Merge pull request #5376 from radarhere/xmp

Tests/test_file_jpeg.py

@@ -805,6 +805,13 @@ def read(n=-1):
         # Assert the entire file has not been read
         assert 0 < buffer.max_pos < size
 
+    def test_getxmp(self):
+        with Image.open("Tests/images/xmp_test.jpg") as im:
+            xmp = im.getxmp()
+
+            assert isinstance(xmp, dict)
+            assert xmp["Description"]["Version"] == "10.4"
+
 
 @pytest.mark.skipif(not is_win32(), reason="Windows only")
 @skip_unless_feature("jpg")

docs/releasenotes/8.2.0.rst

@@ -112,6 +112,20 @@ separate histograms for each color channel, changing the tone of the image. The
 ``preserve_tone`` argument keeps the tone unchanged by using one luminance histogram
 for all channels.
 
+getxmp() for JPEG images
+^^^^^^^^^^^^^^^^^^^^^^^^
+
+A new method has been added to return
+`XMP data <https://en.wikipedia.org/wiki/Extensible_Metadata_Platform>`_ for JPEG
+images. It reads the XML data into a dictionary of names and values.
+
+For example::
+
+    >>> from PIL import Image
+    >>> with Image.open("Tests/images/xmp_test.jpg") as im:
+    >>>     print(im.getxmp())
+    {'RDF': {}, 'Description': {'Version': '10.4', 'ProcessVersion': '10.0', ...}, ...}
+
 Security
 ========
 

src/PIL/Image.py

@@ -528,7 +528,6 @@ def __init__(self):
         self.readonly = 0
         self.pyaccess = None
         self._exif = None
-        self._xmp = None
 
     def __getattr__(self, name):
         if name == "category":
@@ -1340,27 +1339,6 @@ def getexif(self):
 
         return self._exif
 
-    def getxmp(self):
-        """
-        Returns a dictionary containing the xmp tags for a given image.
-        :returns: XMP tags in a dictionary.
-        """
-
-        if self._xmp is None:
-            self._xmp = {}
-
-        for segment, content in self.applist:
-            if segment == "APP1":
-                marker, xmp_tags = content.rsplit(b"\x00", 1)
-                if marker == b"http://ns.adobe.com/xap/1.0/":
-                    root = xml.etree.ElementTree.fromstring(xmp_tags)
-                    for element in root.findall(".//"):
-                        self._xmp[element.tag.split("}")[1]] = {
-                            child.split("}")[1]: value
-                            for child, value in element.attrib.items()
-                        }
-        return self._xmp
-
     def getim(self):
         """
         Returns a capsule that points to the internal image memory.

src/PIL/JpegImagePlugin.py

@@ -39,6 +39,7 @@
 import sys
 import tempfile
 import warnings
+import xml.etree.ElementTree
 
 from . import Image, ImageFile, TiffImagePlugin
 from ._binary import i16be as i16
@@ -358,6 +359,7 @@ def _open(self):
         self.app = {}  # compatibility
         self.applist = []
         self.icclist = []
+        self._xmp = None
 
         while True:
 
@@ -474,6 +476,27 @@ def _getexif(self):
     def _getmp(self):
         return _getmp(self)
 
+    def getxmp(self):
+        """
+        Returns a dictionary containing the XMP tags.
+        :returns: XMP tags in a dictionary.
+        """
+
+        if self._xmp is None:
+            self._xmp = {}
+
+        for segment, content in self.applist:
+            if segment == "APP1":
+                marker, xmp_tags = content.rsplit(b"\x00", 1)
+                if marker == b"http://ns.adobe.com/xap/1.0/":
+                    root = xml.etree.ElementTree.fromstring(xmp_tags)
+                    for element in root.findall(".//"):
+                        self._xmp[element.tag.split("}")[1]] = {
+                            child.split("}")[1]: value
+                            for child, value in element.attrib.items()
+                        }
+        return self._xmp
+
 
 def _getexif(self):
     if "exif" not in self.info: