#!/usr/bin/python
# -*- encoding: utf-8 -*-
#
# exif-rewrite - Rewrites EXIF tags according to a translation dictionary
#
# Usage: exif-rewrite rules.ini JPG_files
#
# Copyright (C) Miguel Ángel Vilela <miguev@gmail.com>
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.

import ConfigParser
import pyexiv2
import re
import sys

MultipleValuesRegex = re.compile('^(\+|\-)[^,]+(, *(\+|\-)[^,]+)*$')

encoding = sys.argv[1]
if encoding in ('latin-1', 'utf-8'):
  config = sys.argv[2]
  images = sys.argv[3:]
else:
  config = sys.argv[1]
  images = sys.argv[2:]
# Load rules from .ini file
conf = ConfigParser.RawConfigParser()
conf.read(config)
search_tags = set(conf.get('General', 'Search.Tags').split(','))
replace_tags = set(conf.get('General', 'Replace.Tags').split(','))
replace_tags_names = dict([(tag.lower().strip(), tag.strip()) for tag in replace_tags])
search_values = set(conf.sections()) - set(['General'])

# Rewrite EXIF tags on each image
for image in images:
  print "Reading '%s'" % image
  im = pyexiv2.Image(image)
  im.readMetadata()
  image_tags = set(im.exifKeys() + im.iptcKeys())
  interesting_tags = list(search_tags.intersection(image_tags))
  already_matched_values = []
  for search_tag in interesting_tags:
    search_tag_modified = False
    original_value = im[search_tag]
    if type(original_value) == type(""):
      original_value = [original_value]
    if encoding == 'latin-1':
      original_value = [v.decode('latin-1').encode('utf-8') for v in original_value]
#    elif encoding == 'utf-8':
#      original_value = [v.decode('utf-8').encode('latin-1') for v in im[search_tag]]
    values = set(original_value)
    for value in values.intersection(search_values):
      if value in already_matched_values:
#        print "Ignoring match: %s contains '%s'" % (search_tag, value)
	continue
#      print "Found match: %s contains '%s'" % (search_tag, value)
      for replace_tag in conf.options(value):
        replace_data = conf.get(value, replace_tag)
        if encoding == 'latin-1':
          replace_data = replace_data.decode('utf-8').encode('latin-1')
        elif encoding == 'utf-8':
          replace_data = replace_data.decode('latin-1').encode('utf-8')
        if MultipleValuesRegex.match(replace_data):
          actual_data = im[replace_tags_names[replace_tag]]
          if type(actual_data) == type(""):
            actual_data = list([actual_data])
          else:
            actual_data = list(actual_data)
          for new_value in replace_data.split(','):
            op, new_value = new_value.strip()[0], new_value.strip()[1:]
            if op == '+':
	      if new_value in actual_data:
#                print "Will not append '%s' to %s AGAIN" % (new_value, replace_tags_names[replace_tag])
                continue
              print "Append '%s' to %s" % (new_value, replace_tags_names[replace_tag])
              # Append current search_tag for re-consideration, but avoid matching just-appended value
	      already_matched_values.append(value)
              interesting_tags.append(replace_tags_names[replace_tag])
              actual_data.append(new_value)
              search_tag_modified = True
            elif op == '-':
              print "Remove '%s' from %s" % (new_value, replace_tags_names[replace_tag])
              if new_value in actual_data:
	        actual_data.remove(new_value)
                search_tag_modified = True
            else:
              print "This should never happen!"
              sys.exit(1)
	  if len(actual_data) == 1:
	    actual_data.append('')
	  actual_data = [e for e in list(set(actual_data)) if len(e) > 1] # Eliminate duplicate and one-char values
	  if set(im[replace_tags_names[replace_tag]]) != set(actual_data):
            im[replace_tags_names[replace_tag]] = actual_data
            search_tag_modified = True
#          print "%s now contains: %s" % (replace_tags_names[replace_tag], im[replace_tags_names[replace_tag]])
        else:
	  if replace_tags_names[replace_tag] in image_tags:
	    if (type(im[replace_tags_names[replace_tag]]) == type(replace_data) and \
	        im[replace_tags_names[replace_tag]] == replace_data) or \
	        str(im[replace_tags_names[replace_tag]]) == str(replace_data) or \
	       (type(im[replace_tags_names[replace_tag]]) == type(()) and \
	        ' '.join([str(x) for x in im[replace_tags_names[replace_tag]]]) == replace_data):
	      continue
          print "Setting %s to '%s'" % (replace_tags_names[replace_tag], replace_data)
          im[replace_tags_names[replace_tag]] = replace_data
          search_tag_modified = True
    # Write metadata after each interesting_tag since the list may get new ones
    if not search_tag_modified:
      continue
    original_value = im[search_tag]
    if type(original_value) == type(""):
      original_value = [original_value]
    im[search_tag] = [d for d in original_value if len(d) > 1]
    print "Writing '%s'" % image
    im.writeMetadata()
  print "Done with '%s'" % image