Add a serialization option for EXIF dictionaries

ChangeLog.rst

@@ -9,6 +9,7 @@
     * fix webp for file conversion by magick (#132) by Marcelo
     * Ignore unknown parsers, fix for #160 (#175) by Herbert Poul
     * Allow extracting thumbnails with details=False (#170)
+    * Add option to return built-in Python types (#129) by Étienne Pelletier
 
 3.0.0 — 2022-05-08
     * **BREAKING CHANGE:** Add type hints, which removes Python2 compatibility

EXIF.py

@@ -51,6 +51,10 @@ def get_args() -> argparse.Namespace:
         '-s', '--strict', action='store_true', dest='strict',
         help='Run in strict mode (stop on errors).',
     )
+    parser.add_argument(
+        '-b', '--builtin', action='store_true', dest='builtin_types',
+        help='Convert IfdTag values to built-in Python variable types',
+    )
     parser.add_argument(
         '-d', '--debug', action='store_true', dest='debug',
         help='Run in debug mode (display extra info).',
@@ -92,7 +96,8 @@ def main(args) -> None:
             details=args.detailed,
             strict=args.strict,
             debug=args.debug,
-            extract_thumbnail=args.detailed
+            extract_thumbnail=args.detailed,
+            builtin_types=args.builtin_types,
         )
 
         tag_stop = timeit.default_timer()
@@ -109,14 +114,15 @@ def main(args) -> None:
             logger.info('File has TIFF thumbnail')
             del data['TIFFThumbnail']
 
-        tag_keys = list(data.keys())
-        tag_keys.sort()
-
-        for i in tag_keys:
+        for field in sorted(data):
+            value = data[field]
             try:
-                logger.info('%s (%s): %s', i, FIELD_TYPES[data[i].field_type][2], data[i].printable)
+                if args.builtin_types:
+                    logger.info('%s (%s): %r', field, type(value).__name__, value)
+                else:
+                    logger.info('%s (%s): %s', field, FIELD_TYPES[value.field_type][2], value.printable)
             except:
-                logger.error("%s : %s", i, str(data[i]))
+                logger.error("%s: %s", field, str(value))
 
         file_stop = timeit.default_timer()
 

README.rst

@@ -79,9 +79,9 @@ You can process the tags as you wish. In particular, you can iterate through all
 
 .. code-block:: python
 
-    for tag in tags.keys():
+    for tag, value in tags.items():
         if tag not in ('JPEGThumbnail', 'TIFFThumbnail', 'Filename', 'EXIF MakerNote'):
-            print "Key: %s, value %s" % (tag, tags[tag])
+            print(f"Key: {tag}, value {value}")
 
 An ``if`` statement is used to avoid printing out a few of the tags that tend to be long or boring.
 
@@ -160,6 +160,23 @@ Pass the ``-s`` or ``--strict`` argument, or as:
 
     tags = exifread.process_file(file_handle, strict=True)
 
+Built-in Types
+==============
+
+For easier serialization and programmatic use, this option returns a dictionary with values in built-in Python types (int, float, str, bytes, list, None) instead of `IfdTag` objects.
+
+Pass the ``-b`` or ``--builtin`` argument, or as:
+
+.. code-block:: python
+
+    tags = exifread.process_file(file_handle, builtin_types=True)
+
+For direct JSON serialization, combine this option with ``details=False`` to avoid bytes in the output:
+
+.. code-block:: python
+
+    json.dumps(exifread.process_file(file_handle, details=False, builtin_types=True))
+
 Usage Example
 =============
 
@@ -171,14 +188,14 @@ This example shows how to use the library to correct the orientation of an image
     import exifread
     from PIL import Image
     import logging
-    
+
     def _read_img_and_correct_exif_orientation(path):
         im = Image.open(path)
         tags = {}
         with open(path, "rb") as file_handle:
             tags = exifread.process_file(file_handle, details=False)
 
-        if "Image Orientation" in tags.keys():
+        if "Image Orientation" in tags:
             orientation = tags["Image Orientation"]
             logging.basicConfig(level=logging.DEBUG)
             logging.debug("Orientation: %s (%s)", orientation, orientation.values)

exifread/__init__.py

@@ -1,5 +1,6 @@
 """
-Read Exif metadata from tiff and jpeg files.
+Read Exif metadata from image files
+Supported formats: TIFF, JPEG, PNG, Webp, HEIC
 """
 
 import struct
@@ -12,6 +13,7 @@
 from exifread.heic import HEICExifFinder
 from exifread.jpeg import find_jpeg_exif
 from exifread.exceptions import InvalidExif, ExifNotFound
+from exifread.serialize import convert_types
 
 __version__ = '3.1.0'
 
@@ -122,7 +124,8 @@ def _determine_type(fh: BinaryIO) -> tuple:
 
 def process_file(fh: BinaryIO, stop_tag=DEFAULT_STOP_TAG,
                  details=True, strict=False, debug=False,
-                 truncate_tags=True, auto_seek=True, extract_thumbnail=True):
+                 truncate_tags=True, auto_seek=True,
+                 extract_thumbnail=True, builtin_types=False) -> dict:
     """
     Process an image file (expects an open file object).
 
@@ -195,4 +198,8 @@ def process_file(fh: BinaryIO, stop_tag=DEFAULT_STOP_TAG,
             xmp_bytes = _get_xmp(fh)
         if xmp_bytes:
             hdr.parse_xmp(xmp_bytes)
+
+    if builtin_types:
+        return convert_types(hdr.tags)
+
     return hdr.tags

exifread/classes.py

@@ -15,7 +15,7 @@ class IfdTag:
     """
 
     def __init__(self, printable: str, tag: int, field_type: int, values,
-                 field_offset: int, field_length: int):
+                 field_offset: int, field_length: int, prefer_printable: bool = True):
         # printable version of data
         self.printable = printable
         # tag ID number
@@ -29,6 +29,8 @@ def __init__(self, printable: str, tag: int, field_type: int, values,
         # either string, bytes or list of data items
         # TODO: sort out this type mess!
         self.values = values
+        # indication if printable version should be used upon serialization
+        self.prefer_printable = prefer_printable
 
     def __str__(self) -> str:
         return self.printable
@@ -263,10 +265,15 @@ def _process_tag(self, ifd, ifd_name: str, tag_entry, entry, tag: int, tag_name,
                 printable = str(values[0:-1])
         else:
             printable = str(values)
+
+        prefer_printable = False
+
         # compute printable version of values
         if tag_entry:
             # optional 2nd tag element is present
             if len(tag_entry) != 1:
+                prefer_printable = True
+
                 if callable(tag_entry[1]):
                     # call mapping function
                     printable = tag_entry[1](values)
@@ -284,7 +291,7 @@ def _process_tag(self, ifd, ifd_name: str, tag_entry, entry, tag: int, tag_name,
                         printable += tag_entry[1].get(val, repr(val))
 
         self.tags[ifd_name + ' ' + tag_name] = IfdTag(
-            printable, tag, field_type, values, field_offset, count * type_length
+            printable, tag, field_type, values, field_offset, count * type_length, prefer_printable
         )
         tag_value = repr(self.tags[ifd_name + ' ' + tag_name])
         logger.debug(' %s: %s', tag_name, tag_value)

exifread/exif_log.py

@@ -28,7 +28,7 @@ def setup_logger(debug, color):
     else:
         log_level = logging.INFO
 
-    logger = logging.getLogger("exifread")
+    logger = get_logger()
     stream = Handler(log_level, debug, color)
     logger.addHandler(stream)
     logger.setLevel(log_level)