Prevent removal of metadata elements if they have only text nodes. Also added some unit tests.

scour.py

@@ -51,7 +51,10 @@
 #   when being run as main (for unit testing)
 # - Removed duplicate gradient stops
 # - Convert all colors to #RRGGBB format
-# - 
+# - prevent metadata from being removed if they contain only text nodes
+# - rework command-line argument processing so that options are configurable
+# - remove unreferenced patterns? https://bugs.edge.launchpad.net/ubuntu/+source/human-icon-theme/+bug/361667/
+
 
 # necessary to get true division
 from __future__ import division
@@ -597,7 +600,7 @@ def scourString(in_string):
 			removeElem = not elem.hasChildNodes()
 			if removeElem == False :
 				for child in elem.childNodes :
-					if child.nodeType in [1, 4, 8] :
+					if child.nodeType in [1, 3, 4, 8] :
 						break
 				else:
 					removeElem = True

testscour.py

@@ -95,6 +95,30 @@ class NoAdobeXPathElements(unittest.TestCase):
 			lambda e: e.namespaceURI != 'http://ns.adobe.com/XPath/1.0/'), False,
 			'Found Adobe XPath elements' )
 
+#class NoInkscapeAttributes(unittest.TestCase):
+#	def runTest(self):
+#		self.assertNotEquals(walkTree(scour.scourXmlFile('unittests/inkscape.svg').documentElement,
+#			lambda e: for a in e.attributes: a.namespaceURI 
+#			False, 
+#			'Found Inkscape attributes')
+
+class DoNotRemoveMetadataWithOnlyText(unittest.TestCase):
+	def runTest(self):
+		doc = scour.scourXmlFile('unittests/metadata-with-text.svg')
+		self.assertEquals(len(doc.getElementsByTagNameNS('http://www.w3.org/2000/svg', 'metadata')), 1,
+			'Removed metadata element with only text child' )
+
+class RemoveEmptyMetadataElement(unittest.TestCase):
+	def runTest(self):
+		doc = scour.scourXmlFile('unittests/empty-metadata.svg')
+		self.assertEquals(len(doc.getElementsByTagNameNS('http://www.w3.org/2000/svg', 'metadata')), 0,
+			'Did not remove empty metadata element' )
+
+class RemoveEmptyGElements(unittest.TestCase):
+	def runTest(self):
+		doc = scour.scourXmlFile('unittests/empty-g.svg')
+		self.assertEquals(len(doc.getElementsByTagNameNS('http://www.w3.org/2000/svg', 'g')), 1,
+			'Did not remove empty g element' )
 
 if __name__ == '__main__':
     unittest.main()

unittests/adobe.svg

@@ -7,6 +7,7 @@
      xmlns:ir="http://ns.adobe.com/ImageReplacement/1.0/"     
      xmlns:custom="http://ns.adobe.com/GenericCustomNamespace/1.0/"
      xmlns:xpath="http://ns.adobe.com/XPath/1.0/"
+     xmlns:ok="A.namespace.we.want.left.in"
      i:viewOrigin="190.2959 599.1841" i:rulerOrigin="0 0" i:pageBounds="0 792 612 0">
 <x:foo>bar</x:foo>
 <i:foo>bar</i:foo>
@@ -26,4 +27,18 @@
 	<slices/>
 	<sliceSourceBounds y="191.664" x="190.296" width="225.72" height="407.52" bottomLeftOrigin="true"/>
 </sfw>
+<rect width="300" height="200" fill="green"
+	x:baz="1"
+	i:baz="1"
+	graph:baz="1"
+	a:baz="1"
+	f:baz="1"
+	ir:baz="1"
+	custom:baz='1'
+	xpath:baz="1"
+	xmlns:v="http://ns.adobe.Variables/1.0/"
+	v:baz="1"
+	xmlns:sfw="http://ns.adobe.com/SaveForWeb/1.0/"
+	sfw:baz="1"
+	ok:baz="1" />
 </svg>

unittests/empty-g.svg

@@ -0,0 +1,6 @@
+<svg xmlns="http://www.w3.org/2000/svg">
+  <g/>
+  <g>
+    <rect width="300" height="200" fill="green" />
+  </g>
+</svg>

unittests/empty-metadata.svg

@@ -0,0 +1,3 @@
+<svg xmlns="http://www.w3.org/2000/svg">
+  <metadata></metadata>
+</svg>

unittests/inkscape.svg

@@ -1,5 +1,5 @@
-<svg xmlns="http://www.w3.org/2000/svg" xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:inkscape="dihttp://www.inkscape.org/namespaces/inkscape"
 	inkscape:version="0.46" version="1.0">
 <inkscape:perspective inkscape:vp_x="0 : 526.18109 : 1" inkscape:vp_y="0 : 1000 : 0" inkscape:vp_z="744.09448 : 526.18109 : 1" inkscape:persp3d-origin="372.04724 : 350.78739 : 1" id="perspective3104"/>
-<rect width="300" height="200" fill="green" />
+<rect width="300" height="200" fill="green" inkscape:collect="always"/>
 </svg>

unittests/metadata-with-text.svg

@@ -0,0 +1,3 @@
+<svg xmlns="http://www.w3.org/2000/svg">
+  <metadata>This is a metadata element with only text node children</metadata>
+</svg>