#!/usr/bin/python
#
# Find directories with duplicate content.
#

import sys
import os
import stat
import itertools
import operator
import subprocess

# should we perform a "deep inspection" of the candidate directories?
DEEP_INSPECT = False


def find(parent_dir):
  finfo = _statall(parent_dir)
  subdirs = [f for f, s in finfo if stat.S_ISDIR(s.st_mode)]

  print 'CHECKING:', len(subdirs), 'subdirectories'
  all_info = [ ]
  for subdir in subdirs:
    path = os.path.join(parent_dir, subdir)
    # to find dup backups of DVDs
    video_ts = os.path.join(path, 'VIDEO_TS')
    if os.path.exists(video_ts):
      path = video_ts
    finfo = frozenset((f, s.st_size)
                      for f, s in _statall(path) if stat.S_ISREG(s.st_mode))
    all_info.append((subdir, finfo, path))

  # Sort based on the subdir file info (names/sizes)
  all_info = sorted(all_info, key=lambda x: hash(x[1]))

  # Locate all subdirs that have the same finfo
  for finfo, group in itertools.groupby(all_info, operator.itemgetter(1)):
    group = list(group)
    candidates = [subdir for subdir, _, _ in group]
    if len(candidates) > 1:
      print 'Candidates:', ', '.join(repr(s) for s in candidates)
      if DEEP_INSPECT:
        path1 = group[0][2]
        files = [f for f, s in group[0][1]]
        for _, _, path2 in group[1:]:
          _deep_inspect(path1, path2, files)


def _deep_inspect(path1, path2, files):
  print '  INSPECTING:', path1, path2
  for f in files:
    print '    Comparing:', f, '...',
    sys.stdout.flush()
    rc = subprocess.call(['cmp', os.path.join(path1, f), os.path.join(path2, f)])
    if rc == 0:
      print 'same'
    else:
      print 'DIFFERENT. (rc=%d)' % (rc,)
      return
    
  print '    ==> DIRECTORY CONTENTS MATCH'


def _statall(dirname):
  files = filter(lambda f: not f.startswith('.'), os.listdir(dirname))
  return [(f, os.lstat(os.path.join(dirname, f))) for f in files]


if __name__ == '__main__':
  if len(sys.argv) > 1 and sys.argv[1] == '--deep':
    DEEP_INSPECT = True
    del sys.argv[1]
  if len(sys.argv) > 2:
    print 'ERROR: %s [PARENT_DIR]' % (os.path.basename(sys.argv[0]),)
    sys.exit(1)
  if len(sys.argv) == 2:
    parent_dir = sys.argv[1]
  else:
    parent_dir = '.'
  find(parent_dir)