From 051aaa5fe57de3a5e461693c4c75d5631382d1df Mon Sep 17 00:00:00 2001
From: Sergei Nizovtsev <snizovtsev@gmail.com>
Date: Sun, 12 Jan 2025 05:30:46 +0300
Subject: [PATCH] Fix git-tag cache-key reader in case of slashes (#10467)
 (#10500)

## Summary

The assumption that all tags are listed under a flat `.git/ref/tags`
structure was wrong. Git creates a hierarchy of directories for tags
containing slashes. To fix the cache key calculation, we need to
recursively traverse all files under that folder instead.

## Test Plan

1. Create an `uv` project with git-tag cache-keys;
2. Add any tag with slash;
3. Run `uv sync` and see uv_cache_info error in verbose log;
4. `uv sync` doesn't trigger reinstall on next tag addition or removal;
5. With fix applied, reinstall triggers on every tag update and there
are no errors in the log.

Fixes #10467

---------

Co-authored-by: Sergei Nizovtsev <sergei.nizovtsev@eqvilent.com>
Co-authored-by: Charlie Marsh <charlie.r.marsh@gmail.com>
---
 Cargo.lock                           |  1 +
 crates/uv-cache-info/Cargo.toml      |  1 +
 crates/uv-cache-info/src/git_info.rs | 32 +++++++++++++++++-----------
 3 files changed, 21 insertions(+), 13 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index d2ba39b1a..c6696c606 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -4739,6 +4739,7 @@ dependencies = [
  "thiserror 2.0.11",
  "toml",
  "tracing",
+ "walkdir",
 ]
 
 [[package]]
diff --git a/crates/uv-cache-info/Cargo.toml b/crates/uv-cache-info/Cargo.toml
index ec17a4416..6b10bbebe 100644
--- a/crates/uv-cache-info/Cargo.toml
+++ b/crates/uv-cache-info/Cargo.toml
@@ -23,3 +23,4 @@ serde = { workspace = true, features = ["derive"] }
 thiserror = { workspace = true }
 toml = { workspace = true }
 tracing = { workspace = true }
+walkdir = { workspace = true }
diff --git a/crates/uv-cache-info/src/git_info.rs b/crates/uv-cache-info/src/git_info.rs
index 9df098e69..ad566d13b 100644
--- a/crates/uv-cache-info/src/git_info.rs
+++ b/crates/uv-cache-info/src/git_info.rs
@@ -1,6 +1,9 @@
 use std::collections::BTreeMap;
 use std::path::{Path, PathBuf};
 
+use tracing::warn;
+use walkdir::WalkDir;
+
 #[derive(Debug, thiserror::Error)]
 pub(crate) enum GitInfoError {
     #[error("The repository at {0} is missing a `.git` directory")]
@@ -80,24 +83,27 @@ impl Tags {
             .find(|git_dir| git_dir.exists())
             .ok_or_else(|| GitInfoError::MissingGitDir(path.to_path_buf()))?;
 
-        let git_refs_path =
-            git_refs(&git_dir).ok_or_else(|| GitInfoError::MissingRefs(git_dir.clone()))?;
+        let git_tags_path = git_refs(&git_dir)
+            .ok_or_else(|| GitInfoError::MissingRefs(git_dir.clone()))?
+            .join("tags");
 
         let mut tags = BTreeMap::new();
 
         // Map each tag to its commit.
-        let read_dir = match fs_err::read_dir(git_refs_path.join("tags")) {
-            Ok(read_dir) => read_dir,
-            Err(err) if err.kind() == std::io::ErrorKind::NotFound => {
-                return Ok(Self(tags));
-            }
-            Err(err) => return Err(err.into()),
-        };
-        for entry in read_dir {
-            let entry = entry?;
+        for entry in WalkDir::new(&git_tags_path).contents_first(true) {
+            let entry = match entry {
+                Ok(entry) => entry,
+                Err(err) => {
+                    warn!("Failed to read Git tags: {err}");
+                    continue;
+                }
+            };
             let path = entry.path();
-            if let Some(tag) = path.file_name().and_then(|name| name.to_str()) {
-                let commit = fs_err::read_to_string(&path)?.trim().to_string();
+            if !entry.file_type().is_file() {
+                continue;
+            }
+            if let Ok(Some(tag)) = path.strip_prefix(&git_tags_path).map(|name| name.to_str()) {
+                let commit = fs_err::read_to_string(path)?.trim().to_string();
 
                 // The commit should be 40 hexadecimal characters.
                 if commit.len() != 40 {