Add support for live photos

Look for MediaGroupUUID and ContentIdentifier EXIF tags. These are
present in media files which are part of live photo pairs. If found, any
future files with the same UUID are named the same as the first one and
placed in the same directory.

With this change, live photos and their associated .mov files are moved
together.

src/exif.py

@@ -28,5 +28,5 @@ def data(self):
     def get_exif_command(filename):
         # Handle all platform variations
         if sys.platform == 'win32':
-            return f'exiftool -time:all -mimetype -j "{filename}"'
-        return f'exiftool -time:all -mimetype -j {shlex.quote(filename)}'
+            return f'exiftool -MediaGroupUUID -ContentIdentifier -time:all -mimetype -j "{filename}"'
+        return f'exiftool -MediaGroupUUID -ContentIdentifier -time:all -mimetype -j {shlex.quote(filename)}'

src/phockup.py

@@ -12,6 +12,7 @@
 
 from src.date import Date
 from src.exif import Exif
+from collections import defaultdict
 
 logger = logging.getLogger('phockup')
 ignored_files = ('.DS_Store', 'Thumbs.db')
@@ -28,6 +29,7 @@ def __init__(self, input_dir, output_dir, **args):
         self.unknown_found = 0
         self.files_moved = 0
         self.files_copied = 0
+        self.content_ids = defaultdict(list)
 
         input_dir = os.path.expanduser(input_dir)
         output_dir = os.path.expanduser(output_dir)
@@ -359,16 +361,36 @@ def get_file_name_and_path(self, filename):
         if target_file_type in ['image', 'video']:
             date = Date(filename).from_exif(exif_data, self.timestamp, self.date_regex,
                                             self.date_field)
-            output = self.get_output_dir(date)
+            output_dir = self.get_output_dir(date)
             target_file_name = self.get_file_name(filename, date)
             if not self.original_filenames:
                 target_file_name = target_file_name.lower()
         else:
-            output = self.get_output_dir([])
+            output_dir = self.get_output_dir([])
             target_file_name = os.path.basename(filename)
 
-        target_file_path = os.path.sep.join([output, target_file_name])
-        return output, target_file_name, target_file_path, target_file_type, date
+        target_file_path = os.path.sep.join([output_dir, target_file_name])
+
+        if exif_data and ('ContentIdentifier' in exif_data or 'MediaGroupUUID' in exif_data):
+            uuid = exif_data['ContentIdentifier'] if 'ContentIdentifier' in exif_data else exif_data['MediaGroupUUID']
+            logger.debug(f"{filename} has media group ID: {uuid}")
+            cids = self.content_ids[uuid]
+
+            # If a file with this same ContentId UUID exists already, use its
+            # name and put it in the same directory
+            if cids:
+                logger.debug(f"Found existing files with same ID: {cids}")
+                logger.debug(f"* Previous target dir for current file:  {output_dir}")
+                logger.debug(f"* Previous target path for current file: {target_file_path}")
+                output_dir, basename = os.path.split(cids[0])
+                target_file_name = os.path.splitext(basename)[0] + os.path.splitext(filename)[1]
+                target_file_name = target_file_name.lower()
+                target_file_path = os.path.sep.join([output_dir, target_file_name])
+                logger.debug(f"* New target dir for current file:  {output_dir}")
+                logger.debug(f"* New target path for current file: {target_file_path}")
+            cids.append(target_file_path)
+
+        return output_dir, target_file_name, target_file_path, target_file_type, date
 
     def process_xmp(self, original_filename, file_name, suffix, output):
         """