Preserve attribute order when parsing

Author: kovidgoyal

html5lib/constants.py

@@ -431,6 +431,73 @@
     (namespaces["mathml"], "mtext")
 ])
 
+adjustSVGAttributes = {
+    "attributename": "attributeName",
+    "attributetype": "attributeType",
+    "basefrequency": "baseFrequency",
+    "baseprofile": "baseProfile",
+    "calcmode": "calcMode",
+    "clippathunits": "clipPathUnits",
+    "contentscripttype": "contentScriptType",
+    "contentstyletype": "contentStyleType",
+    "diffuseconstant": "diffuseConstant",
+    "edgemode": "edgeMode",
+    "externalresourcesrequired": "externalResourcesRequired",
+    "filterres": "filterRes",
+    "filterunits": "filterUnits",
+    "glyphref": "glyphRef",
+    "gradienttransform": "gradientTransform",
+    "gradientunits": "gradientUnits",
+    "kernelmatrix": "kernelMatrix",
+    "kernelunitlength": "kernelUnitLength",
+    "keypoints": "keyPoints",
+    "keysplines": "keySplines",
+    "keytimes": "keyTimes",
+    "lengthadjust": "lengthAdjust",
+    "limitingconeangle": "limitingConeAngle",
+    "markerheight": "markerHeight",
+    "markerunits": "markerUnits",
+    "markerwidth": "markerWidth",
+    "maskcontentunits": "maskContentUnits",
+    "maskunits": "maskUnits",
+    "numoctaves": "numOctaves",
+    "pathlength": "pathLength",
+    "patterncontentunits": "patternContentUnits",
+    "patterntransform": "patternTransform",
+    "patternunits": "patternUnits",
+    "pointsatx": "pointsAtX",
+    "pointsaty": "pointsAtY",
+    "pointsatz": "pointsAtZ",
+    "preservealpha": "preserveAlpha",
+    "preserveaspectratio": "preserveAspectRatio",
+    "primitiveunits": "primitiveUnits",
+    "refx": "refX",
+    "refy": "refY",
+    "repeatcount": "repeatCount",
+    "repeatdur": "repeatDur",
+    "requiredextensions": "requiredExtensions",
+    "requiredfeatures": "requiredFeatures",
+    "specularconstant": "specularConstant",
+    "specularexponent": "specularExponent",
+    "spreadmethod": "spreadMethod",
+    "startoffset": "startOffset",
+    "stddeviation": "stdDeviation",
+    "stitchtiles": "stitchTiles",
+    "surfacescale": "surfaceScale",
+    "systemlanguage": "systemLanguage",
+    "tablevalues": "tableValues",
+    "targetx": "targetX",
+    "targety": "targetY",
+    "textlength": "textLength",
+    "viewbox": "viewBox",
+    "viewtarget": "viewTarget",
+    "xchannelselector": "xChannelSelector",
+    "ychannelselector": "yChannelSelector",
+    "zoomandpan": "zoomAndPan"
+}
+
+adjustMathMLAttributes = {"definitionurl": "definitionURL"}
+
 adjustForeignAttributes = {
     "xlink:actuate": ("xlink", "actuate", namespaces["xlink"]),
     "xlink:arcrole": ("xlink", "arcrole", namespaces["xlink"]),

html5lib/html5parser.py

@@ -2,6 +2,7 @@
 from six import with_metaclass
 
 import types
+from collections import OrderedDict
 
 from . import inputstream
 from . import tokenizer
@@ -17,6 +18,7 @@
     namespaces,
     htmlIntegrationPointElements, mathmlTextIntegrationPointElements,
     adjustForeignAttributes as adjustForeignAttributesMap,
+    adjustMathMLAttributes, adjustSVGAttributes,
     E,
     ReparseException
 )
@@ -263,96 +265,18 @@ def normalizeToken(self, token):
         """ HTML5 specific normalizations to the token stream """
 
         if token["type"] == tokenTypes["StartTag"]:
-            token["data"] = dict(token["data"][::-1])
+            token["data"] = OrderedDict(token['data'])
 
         return token
 
     def adjustMathMLAttributes(self, token):
-        replacements = {"definitionurl": "definitionURL"}
-        for k, v in replacements.items():
-            if k in token["data"]:
-                token["data"][v] = token["data"][k]
-                del token["data"][k]
+        adjust_attributes(token, adjustMathMLAttributes)
 
     def adjustSVGAttributes(self, token):
-        replacements = {
-            "attributename": "attributeName",
-            "attributetype": "attributeType",
-            "basefrequency": "baseFrequency",
-            "baseprofile": "baseProfile",
-            "calcmode": "calcMode",
-            "clippathunits": "clipPathUnits",
-            "contentscripttype": "contentScriptType",
-            "contentstyletype": "contentStyleType",
-            "diffuseconstant": "diffuseConstant",
-            "edgemode": "edgeMode",
-            "externalresourcesrequired": "externalResourcesRequired",
-            "filterres": "filterRes",
-            "filterunits": "filterUnits",
-            "glyphref": "glyphRef",
-            "gradienttransform": "gradientTransform",
-            "gradientunits": "gradientUnits",
-            "kernelmatrix": "kernelMatrix",
-            "kernelunitlength": "kernelUnitLength",
-            "keypoints": "keyPoints",
-            "keysplines": "keySplines",
-            "keytimes": "keyTimes",
-            "lengthadjust": "lengthAdjust",
-            "limitingconeangle": "limitingConeAngle",
-            "markerheight": "markerHeight",
-            "markerunits": "markerUnits",
-            "markerwidth": "markerWidth",
-            "maskcontentunits": "maskContentUnits",
-            "maskunits": "maskUnits",
-            "numoctaves": "numOctaves",
-            "pathlength": "pathLength",
-            "patterncontentunits": "patternContentUnits",
-            "patterntransform": "patternTransform",
-            "patternunits": "patternUnits",
-            "pointsatx": "pointsAtX",
-            "pointsaty": "pointsAtY",
-            "pointsatz": "pointsAtZ",
-            "preservealpha": "preserveAlpha",
-            "preserveaspectratio": "preserveAspectRatio",
-            "primitiveunits": "primitiveUnits",
-            "refx": "refX",
-            "refy": "refY",
-            "repeatcount": "repeatCount",
-            "repeatdur": "repeatDur",
-            "requiredextensions": "requiredExtensions",
-            "requiredfeatures": "requiredFeatures",
-            "specularconstant": "specularConstant",
-            "specularexponent": "specularExponent",
-            "spreadmethod": "spreadMethod",
-            "startoffset": "startOffset",
-            "stddeviation": "stdDeviation",
-            "stitchtiles": "stitchTiles",
-            "surfacescale": "surfaceScale",
-            "systemlanguage": "systemLanguage",
-            "tablevalues": "tableValues",
-            "targetx": "targetX",
-            "targety": "targetY",
-            "textlength": "textLength",
-            "viewbox": "viewBox",
-            "viewtarget": "viewTarget",
-            "xchannelselector": "xChannelSelector",
-            "ychannelselector": "yChannelSelector",
-            "zoomandpan": "zoomAndPan"
-        }
-        for originalName in list(token["data"].keys()):
-            if originalName in replacements:
-                svgName = replacements[originalName]
-                token["data"][svgName] = token["data"][originalName]
-                del token["data"][originalName]
+        adjust_attributes(token, adjustSVGAttributes)
 
     def adjustForeignAttributes(self, token):
-        replacements = adjustForeignAttributesMap
-
-        for originalName in token["data"].keys():
-            if originalName in replacements:
-                foreignName = replacements[originalName]
-                token["data"][foreignName] = token["data"][originalName]
-                del token["data"][originalName]
+        adjust_attributes(token, adjustForeignAttributesMap)
 
     def reparseTokenNormal(self, token):
         self.parser.phase()
@@ -2702,6 +2626,12 @@ def processEndTag(self, token):
     }
 
 
+def adjust_attributes(token, replacements):
+    if frozenset(token['data']) & frozenset(replacements):
+        token['data'] = OrderedDict(
+            (replacements.get(k, k), v) for k, v in token['data'].iteritems())
+
+
 def impliedTagToken(name, type="EndTag", attributes=None,
                     selfClosing=False):
     if attributes is None: