mediagoblin/media_types/image/processing.py

   1 # GNU MediaGoblin -- federated, autonomous media hosting
   2 # Copyright (C) 2011, 2012 MediaGoblin contributors.  See AUTHORS.
   3 #
   4 # This program is free software: you can redistribute it and/or modify
   5 # it under the terms of the GNU Affero General Public License as published by
   6 # the Free Software Foundation, either version 3 of the License, or
   7 # (at your option) any later version.
   8 #
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 # GNU Affero General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU Affero General Public License
  15 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  16
  17 try:
  18     from PIL import Image
  19 except ImportError:
  20     import Image
  21 import os
  22 import logging
  23 import argparse
  24
  25 from mediagoblin import mg_globals as mgg
  26 from mediagoblin.db.models import MediaEntry
  27 from mediagoblin.processing import BadMediaFail, FilenameBuilder
  28 from mediagoblin.submit.lib import run_process_media
  29 from mediagoblin.tools.exif import exif_fix_image_orientation, \
  30     extract_exif, clean_exif, get_gps_data, get_useful, \
  31     exif_image_needs_rotation
  32 from mediagoblin.tools.translate import lazy_pass_to_ugettext as _
  33
  34 _log = logging.getLogger(__name__)
  35
  36 PIL_FILTERS = {
  37     'NEAREST': Image.NEAREST,
  38     'BILINEAR': Image.BILINEAR,
  39     'BICUBIC': Image.BICUBIC,
  40     'ANTIALIAS': Image.ANTIALIAS}
  41
  42 MEDIA_TYPE = 'mediagoblin.media_types.image'
  43
  44
  45 def resize_image(proc_state, resized, keyname, target_name, new_size,
  46                  exif_tags, workdir):
  47     """
  48     Store a resized version of an image and return its pathname.
  49
  50     Arguments:
  51     proc_state -- the processing state for the image to resize
  52     resized -- an image from Image.open() of the original image being resized
  53     keyname -- Under what key to save in the db.
  54     target_name -- public file path for the new resized image
  55     exif_tags -- EXIF data for the original image
  56     workdir -- directory path for storing converted image files
  57     new_size -- 2-tuple size for the resized image
  58     """
  59     config = mgg.global_config['media_type:mediagoblin.media_types.image']
  60
  61     resized = exif_fix_image_orientation(resized, exif_tags)  # Fix orientation
  62
  63     filter_config = config['resize_filter']
  64     try:
  65         resize_filter = PIL_FILTERS[filter_config.upper()]
  66     except KeyError:
  67         raise Exception('Filter "{0}" not found, choose one of {1}'.format(
  68             unicode(filter_config),
  69             u', '.join(PIL_FILTERS.keys())))
  70
  71     resized.thumbnail(new_size, resize_filter)
  72
  73     # Copy the new file to the conversion subdir, then remotely.
  74     tmp_resized_filename = os.path.join(workdir, target_name)
  75     with file(tmp_resized_filename, 'w') as resized_file:
  76         resized.save(resized_file, quality=config['quality'])
  77     proc_state.store_public(keyname, tmp_resized_filename, target_name)
  78
  79
  80 def resize_tool(proc_state, force, keyname, target_name,
  81                 conversions_subdir, exif_tags, new_size=None):
  82     # filename -- the filename of the original image being resized
  83     filename = proc_state.get_orig_filename()
  84
  85     # Use the default size if new_size was not given
  86     if not new_size:
  87         max_width = mgg.global_config['media:' + keyname]['max_width']
  88         max_height = mgg.global_config['media:' + keyname]['max_height']
  89         new_size = (max_width, max_height)
  90
  91     # If the size of the original file exceeds the specified size for the desized
  92     # file, a target_name file is created and later associated with the media
  93     # entry.
  94     # Also created if the file needs rotation, or if forced.
  95     try:
  96         im = Image.open(filename)
  97     except IOError:
  98         raise BadMediaFail()
  99     if force \
 100         or im.size[0] > new_size[0]\
 101         or im.size[1] > new_size[1]\
 102         or exif_image_needs_rotation(exif_tags):
 103         resize_image(
 104             proc_state, im, unicode(keyname), target_name,
 105             new_size,
 106             exif_tags, conversions_subdir)
 107
 108
 109 SUPPORTED_FILETYPES = ['png', 'gif', 'jpg', 'jpeg', 'tiff']
 110
 111
 112 def sniff_handler(media_file, **kw):
 113     _log.info('Sniffing {0}'.format(MEDIA_TYPE))
 114     if kw.get('media') is not None:  # That's a double negative!
 115         name, ext = os.path.splitext(kw['media'].filename)
 116         clean_ext = ext[1:].lower()  # Strip the . from ext and make lowercase
 117
 118         if clean_ext in SUPPORTED_FILETYPES:
 119             _log.info('Found file extension in supported filetypes')
 120             return MEDIA_TYPE
 121         else:
 122             _log.debug('Media present, extension not found in {0}'.format(
 123                     SUPPORTED_FILETYPES))
 124     else:
 125         _log.warning('Need additional information (keyword argument \'media\')'
 126                      ' to be able to handle sniffing')
 127
 128     return None
 129
 130
 131 class ProcessImage(object):
 132     """Code to process an image. Will be run by celery.
 133
 134     A Workbench() represents a local tempory dir. It is automatically
 135     cleaned up when this function exits.
 136     """
 137     def init(self, proc_state):
 138         self.proc_state = proc_state
 139         self.entry = proc_state.entry
 140         self.workbench = proc_state.workbench
 141
 142         # Conversions subdirectory to avoid collisions
 143         self.conversions_subdir = os.path.join(
 144             self.workbench.dir, 'convirsions')
 145
 146         self.orig_filename = proc_state.get_orig_filename()
 147         self.name_builder = FilenameBuilder(self.orig_filename)
 148
 149         # Exif extraction
 150         self.exif_tags = extract_exif(self.orig_filename)
 151
 152         os.mkdir(self.conversions_subdir)
 153
 154     def reprocess_action(self, args):
 155         """
 156         List the available actions for media in a given state
 157         """
 158         if args[0].state == 'processed':
 159             print _('\n Available reprocessing actions for processed images:'
 160                     '\n \t --resize: thumb or medium'
 161                     '\n Options:'
 162                     '\n \t --size: max_width max_height (defaults to'
 163                     'config specs)')
 164             return True
 165
 166     def _parser(self, args):
 167         """
 168         Parses the unknown args from the gmg parser
 169         """
 170         parser = argparse.ArgumentParser()
 171         parser.add_argument(
 172             '--resize',
 173             choices=['thumb', 'medium'])
 174         parser.add_argument(
 175             '--size',
 176             nargs=2,
 177             type=int)
 178         parser.add_argument(
 179             '--initial_processing',
 180             action='store_true')
 181
 182         return parser.parse_args(args[1])
 183
 184     def _check_eligible(self, entry_args, reprocess_args):
 185         """
 186         Check to see if we can actually process the given media as requested
 187         """
 188
 189         if entry_args.state == 'processed':
 190             if reprocess_args.initial_processing:
 191                 raise Exception(_('You can not run --initial_processing on'
 192                                   ' media that has already been processed.'))
 193
 194         if entry_args.state == 'failed':
 195             if reprocess_args.resize:
 196                 raise Exception(_('You can not run --resize on media that has'
 197                                   ' not been processed.'))
 198             if reprocess_args.size:
 199                 _log.warn('With --initial_processing, the --size flag will be'
 200                           ' ignored.')
 201
 202         if entry_args.state == 'processing':
 203             raise Exception(_('We currently do not support reprocessing on'
 204                               ' media that is in the "processing" state.'))
 205
 206     def initial_processing(self):
 207         # Is there any GPS data
 208         gps_data = get_gps_data(self.exif_tags)
 209
 210          # Always create a small thumbnail
 211         resize_tool(self.proc_state, True, 'thumb', self.orig_filename,
 212                     self.name_builder.fill('{basename}.thumbnail{ext}'),
 213                     self.conversions_subdir, self.exif_tags)
 214
 215         # Possibly create a medium
 216         resize_tool(self.proc_state, False, 'medium', self.orig_filename,
 217                     self.name_builder.fill('{basename}.medium{ext}'),
 218                     self.conversions_subdir, self.exif_tags)
 219
 220         # Copy our queued local workbench to its final destination
 221         self.proc_state.copy_original(self.name_builder.fill('{basename}{ext}'))
 222
 223         # Remove queued media file from storage and database
 224         self.proc_state.delete_queue_file()
 225
 226         # Insert exif data into database
 227         exif_all = clean_exif(self.exif_tags)
 228
 229         if len(exif_all):
 230             self.entry.media_data_init(exif_all=exif_all)
 231
 232         if len(gps_data):
 233             for key in list(gps_data.keys()):
 234                 gps_data['gps_' + key] = gps_data.pop(key)
 235             self.entry.media_data_init(**gps_data)
 236
 237     def reprocess(self, reprocess_info):
 238         """
 239         This function actually does the reprocessing when called by
 240         ProcessMedia in gmg/processing/task.py
 241         """
 242         new_size = None
 243
 244         # Did they specify a size? They must specify either both or none, so
 245         # we only need to check if one is present
 246         if reprocess_info.get('max_width'):
 247             max_width = reprocess_info['max_width']
 248             max_height = reprocess_info['max_height']
 249
 250             new_size = (max_width, max_height)
 251
 252         resize_tool(self.proc_state, False, reprocess_info['resize'],
 253                     self.name_builder.fill('{basename}.medium{ext}'),
 254                     self.conversions_subdir, self.exif_tags, new_size)
 255
 256     def media_reprocess(self, args):
 257         """
 258         This function handles the all of the reprocessing logic, before calling
 259         gmg/submit/lib/run_process_media
 260         """
 261         reprocess_args = self._parser(args)
 262         entry_args = args[0]
 263
 264         # Can we actually process the given media as requested?
 265         self._check_eligible(entry_args, reprocess_args)
 266
 267         # Do we want to re-try initial processing?
 268         if reprocess_args.initial_processing:
 269             for id in entry_args.media_id:
 270                 entry = MediaEntry.query.filter_by(id=id).first()
 271                 run_process_media(entry)
 272
 273         # Are we wanting to resize the thumbnail or medium?
 274         elif reprocess_args.resize:
 275
 276             # reprocess all given media entries
 277             for id in entry_args.media_id:
 278                 entry = MediaEntry.query.filter_by(id=id).first()
 279
 280                 # For now we can only reprocess with the original file
 281                 if not entry.media_files.get('original'):
 282                     raise Exception(_('The original file for this media entry'
 283                                       ' does not exist.'))
 284
 285                 reprocess_info = self._get_reprocess_info(reprocess_args)
 286                 run_process_media(entry, reprocess_info=reprocess_info)
 287
 288         # If we are here, they forgot to tell us how to reprocess
 289         else:
 290             _log.warn('You must set either --resize or --initial_processing'
 291                       ' flag to reprocess an image.')
 292
 293     def _get_reprocess_info(self, args):
 294         """ Returns a dict with the info needed for reprocessing"""
 295         reprocess_info = {'resize': args.resize}
 296
 297         if args.size:
 298             reprocess_info['max_width'] = args.size[0]
 299             reprocess_info['max_height'] = args.size[1]
 300
 301         return reprocess_info
 302
 303 if __name__ == '__main__':
 304     import sys
 305     import pprint
 306
 307     pp = pprint.PrettyPrinter()
 308
 309     result = extract_exif(sys.argv[1])
 310     gps = get_gps_data(result)
 311     clean = clean_exif(result)
 312     useful = get_useful(clean)
 313
 314     print pp.pprint(
 315         clean)