Commit Message Ratio - a made up ratio defined as the ratio of the number of words in the message body to the lines of code changed

commit-message-ratio.rb

#!/usr/bin/env ruby

# Define the "commit message ratio" as (commit_body_words / commit_lines_changed).
#
# With no args, this script will find the commit message ratio of all commits in
# the git repo in the pwd, and will sort them by that ratio.
# (Commits with a high commit message ratio tend to be interesting commits!)
#
# With a commit hash as the only arg, provide the commit message ratio of that commit.
#
# Runs using git in the PWD.

def all_ratios
    # Adapted from
    # https://gist.github.com/samvrlewis/a4545386adb99556e7ff629d377e1769
    msg_wc_out = `git rev-list --no-merges "${1:-HEAD}" | while read rev; do echo "$(git show -s --format='%b' $rev | wc -w) words: $rev"; done`

    msg_wc_hash = msg_wc_out.lines.map { |line|
        parts = line.split(":").map(&:strip)
        words = parts.first.match(/(\d+) words/)[1].to_i
        [parts.last, words]
    }.to_h

    # Adapted from
    # https://stackoverflow.com/questions/21137477/how-to-get-git-log-with-short-stat-in-one-line
    loc_out = `git log --oneline --pretty="@%H" --stat | grep -v \\| | tr "\\n" " " | tr "@" "\\n"`

    loc_hash = loc_out.lines.map { |line|
        line.match(/(?'commit'[0-9a-z]+)   \d+ files? changed(?:, (?'ins'\d+) insertions?\(\+\))?(?:, (?'del'\d+) deletions?\(\-\))?/)
    }.select { |m| !m.nil? }.map { |matches|
        insertions = matches.named_captures['ins'].to_i || 0
        deletions = matches.named_captures['del'].to_i || 0
        [matches['commit'], insertions + deletions]
    }.select { |_, lines| lines > 0 }.to_h

    subject_out = `git log --pretty="%H::::%s"`
    subjects = subject_out.lines.map { |line|
        parts = line.split("::::").map(&:strip)
        [parts.first, parts.last]
    }.to_h

    ratios = loc_hash.map { |commit, lines_changed|
        msg_wc = msg_wc_hash[commit]
        {commit: commit, ratio: (msg_wc.to_f / lines_changed).round(2), subject: subjects[commit]}
    }

    pp ratios.sort_by { |data| - data[:ratio] }
end

def single_ratio(commit)
    msg_wc = `git show -s --format='%b' #{commit} | wc -w`.to_i
    loc_out = `git log --oneline --pretty="@%H" --stat #{commit} ^#{commit}^ | grep -v \\| | tr "\\n" " " | tr "@" "\\n"`.strip
    matches = loc_out.match(/(?'commit'[0-9a-z]+)   \d+ files? changed(?:, (?'ins'\d+) insertions?\(\+\))?(?:, (?'del'\d+) deletions?\(\-\))?/)
    insertions = matches.named_captures['ins'].to_i || 0
    deletions = matches.named_captures['del'].to_i || 0
    loc = insertions + deletions

    puts "          Word Count: #{msg_wc}"
    puts "                 LOC: #{loc}"
    puts "Commit Message Ratio: #{msg_wc.to_f / loc}"
end

if ARGV.size == 0
    all_ratios
else
    single_ratio(ARGV.first)
end