chore: release v0.1.8

[onsails]

🤖 New release

• ccsync-core: 0.1.7 -> 0.1.8 (✓ API compatible changes)
• ccsync: 0.1.7 -> 0.1.8

Changelog

ccsync-core

0.1.8 - 2025-11-24

Fixed

• override r--r--r--

Other

• x.x versions

ccsync

0.1.8 - 2025-11-24

Other

• update Cargo.toml dependencies

---

This PR was generated with release-plz.

[claude]

Code Review: PR #22 - Release v0.1.8

Summary

This is a routine release PR bumping versions from v0.1.7 to v0.1.8. However, there is a critical violation of the project's dependency versioning standards that must be fixed before merging.

---

❌ Critical Issue: Dependency Versioning Format Violation

Location: crates/ccsync/Cargo.toml:17

Current (Incorrect):

ccsync-core = { version = "0.1.8", path = "../ccsync-core" }

Required (Per CLAUDE.rust.md §1):

ccsync-core = { version = "0.1", path = "../ccsync-core" }

Rationale:
According to the project's Rust standards in CLAUDE.rust.md, Section 1 states:

Always use requirements: x.x (ensures patch compatibility)
Example: serde = "1.0"

Using the three-part version "0.1.8" violates this standard. The internal workspace dependency should use the two-part format "0.1" to ensure patch compatibility, meaning any 0.1.z version will satisfy the requirement.

Impact:

• ✅ All workspace dependencies in root Cargo.toml correctly use x.x format (clap = "4.5", anyhow = "1.0", etc.)
• ❌ Only the internal ccsync-core dependency violates this standard
• This creates inconsistency between external and internal dependency versioning

---

✅ What's Correct

1. Semantic Versioning: Correctly uses patch version bump (0.1.7 → 0.1.8) for:

   • Metadata changes in ccsync-core
   • Internal dependency updates in ccsync

2. Consistency: All related files updated consistently:

   • ✅ Workspace version in root Cargo.toml
   • ✅ Cargo.lock reflects new versions
   • ✅ Both CHANGELOGs updated with matching dates (2025-11-24)

3. Standards Compliance:

   • ✅ Uses Cargo edition 2024
   • ✅ Proper workspace architecture maintained
   • ✅ External dependencies follow x.x format

---

💭 Observations

CHANGELOG Quality:
The CHANGELOG entries are minimal:

• ccsync-core: "x.x versions"
• ccsync: "update Cargo.toml dependencies"

The entry "x.x versions" suggests this release was intended to standardize versioning format, which makes it ironic that the PR itself violates that standard by using "0.1.8" instead of "0.1".

Consider enhancing CHANGELOG entries to better communicate purpose to users:

### Other
- Standardize internal dependency versioning to x.x format

---

🔧 Required Action

Fix the dependency version format:

- ccsync-core = { version = "0.1.8", path = "../ccsync-core" }
+ ccsync-core = { version = "0.1", path = "../ccsync-core" }

This ensures:

• ✅ Compliance with project standards
• ✅ Patch compatibility (any 0.1.z version satisfies)
• ✅ Consistency with external dependency versioning
• ✅ Alignment with workspace version control

---

Recommendation

Request Changes: Please update the dependency version format before merging.

Once fixed, this will be a clean release PR that properly aligns with project standards.

[claude]

PR Review - Release v0.1.8

Summary

This is a clean automated release PR generated by release-plz. The changes are minimal and follow proper versioning practices.

Code Quality & Best Practices

• Version Management: Properly bumps workspace version from 0.1.7 to 0.1.8
• Dependency Consistency: Correctly updates ccsync-core dependency version in crates/ccsync/Cargo.toml to match the new release
• Changelog Maintenance: Follows Keep a Changelog format with proper version links
• Semantic Versioning: Version bump is appropriate for the changes described

Potential Issues

• None identified: These are straightforward version number changes with no code modifications

Performance Considerations

• N/A: No code changes that would impact performance

Security Concerns

• None: No code changes that would introduce security vulnerabilities

Test Coverage

• N/A: Version bumps do not require additional test coverage

Changes Overview

1. Workspace version: 0.1.7 to 0.1.8 in root Cargo.toml
2. Dependency update: ccsync-core version in crates/ccsync/Cargo.toml
3. Changelog entries added for both ccsync-core and ccsync crates

Recommendation

APPROVE - This PR is ready to merge. All changes are consistent, follow project conventions, and maintain proper version synchronization across the workspace.

Note: This release PR does not conflict with pending taskmaster tasks (7, 8) which are feature additions beyond this version bump.

---

Review generated by Claude Code following project CLAUDE.md guidelines

[claude]

Code Review - PR #22: Release v0.1.8

Summary

✅ APPROVED with recommendations for follow-up

This release PR looks good for merging. The version bump from 0.1.7 → 0.1.8 is straightforward and includes the important fix for handling read-only files during sync operations.

---

Critical Issues Found (Not in this PR, but in codebase)

While reviewing the code context, I found several critical issues that should be addressed in follow-up PRs:

🔴 CRITICAL: Platform-Specific Test Missing #[cfg(unix)]

Location: crates/ccsync-core/src/sync/executor.rs:327-348

The test test_copy_file_overwrites_readonly_destination uses Unix-specific APIs without proper conditional compilation:

#[test]
fn test_copy_file_overwrites_readonly_destination() {
    use std::os::unix::fs::PermissionsExt;  // ❌ Will fail on Windows\!
    // ...
    perms.set_mode(0o444);  // ❌ Unix-only API
}

Fix:

#[test]
#[cfg(unix)]  // ✅ Add this
fn test_copy_file_overwrites_readonly_destination() {
    // ... existing test
}

#[test]
#[cfg(windows)]  // ✅ Add Windows version
fn test_copy_file_overwrites_readonly_destination_windows() {
    // ... use set_readonly(true) instead
}

---

🔴 CRITICAL: Symlink Attack Vector

Location: crates/ccsync-core/src/sync/executor.rs:189-190, 211-212

fs::remove_dir_all() follows symlinks, which could be exploited:

if dest.exists() {
    fs::remove_dir_all(dest)?;  // ❌ DANGEROUS if dest is a symlink\!
}

Fix:

if dest.exists() {
    let metadata = fs::symlink_metadata(dest)?;  // ✅ Don't follow symlinks
    if metadata.is_symlink() {
        fs::remove_file(dest)?;
    } else if metadata.is_dir() {
        fs::remove_dir_all(dest)?;
    } else {
        fs::remove_file(dest)?;
    }
}

---

🔴 CRITICAL: Known Bug in Interactive Approval

Location: crates/ccsync-core/src/sync/orchestrator.rs:432-444

There's a documented bug in the test suite:

// BUG: Currently this test fails because even though user approved,
// the executor sees ConflictStrategy::Fail and bails

This needs investigation and fixing.

---

Security Issues (Medium Priority)

🟠 TOCTOU Race Condition

Location: crates/ccsync-core/src/sync/executor.rs:134-160

Time-of-check-time-of-use vulnerability between checking if a file is read-only and modifying it. Another process could modify the file in between.

🟠 Unwrap in Production Code

Location: crates/ccsync-core/src/sync/executor.rs:252

let file_name = path.file_name().unwrap();  // ❌ Can panic

Should use proper error handling:

let file_name = path.file_name().ok_or_else(|| {
    anyhow::anyhow\!("Invalid path: no file name in {}", path.display())
})?;

---

Code Quality Issues

🟡 Missing Edge Case Tests

The test suite should include:

• Unicode/special characters in filenames
• Very long paths (>260 chars on Windows)
• Disk space errors
• Concurrent access scenarios
• Deeply nested directories

🟡 Potential Stack Overflow

Location: crates/ccsync-core/src/comparison/directory.rs:129-147

Recursive directory traversal without depth limit could overflow on extremely deep directory trees. Consider adding a max depth parameter.

---

Performance Considerations

• ✅ File hashing uses streaming (8KB buffer) - good for memory
• ✅ Proper error contexts throughout
• ⚠️ Directory copying is sequential - could benefit from parallelization for large trees
• ⚠️ No progress indication for long operations

---

Test Coverage Assessment

Current Coverage: ~70% (estimated)

Well-covered:

• ✅ Basic file operations
• ✅ Conflict strategies
• ✅ Dry-run mode
• ✅ Pattern matching

Missing:

• ❌ Cross-platform permission handling
• ❌ Symlink edge cases
• ❌ Large files (>1GB)
• ❌ Network filesystems

---

Recommendations

For this PR:

• ✅ Approve and merge - the version bump and changelog are correct

Immediate follow-up (v0.1.9):

1. Add #[cfg(unix)] to the readonly test
2. Fix symlink attack vector in remove_dir_all
3. Fix documented bug in interactive approval
4. Replace .unwrap() at executor.rs:252

Future improvements:
5. Add TOCTOU mitigation for permission checking
6. Add depth limit to recursive traversal
7. Expand test coverage for edge cases
8. Consider parallelization for large operations

---

Positive Aspects ✨

• Excellent error handling with anyhow::Context
• Clean separation of concerns
• Good use of #[must_use] and documentation
• No unsafe code
• Proper workspace lints configured

---

Overall Assessment

Grade: B+ (Good, with room for improvement)

The code is well-structured and follows Rust best practices. The read-only file fix is functionally correct. The critical issues found are not introduced by this PR but exist in the current codebase and should be addressed soon.

Recommendation: ✅ APPROVE and MERGE