git-swear-stats 0.1.0

data/LICENSE

@@ -0,0 +1,14 @@
+            DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE
+                    Version 2, December 2004
+
+ Copyright (C) 2013 James Pearson
+
+ Everyone is permitted to copy and distribute verbatim or modified
+ copies of this license document, and changing it is allowed as long
+ as the name is changed.
+
+            DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE
+   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+  0. You just DO WHAT THE FUCK YOU WANT TO.
+

data/README.md

@@ -0,0 +1,44 @@
+Inspired by [the Linux Kernel Fuck Count] and [its descendent],
+`git-swear-stats` will give you some interesting statistics on swearing in a
+git repository's commit messages.
+
+[the Linux Kernel Fuck Count]: http://durak.org/sean/pubs/kfc/
+[its descendent]: http://www.vidarholen.net/contents/wordcount/
+
+# Installation
+
+    [$]> gem install git-swear-stats
+
+# Usage
+
+    [$]> git-swear-stats --help
+    git-swear-stats
+    
+    Usage:
+       git-swear-stats [options] [messages|diffs]
+    
+    Options:
+       -h, --help        Show this screen.
+       --debug           Print out debug messages.
+       --include-merges  Look at merge commits.
+       --no-progress     Don't print out progress information.
+
+# Example Output
+
+    [$]> git swear-stats 
+    Reading in git log... done!
+    Parsing git log.
+    Overall:
+    +-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+    | clusterfuck | fuck | fsck | eff | shit | cunt | cock | ass | arse | bitch | wtf | fag | damn | bastard | hell |
+    +-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+    | 1           | 3    | 2    | 2   | 2    | 1    | 1    | 1   | 1    | 1     | 3   | 1   | 1    | 1       | 1    |
+    +-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+    Authors:
+    +---------------+-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+    |               | clusterfuck | fuck | fsck | eff | shit | cunt | cock | ass | arse | bitch | wtf | fag | damn | bastard | hell |
+    +---------------+-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+    | xiongchiamiov | 0           | 1    | 0    | 0   | 0    | 0    | 0    | 0   | 0    | 0     | 2   | 0   | 0    | 0       | 0    |
+    | James Pearson | 1           | 2    | 2    | 2   | 2    | 1    | 1    | 1   | 1    | 1     | 1   | 1   | 1    | 1       | 1    |
+    +---------------+-------------+------+------+-----+------+------+------+-----+------+-------+-----+-----+------+---------+------+
+

data/Rakefile

@@ -0,0 +1,3 @@
+require 'mg'
+MG.new 'git-swear-stats.gemspec'
+

data/bin/git-swear-stats

@@ -0,0 +1,155 @@
+#!/usr/bin/env ruby
+# encoding: utf-8
+
+# May you recognize your weaknesses and share your strengths.
+# May you share freely, never taking more than you give.
+# May you find love and love everyone you find.
+
+require 'docopt'
+require 'git'
+require 'pp'
+require 'terminal-table'
+
+# Git intercepts `git swear-stats --help`, so unfortunately most people won't
+# see this usage message.  Docopt is still a great way to do parsing, though.
+doc = <<DOCOPT
+git-swear-stats
+
+Usage:
+   git-swear-stats [options] [messages|diffs]
+
+Options:
+   -h, --help        Show this screen.
+   --debug           Print out debug messages.
+   --include-merges  Look at merge commits.
+   --no-progress     Don't print out progress information.
+
+DOCOPT
+begin
+   options = Docopt::docopt doc
+rescue Docopt::Exit => e
+   puts e.message
+   exit 1
+end
+pp options if options['--debug']
+
+swears = [
+   %r{^(dip)?shit(ty)?$},
+   %r{^(cluster|mother)?(fuck|fsck|eff)(ing|er|ed|head)?$},
+   %r{^cunt$},
+   %r{^cock(suck|sucker)?$},
+   %r{^(ass|arse)(hole)?$},
+   %r{^bitch(y)?$},
+   %r{^wtf$},
+   %r{^fag(got)?$},
+   %r{^(god)?damn(ed|able)?$},
+   %r{^bastard$},
+   %r{^hell(ish)?$},
+]
+wordDelimiters = %r{[ \-_.,!?;:*"'|{}()\n]}
+
+wordStats = {}
+wordStats.default = 0
+authorStats = {}
+
+repo = Git.open '.'
+# By default, we only get up to 30 commits in the log.  By digging through the
+# source, I found that passing nil to repo.log() ends up setting the limit to
+# nil, and thus giving us all commits.
+print 'Reading in git log... ' if not options['--no-progress']
+log = repo.log nil
+puts 'done!' if not options['--no-progress']
+
+print 'Parsing git log' if not options['--no-progress']
+log.each_with_index do |commit, i|
+   # Merges often have data in them from other commits, pull request
+   # descriptions, etc., so it's not really fair to count them.
+   # Git::Log unfortunately appears to have no way of passing --no-merges
+   # through to the underlying binary, so we have to check merges on a
+   # commit-by-commit basis instead.
+   next if not options['--include-merges'] and commit.parents.count > 1
+   
+   author = commit.author.name
+   if not authorStats.key? author
+      authorStats[author] = {}
+      authorStats[author].default = 0
+   end
+   
+   if !options['diffs']
+      # Splitting on words makes our regexes a little simpler.
+      commit.message.split(wordDelimiters).each do |word|
+         word.downcase! # A little bit of normalization.
+         swears.each do |swear|
+            if word =~ swear
+               wordStats[word] += 1
+               authorStats[author][word] += 1
+               if options['--debug']
+                  puts '-'*80
+                  puts author
+                  puts commit.message
+                  puts '-'*80
+               end
+            end
+         end
+      end
+   end
+   
+   if !options['messages']
+      # The git module doesn't handle a diff on the initial commit very well - by
+      # which I mean it throws an exception.
+      # TODO: Monkeypatch the shit out of Git::Commit.
+      next if commit.parents.count == 0
+      
+      diff = commit.diff_parent.to_s.force_encoding 'iso-8859-1'
+      # We only want to look at stuff they did, not surrounding things.  We'll
+      # approximate this by grepping through lines "added".
+      lines = diff.split("\n")
+      # Drop the first four lines to get rid of the diff header.
+      lines.drop(4).keep_if {|line| line =~ /^\+/}.each do |line|
+         line[1..-1].split(wordDelimiters).each do |word|
+            word.downcase! # A little bit of normalization.
+            swears.each do |swear|
+               if word =~ swear
+                  wordStats[word] += 1
+                  authorStats[author][word] += 1
+                  if options['--debug']
+                     puts '-'*80
+                     puts author
+                     puts line
+                     puts '-'*80
+                  end
+               end
+            end
+         end
+      end
+   end
+   
+   # For large repos, parsing can take a while; let users know something is
+   # happening.
+   print '.' if i % 100 == 0 and not options['--no-progress']
+end
+puts if not options['--no-progress']
+
+# People who haven't sworn at all aren't very interesting.
+authorStats.delete_if {|author, swears| swears.empty?}
+
+pp wordStats if options['--debug']
+pp authorStats if options['--debug']
+
+puts 'Overall:'
+table = Terminal::Table.new
+table.headings = wordStats.keys
+table.add_row wordStats.values
+puts table
+
+puts 'Authors:'
+table = Terminal::Table.new
+table.headings = header = [''] + wordStats.keys
+authorStats.each do |author, stats|
+   row = []
+   stats[''] = author # Cheat a little for the first column.
+   header.each {|swear| row << stats[swear]}
+   table.add_row row
+end
+puts table
+

metadata

@@ -0,0 +1,93 @@
+--- !ruby/object:Gem::Specification 
+name: git-swear-stats
+version: !ruby/object:Gem::Version 
+  prerelease: 
+  version: 0.1.0
+platform: ruby
+authors: 
+- James Pearson
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2013-03-24 00:00:00 -07:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: docopt
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+  type: :runtime
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: git
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+  type: :runtime
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: terminal-table
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+  type: :runtime
+  version_requirements: *id003
+description: Gather statistics on swearing in a repo.
+email: 
+- xiong.chiamiov@gmail.com
+executables: 
+- git-swear-stats
+extensions: []
+
+extra_rdoc_files: 
+- LICENSE
+- README.md
+files: 
+- LICENSE
+- Rakefile
+- README.md
+- bin/git-swear-stats
+has_rdoc: true
+homepage: https://github.com/xiongchiamiov/git-swear-stats
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.5.3
+signing_key: 
+specification_version: 3
+summary: Gather statistics on swearing in a repo.
+test_files: []
+