tools-count-sloc

Count lines of source code and present complexity metrics for the codebase.

Input

Optional argument: a subdirectory to scope the analysis (e.g. components/dispatcher/v0). Defaults to the repository root.

Steps

1. Determine the target directory:

REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || pwd)
if [[ -n "$ARGUMENTS" ]]; then
    TARGET="$REPO_ROOT/$ARGUMENTS"
    if [[ ! -d "$TARGET" ]]; then
        echo "Error: directory not found: $TARGET" >&2
        exit 1
    fi
else
    TARGET="$REPO_ROOT"
fi

2. Ensure tokei is installed (it's a fast, accurate SLOC counter written in Rust):

if ! command -v tokei &>/dev/null; then
    echo "Installing tokei..."
    cargo install tokei --quiet
fi

3. Run tokei for the SLOC summary, excluding build artifacts and vendored dependencies:

tokei "$TARGET" --exclude deps/spdk-build --exclude target --sort code

4. Run the complexity analysis script to produce per-component metrics:

SCRIPT="$(git rev-parse --show-toplevel)/.claude/skills/tools-count-sloc/complexity.sh"
bash "$SCRIPT" "$TARGET"

5. Present the results to the user in a clear summary with:

   • Total lines of code (by language)
   • Breakdown by component/directory
   • Complexity indicators (files with highest function counts, deepest nesting)
   • Comparison of code vs tests vs comments ratios

6. Write out the presented results to SUMMARY.md file. Interactively allow user to specify a different directory.