false
+doc = false

fuzz/README.md

@@ -0,0 +1,52 @@
+# Fuzz Testing
+
+## Introduction
+
+Fuzz testing produces pseudo-random / arbitrary data that is used to find
+stability issues within a code base. While Rust provides a strong type system,
+this does not guarantee that an object will convert properly from one struct
+to another. It is the responsibility of the developer to ensure that a struct
+is converted properly. Fuzz testing will generate input within the domain of
+each property. This arbitrary data can then be used to convert from ObjectA
+to ObjectB and then back. This type of testing will help catch bugs that the
+type system is not able to see.
+
+## Installation
+
+This crate relies on the `cargo-fuzz` component. To install this component,
+run the following from the `fuzz` directory:
+
+```bash
+cargo install cargo-fuzz
+```
+
+## Listing Targets
+
+Once installed, fuzz targets can be listed by running the following command:
+
+```bash
+cargo fuzz list
+```
+
+This command will print out a list of all targets that can be tested.
+
+## Running Fuzz Tests
+
+To run a fuzz test, the target must be specified:
+
+```bash
+cargo fuzz run <target>
+```
+
+Note that the above will run the fuzz test indefinitely. Use the
+`-max_total_time=<num seconds>` flag to specify how many seconds the test
+should run for:
+
+```bash
+cargo fuzz run <target> -- -max_total_time=5
+```
+
+The above command will run the fuzz test for five seconds. If the test
+completes without error it will show how many tests were run successfully.
+The test will abort and return a non-zero error code if it is able to produce
+an error. The arbitrary input will be displayed in the event of a failure.

fuzz/fuzz_targets/fuzz_glob.rs

@@ -0,0 +1,22 @@
+#![no_main]
+
+use std::str::FromStr;
+
+use globset::Glob;
+
+libfuzzer_sys::fuzz_target!(|glob_str: &str| {
+    let Ok(glob) = Glob::new(glob_str) else {
+        return;
+    };
+
+    let Ok(glob2) = Glob::from_str(glob_str) else {
+        return;
+    };
+
+    // Verify that a `Glob` constructed with `new` is the same as a `Glob`` constructed
+    // with `from_str`.
+    assert_eq!(glob, glob2);
+
+    // Verify that `Glob::glob` produces the same string as the original.
+    assert_eq!(glob.glob(), glob_str);
+});