Do not go too deep in Git history

.github/workflows/bench.yml

@@ -0,0 +1,166 @@
+name: Bench
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+env:
+  CARGO_TERM_COLOR: always
+  RUST_BACKTRACE: short
+  OPENSSL_STATIC: true
+  PKG_CONFIG_ALLOW_CROSS: true
+
+jobs:
+  # Ensure various stress tests pass on all platforms
+  stress:
+    name: Stress
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        include:
+          # Linux builds using cross
+          - os: ubuntu-latest
+            target: x86_64-unknown-linux-gnu
+            use-cross: true
+            artifact_name: yek
+            asset_name: yek-x86_64-unknown-linux-gnu.tar.gz
+          - os: ubuntu-latest
+            target: aarch64-unknown-linux-gnu
+            use-cross: true
+            artifact_name: yek
+            asset_name: yek-aarch64-unknown-linux-gnu.tar.gz
+          - os: ubuntu-latest
+            target: x86_64-unknown-linux-musl
+            use-cross: true
+            artifact_name: yek
+            asset_name: yek-x86_64-unknown-linux-musl.tar.gz
+          - os: ubuntu-latest
+            target: aarch64-unknown-linux-musl
+            use-cross: true
+            artifact_name: yek
+            asset_name: yek-aarch64-unknown-linux-musl.tar.gz
+
+          # Native macOS builds
+          - os: macos-latest
+            target: x86_64-apple-darwin
+            artifact_name: yek
+            asset_name: yek-x86_64-apple-darwin.tar.gz
+          - os: macos-latest
+            target: aarch64-apple-darwin
+            artifact_name: yek
+            asset_name: yek-aarch64-apple-darwin.tar.gz
+
+          # Native Windows builds
+          - os: windows-latest
+            target: x86_64-pc-windows-msvc
+            artifact_name: yek.exe
+            asset_name: yek-x86_64-pc-windows-msvc.zip
+          - os: windows-latest
+            target: aarch64-pc-windows-msvc
+            artifact_name: yek.exe
+            asset_name: yek-aarch64-pc-windows-msvc.zip
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Checkout VSCode repository
+        uses: actions/checkout@v4
+        with:
+          repository: microsoft/vscode
+          path: vscode
+          fetch-depth: 1
+
+      - name: Install cross (Linux)
+        if: matrix.use-cross
+        run: cargo install cross
+
+      - name: Setup Rust (Native builds)
+        if: ${{ !matrix.use-cross }}
+        uses: actions-rust-lang/setup-rust-toolchain@v1
+        with:
+          components: rustfmt,clippy
+
+      - name: Install Rust target
+        if: ${{ !matrix.use-cross }}
+        run: rustup target add ${{ matrix.target }}
+
+      - name: Build with cross (Linux)
+        if: matrix.use-cross
+        run: cross build --release --target ${{ matrix.target }}
+
+      - name: Native build (macOS/Windows)
+        if: ${{ !matrix.use-cross }}
+        run: cargo build --release --target ${{ matrix.target }}
+
+      - name: Install yek
+        run: cargo install --path . --all-features
+
+      - name: Run yek
+        timeout-minutes: 1
+        run: yek
+
+  benchmark:
+    name: Benchmark / ${{ matrix.benchmark_group.name }}
+    if: github.event_name == 'pull_request'
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        benchmark_group:
+          - group: "SingleFile_ByteMode"
+            name: "Single File Byte Mode"
+          - group: "SingleFile_ByteMode_Large"
+            name: "Single File Byte Mode Large"
+          - group: "SingleFile_TokenMode_Large"
+            name: "Single File Token Mode Large"
+          - group: "MultipleFiles_Small"
+            name: "Multiple Files Small"
+          - group: "MultipleFiles_Medium"
+            name: "Multiple Files Medium"
+          - group: "MultipleFiles_Large"
+            name: "Multiple Files Large"
+          - group: "MultipleFiles_TokenMode"
+            name: "Multiple Files Token Mode"
+          - group: "CustomConfig"
+            name: "Custom Config"
+
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      - name: Install Rust
+        uses: dtolnay/rust-toolchain@stable
+
+      - uses: Swatinem/rust-cache@v2
+        with:
+          cache-on-failure: true
+
+      - name: Build benchmarks on target branch
+        run: |
+          git fetch origin ${{ github.base_ref }}
+          git checkout ${{ github.base_ref }}
+          cargo bench --bench serialization --no-run
+
+      - name: Run benchmark on target branch
+        run: cargo bench --bench serialization -- --save-baseline ${{ github.base_ref }} '${{ matrix.benchmark_group.group }}/'
+
+      - name: Build benchmarks on PR branch
+        run: |
+          git checkout ${{ github.head_ref }}
+          cargo bench --bench serialization --no-run
+
+      - name: Compare benchmarks
+        run: |
+          cargo bench --bench serialization -- --baseline ${{ github.base_ref }} --noise-threshold 2 '${{ matrix.benchmark_group.group }}/' > benchmark_results.md
+          echo "## Benchmark Results for ${{ matrix.benchmark_group.name }}" >> $GITHUB_STEP_SUMMARY
+          cat benchmark_results.md >> $GITHUB_STEP_SUMMARY
+
+      - name: Upload benchmark results
+        uses: actions/upload-artifact@v4
+        with:
+          name: criterion-${{ matrix.benchmark_group.group }}-results
+          path: benchmark_results.md
+          if-no-files-found: error

.github/workflows/ci.yml

@@ -172,70 +172,6 @@ jobs:
           path: release-artifacts/${{ matrix.asset_name }}
           if-no-files-found: error
 
-  benchmark:
-    name: Benchmark / ${{ matrix.benchmark_group.name }}
-    if: github.event_name == 'pull_request'
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        benchmark_group:
-          - group: "SingleFile_ByteMode"
-            name: "Single File Byte Mode"
-          - group: "SingleFile_ByteMode_Large"
-            name: "Single File Byte Mode Large"
-          - group: "SingleFile_TokenMode_Large"
-            name: "Single File Token Mode Large"
-          - group: "MultipleFiles_Small"
-            name: "Multiple Files Small"
-          - group: "MultipleFiles_Medium"
-            name: "Multiple Files Medium"
-          - group: "MultipleFiles_Large"
-            name: "Multiple Files Large"
-          - group: "MultipleFiles_TokenMode"
-            name: "Multiple Files Token Mode"
-          - group: "CustomConfig"
-            name: "Custom Config"
-      fail-fast: false
-
-    steps:
-      - uses: actions/checkout@v3
-        with:
-          fetch-depth: 0
-
-      - name: Install Rust
-        uses: dtolnay/rust-toolchain@stable
-
-      - uses: Swatinem/rust-cache@v2
-        with:
-          cache-on-failure: true
-
-      - name: Build benchmarks on target branch
-        run: |
-          git fetch origin ${{ github.base_ref }}
-          git checkout ${{ github.base_ref }}
-          cargo bench --bench serialization --no-run
-
-      - name: Run benchmark on target branch
-        run: cargo bench --bench serialization -- --save-baseline ${{ github.base_ref }} '${{ matrix.benchmark_group.group }}/'
-
-      - name: Build benchmarks on PR branch
-        run: |
-          git checkout ${{ github.head_ref }}
-          cargo bench --bench serialization --no-run
-
-      - name: Compare benchmarks
-        run: |
-          cargo bench --bench serialization -- --baseline ${{ github.base_ref }} --noise-threshold 2 '${{ matrix.benchmark_group.group }}/' > benchmark_results.md
-          echo "## Benchmark Results for ${{ matrix.benchmark_group.name }}" >> $GITHUB_STEP_SUMMARY
-          cat benchmark_results.md >> $GITHUB_STEP_SUMMARY
-
-      - name: Upload benchmark results
-        uses: actions/upload-artifact@v4
-        with:
-          name: criterion-${{ matrix.benchmark_group.group }}-results
-          path: benchmark_results.md
-          if-no-files-found: error
-
   release:
     name: Release
     needs: [test, lint, build]

src/config.rs

@@ -78,6 +78,10 @@ pub struct YekConfig {
 
     /// Final resolved output file path (only used if not streaming)
     pub output_file_full_path: Option<String>,
+
+    /// Maximum depth to search for Git commit times
+    #[config_arg(accept_from = "config_only", default_value = "100")]
+    pub max_git_depth: i32,
 }
 
 /// Provide defaults so tests or other callers can create a baseline YekConfig easily.
@@ -104,6 +108,7 @@ impl Default for YekConfig {
             stream: false,
             token_mode: false,
             output_file_full_path: None,
+            max_git_depth: 100,
         }
     }
 }

src/lib.rs

@@ -45,7 +45,7 @@ pub fn serialize_repo(config: &YekConfig) -> Result<(String, Vec<ProcessedFile>)
         .par_iter()
         .filter_map(|dir| {
             let repo_path = Path::new(dir);
-            priority::get_recent_commit_times_git2(repo_path)
+            priority::get_recent_commit_times_git2(repo_path, config.max_git_depth as usize)
         })
         .flatten()
         .collect::<HashMap<String, u64>>();

src/priority.rs

@@ -1,4 +1,4 @@
-use git2::Repository;
+use git2;
 use regex;
 use serde::{Deserialize, Serialize};
 use std::{collections::HashMap, path::Path};
@@ -61,7 +61,11 @@ pub fn compute_recentness_boost(
 /// Get the commit time of the most recent change to each file using git2.
 /// Returns a map from file path (relative to the repo root) → last commit Unix time.
 /// If Git or .git folder is missing, returns None instead of erroring.
-pub fn get_recent_commit_times_git2(repo_path: &Path) -> Option<HashMap<String, u64>> {
+/// Only considers up to `max_commits` most recent commits.
+pub fn get_recent_commit_times_git2(
+    repo_path: &Path,
+    max_commits: usize,
+) -> Option<HashMap<String, u64>> {
     // Walk up until you find a .git folder but not higher than the base of the given repo_path
     let mut current_path = repo_path.to_path_buf();
     while current_path.components().count() > 1 {
@@ -71,7 +75,7 @@ pub fn get_recent_commit_times_git2(repo_path: &Path) -> Option<HashMap<String,
         current_path = current_path.parent()?.to_path_buf();
     }
 
-    let repo = match Repository::open(&current_path) {
+    let repo = match git2::Repository::open(&current_path) {
         Ok(repo) => repo,
         Err(_) => {
             debug!("Not a Git repository or unable to open: {:?}", current_path);
@@ -97,14 +101,15 @@ pub fn get_recent_commit_times_git2(repo_path: &Path) -> Option<HashMap<String,
     revwalk.set_sorting(git2::Sort::TIME).ok()?;
 
     let mut commit_times = HashMap::new();
-    for oid in revwalk {
-        let oid = match oid {
+    for oid_result in revwalk.take(max_commits) {
+        let oid = match oid_result {
             Ok(oid) => oid,
             Err(e) => {
                 debug!("Error during revwalk iteration: {:?}", e);
                 continue;
             }
         };
+
         let commit = match repo.find_commit(oid) {
             Ok(commit) => commit,
             Err(e) => {
@@ -119,8 +124,8 @@ pub fn get_recent_commit_times_git2(repo_path: &Path) -> Option<HashMap<String,
                 continue;
             }
         };
-        let time = commit.time().seconds() as u64;
 
+        let time = commit.time().seconds() as u64;
         tree.walk(git2::TreeWalkMode::PreOrder, |root, entry| {
             if let Some(name) = entry.name() {
                 if entry.kind() == Some(git2::ObjectType::Blob) {

tests/priority_test.rs

@@ -163,7 +163,7 @@ mod priority_tests {
     #[test]
     fn test_get_recent_commit_times_no_git() {
         let dir = tempdir().unwrap();
-        let times = get_recent_commit_times_git2(dir.path());
+        let times = get_recent_commit_times_git2(dir.path(), 100);
         assert!(times.is_none());
     }
 
@@ -216,7 +216,7 @@ mod priority_tests {
             .output()
             .unwrap();
 
-        let times = get_recent_commit_times_git2(repo_path).unwrap();
+        let times = get_recent_commit_times_git2(repo_path, 100).unwrap();
         assert_eq!(times.len(), 2);
         assert!(times.contains_key("file1.txt"));
         assert!(times.contains_key("file2.txt"));
@@ -234,7 +234,7 @@ mod priority_tests {
             .output()
             .unwrap();
 
-        let times = get_recent_commit_times_git2(repo_path);
+        let times = get_recent_commit_times_git2(repo_path, 100);
         assert!(times.is_none(), "Expected no times for empty repo");
     }
 
@@ -254,7 +254,7 @@ mod priority_tests {
         fs::remove_dir_all(repo_path.join(".git")).unwrap();
         fs::create_dir(repo_path.join(".git")).unwrap(); // Create an empty directory
 
-        let times = get_recent_commit_times_git2(repo_path);
+        let times = get_recent_commit_times_git2(repo_path, 100);
         assert!(times.is_none(), "Expected no times on Git failure");
     }
 }