Merge pull request #5455 from radarhere/xmp

Tests/test_file_jpeg.py

@@ -828,7 +828,23 @@ class TestFileJpeg:
             xmp = im.getxmp()
 
             assert isinstance(xmp, dict)
-            assert xmp["Description"]["Version"] == "10.4"
+
+            description = xmp["xmpmeta"]["RDF"]["Description"]
+            assert description["DerivedFrom"] == {
+                "documentID": "8367D410E636EA95B7DE7EBA1C43A412",
+                "originalDocumentID": "8367D410E636EA95B7DE7EBA1C43A412",
+            }
+            assert description["Look"]["Description"]["Group"]["Alt"]["li"] == {
+                "lang": "x-default",
+                "text": "Profiles",
+            }
+            assert description["ToneCurve"]["Seq"]["li"] == ["0, 0", "255, 255"]
+
+            # Attribute
+            assert description["Version"] == "10.4"
+
+        with Image.open("Tests/images/hopper.jpg") as im:
+            assert im.getxmp() == {}
 
 
 @pytest.mark.skipif(not is_win32(), reason="Windows only")

Tests/test_file_png.py

@@ -651,6 +651,16 @@ class TestFilePng:
         with Image.open(out) as reloaded:
             assert len(reloaded.png.im_palette[1]) == 3
 
+    def test_xmp(self):
+        with Image.open("Tests/images/color_snakes.png") as im:
+            xmp = im.getxmp()
+
+            assert isinstance(xmp, dict)
+
+            description = xmp["xmpmeta"]["RDF"]["Description"]
+            assert description["PixelXDimension"] == "10"
+            assert description["subject"]["Seq"] is None
+
     def test_exif(self):
         # With an EXIF chunk
         with Image.open("Tests/images/exif.png") as im:

Tests/test_file_tiff.py

@@ -643,6 +643,16 @@ class TestFileTiff:
         with Image.open(outfile) as reloaded:
             assert "icc_profile" not in reloaded.info
 
+    def test_xmp(self):
+        with Image.open("Tests/images/lab.tif") as im:
+            xmp = im.getxmp()
+
+            assert isinstance(xmp, dict)
+
+            description = xmp["xmpmeta"]["RDF"]["Description"]
+            assert description[0]["format"] == "image/tiff"
+            assert description[3]["BitsPerSample"]["Seq"]["li"] == ["8", "8", "8"]
+
     def test_close_on_load_exclusive(self, tmp_path):
         # similar to test_fd_leak, but runs on unixlike os
         tmpfile = str(tmp_path / "temp.tif")

src/PIL/Image.py

@@ -1318,6 +1318,33 @@ class Image:
             return tuple(extrema)
         return self.im.getextrema()
 
+    def _getxmp(self, xmp_tags):
+        def get_name(tag):
+            return tag.split("}")[1]
+
+        def get_value(element):
+            value = {get_name(k): v for k, v in element.attrib.items()}
+            children = list(element)
+            if children:
+                for child in children:
+                    name = get_name(child.tag)
+                    child_value = get_value(child)
+                    if name in value:
+                        if not isinstance(value[name], list):
+                            value[name] = [value[name]]
+                        value[name].append(child_value)
+                    else:
+                        value[name] = child_value
+            elif value:
+                if element.text:
+                    value["text"] = element.text
+            else:
+                return element.text
+            return value
+
+        root = xml.etree.ElementTree.fromstring(xmp_tags)
+        return {get_name(root.tag): get_value(root)}
+
     def getexif(self):
         if self._exif is None:
             self._exif = Exif()
@@ -1338,15 +1365,15 @@ class Image:
         if 0x0112 not in self._exif:
             xmp_tags = self.info.get("XML:com.adobe.xmp")
             if xmp_tags:
-                root = xml.etree.ElementTree.fromstring(xmp_tags)
-                for elem in root.iter():
-                    if elem.tag.endswith("}Description"):
-                        orientation = elem.attrib.get(
-                            "{http://ns.adobe.com/tiff/1.0/}Orientation"
-                        )
-                        if orientation:
-                            self._exif[0x0112] = int(orientation)
-                        break
+                xmp = self._getxmp(xmp_tags)
+                if (
+                    "xmpmeta" in xmp
+                    and "RDF" in xmp["xmpmeta"]
+                    and "Description" in xmp["xmpmeta"]["RDF"]
+                ):
+                    description = xmp["xmpmeta"]["RDF"]["Description"]
+                    if "Orientation" in description:
+                        self._exif[0x0112] = int(description["Orientation"])
 
         return self._exif
 

src/PIL/JpegImagePlugin.py

@@ -40,7 +40,6 @@ import subprocess
 import sys
 import tempfile
 import warnings
-import xml.etree.ElementTree
 
 from . import Image, ImageFile, TiffImagePlugin
 from ._binary import i16be as i16
@@ -362,7 +361,6 @@ class JpegImageFile(ImageFile.ImageFile):
         self.app = {}  # compatibility
         self.applist = []
         self.icclist = []
-        self._xmp = None
 
         while True:
 
@@ -485,20 +483,12 @@ class JpegImageFile(ImageFile.ImageFile):
         :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
+                    return self._getxmp(xmp_tags)
+        return {}
 
 
 def _getexif(self):

src/PIL/PngImagePlugin.py

@@ -978,6 +978,17 @@ class PngImageFile(ImageFile.ImageFile):
 
         return super().getexif()
 
+    def getxmp(self):
+        """
+        Returns a dictionary containing the XMP tags.
+        :returns: XMP tags in a dictionary.
+        """
+        return (
+            self._getxmp(self.info["XML:com.adobe.xmp"])
+            if "XML:com.adobe.xmp" in self.info
+            else {}
+        )
+
     def _close__fp(self):
         try:
             if self.__fp != self.fp:

src/PIL/TiffImagePlugin.py

@@ -1105,6 +1105,13 @@ class TiffImageFile(ImageFile.ImageFile):
         """Return the current frame number"""
         return self.__frame
 
+    def getxmp(self):
+        """
+        Returns a dictionary containing the XMP tags.
+        :returns: XMP tags in a dictionary.
+        """
+        return self._getxmp(self.tag_v2[700]) if 700 in self.tag_v2 else {}
+
     def load(self):
         if self.tile and self.use_load_libtiff:
             return self._load_libtiff()