Initial elementtree support (not yet unit-tested), and changes to parse.py to support multiple treebuilders

jgraham · jgraham · commit 36517ca59da6 · 2006-12-30T21:12:11.000Z
--HG--
extra : convert_revision : svn%3Aacbfec75-9323-0410-a652-858a13e371e0/trunk%40387
diff --git a/parse.py b/parse.py
@@ -1,3 +1,4 @@
+#!/usr/bin/env python
 """usage: %prog [options] filename
 
 Parse a document to a DOMlite tree, with optional profiling
@@ -7,7 +8,7 @@
 import os
 from optparse import OptionParser
 
-from src import parser
+from src import parser, treebuilders
 
 def convertTreeDump(treedump):
     """convert the output of str(document) to something more readable
@@ -25,10 +26,23 @@ def parse():
     optParser = getOptParser()
     opts,args = optParser.parse_args()
 
-    p = parser.HTMLParser()
-    # Don't try to open args[0]. It should be possible to pass a string or file
-    # reference. HTMLInputStream takes care of the difference.
-    f = args[0]
+    try:
+        f = args[-1]
+    except IndexError:
+        print "No filename provided. Use -h for help"
+        sys.exit(1)
+    if hasattr(opts, "treebuilder"):
+        try:
+            #This isn't a great way to do this
+            exec("import treebuilders.%s")%opts.treebuilder.split(".")[0]
+            treebuilder = eval("treebuilders.%s"%opts.treebuilder)
+        except NameError:
+            print "Treebuilder %s not found"%opts.treebuilder 
+            raise
+    else:
+        treebuilder = treebuilders.DOMlite.TreeBuilder
+    p = parser.HTMLParser(tree=treebuilder)
+
     if opts.profile:
         import hotshot
         import hotshot.stats
@@ -45,25 +59,28 @@ def parse():
         t0 = time.time()
         document = p.parse(f)
         t1 = time.time()
-        print convertTreeDump(document.printTree())
+        print p.tree.testSerializer(document)
         t2 = time.time()
         print "\n\nRun took: %fs (plus %fs to print the output)"%(t1-t0, t2-t1)
     else:
         document = p.parse(f)
-        print convertTreeDump(document.printTree())
+        print p.tree.testSerializer(document)
         print "\nParse errors:\n" + "\n".join(p.errors)
 
 def getOptParser():
     parser = OptionParser(usage=__doc__)
 
     parser.add_option("-p", "--profile", action="store_true", default=False,
-                      dest="profile", help="Use the hotdhot profiler to "
+                      dest="profile", help="Use the hotshot profiler to "
                       "produce a detailed log of the run")
 
     parser.add_option("-t", "--time",
                       action="store_true", default=False, dest="time",
                       help="Time the run using time.time (may not be accurate on all platforms, especially for short runs)")
 
+    parser.add_option("-b", "--treebuilder", action="store", type="string",
+                      dest="treebuilder")
+
     return parser
 
 if __name__ == "__main__":
diff --git a/src/treebuilders/DOMlite.py b/src/treebuilders/DOMlite.py
@@ -119,3 +119,5 @@ class TreeBuilder(base.TreeBuilder):
     elementClass = Element
     commentClass = CommentNode
     
+    def testSerializer(self, node):
+        node.printTree()
diff --git a/src/treebuilders/etree.py b/src/treebuilders/etree.py
@@ -7,9 +7,15 @@
 
 class Element(object):
     def __init__(self, name):
-        self._element = ElementTree.Element()
+        self._element = ElementTree.Element(name)
         self.name = name
         self.parent = None
+        self._flags = []
+
+        #Set the element text and tail to the empty string rather than None
+        #XXX - is this desirable or should we do it on a case by case basis?
+        self._element.text = ""
+        self._element.tail = ""
 
     def _setName(self, name):
         self._element.tag = name
@@ -28,7 +34,7 @@ def _setAttributes(self, attributes):
         for key in self._element.attrib.keys():
             del self._element.attrib[key]
         for key, value in attributes.iteritems():
-            self._elements.set(key, value)
+            self._element.set(key, value)
 
     attributes = property(_getAttributes, _setAttributes)
 
@@ -37,23 +43,23 @@ def appendChild(self, node):
         node.parent = self
 
     def insertBefore(self, node, refNode):
-        index = self._element.getChildren().index(refNode._element)
+        index = self._element.getchildren().index(refNode._element)
         self._element.insert(index, node._element)
         node.parent = self
 
     def removeChild(self, node):
         self._element.remove(node._element)
         node.parent=None
 
-    def insertText(self, text, insertBefore):
+    def insertText(self, data, insertBefore=None):
         if not(len(self._element)):
             self._element.text += data
         elif insertBefore is None:
             #Insert the text as the tail of the last child element
             self._element[-1].tail += data
         else:
             #Insert the text before the specified node
-            children = self._element.getChildren()
+            children = self._element.getchildren()
             index = children.index(insertBefore._element)
             if index > 0:
                 self._element[index-1].tail += data
@@ -83,11 +89,18 @@ def _setData(self, value):
 class DocumentType(Element):
     pass
 
+class Document(Element):
+    def __init__(self):
+        Element.__init__(self, "")
+
 class TreeBuilder(base.TreeBuilder):
-    documentClass = Element
+    documentClass = Document
     doctypeClass = DocumentType
     elementClass = Element
     commentClass = Comment
 
+    def testSerializer(self, element):
+        ElementTree.tostring(element)
+
     def getDocument(self):
-        return self.document._element
+        return self.document.getchildren()[0]._element
