Mon, 14 Jan 08 - Searching subversion history

I’ve often felt the need to search through our subversion repository to find, for example, when a method was removed/renamed. As far as I’m aware there’s no easy way to do this using the standard subversion, or trac, tools. A few days ago the need arose once more: I wanted to know when a particular class had stopped being used within our codebase. It struck me that if I had a diff file for each changeset that I’d be able to grep for the change I was looking for. I hacked together a script to produce a diff of each changeset (actually, as I knew when the class was added, I only produced a diff for a subset of the changesets) and was able to find the information I needed. The script is pasted below and on google code. It assumes that you are running it from within an svn working direction. It took a while to produce the diff files (I’m afraid I don’t have any stats) but the result was a really fast way of finding, otherwise hard to find, information. I guess it’d be quite easy to automate the diff creation process so that you had an always up-to-date bunch of diff fies…

Is this useful, or have I missed something obvious in Subversion or Trac that does exactly this?

require 'fileutils'

stop_at_revision = ARGV[0]
raise "Please specify the revision that you wish to go back to as the only argument to this script." unless stop_at_revision
stop_at_revision = Integer(stop_at_revision)

class Time
  def friendly_format
    strftime("%Y-%m-%d %H:%M:%S")
  end
end

def msg(message)
  puts "#{Time.now.friendly_format} - #{message}"
end

def produce_diff(revisions, earlier_revision = nil)
  if revisions.empty?
    msg "All finished."
    exit
  end

  later_revision = earlier_revision ? earlier_revision : revisions.shift
  earlier_revision = revisions.shift

  msg "Creating diff from revision #{earlier_revision} to #{later_revision}."
  `svn diff -r#{earlier_revision}:#{later_revision} > patches/#{later_revision}.patch`
  produce_diff(revisions, earlier_revision)
end

FileUtils.mkdir_p 'patches'

msg "Scanning output of svn log to find all revisions that we care about..."
log = `svn log -rHEAD:#{stop_at_revision} -q`
revisions = log.scan(/^r(\d+)/)
revisions = revisions.flatten.collect { |revision| Integer(revision) }

produce_diff(revisions)