A few minor fixes; fix up etree serialization a bit, allow a single parser instance to parse multiple documents

jgraham · jgraham · commit e5788d7b2d84 · 2007-01-05T21:56:15.000Z
--HG--
extra : convert_revision : svn%3Aacbfec75-9323-0410-a652-858a13e371e0/trunk%40417
diff --git a/setup.py b/setup.py
@@ -5,6 +5,7 @@
       license="MIT License",
       description='HTML parser based on the WHAT-WG Web Applications 1.0' 
                   '("HTML5") specifcation',
-      packages=['html5lib'],
-      package_dir = {'html5lib': 'src'}
+      packages=['html5lib', 'html5lib.treebuilders'],
+      package_dir = {'html5lib': 'src', 
+                     'html5lib.treebuilders': 'src/treebuilders'}
       )
diff --git a/src/__init__.py b/src/__init__.py
@@ -27,7 +27,7 @@
 
 Note: Because HTML documents support various features not in the
 default ElementTree (e.g. doctypes), we suppy our own simple
-serializer; html5lib.treebuilders.etree.write At present this does not
+serializer; html5lib.treebuilders.etree.tostring At present this does not
 have the encoding support offered by the elementtree serializer.
 
 """
diff --git a/src/parser.py b/src/parser.py
@@ -46,9 +46,9 @@ def __init__(self, strict = False, tree=simpletree.TreeBuilder):
 
         # Raise an exception on the first error encountered
         self.strict = strict
-        self.errors = []
 
         self.tree = tree()
+        self.errors = []
 
         self.phases = {
             "initial": InitialPhase(self, self.tree),
@@ -69,10 +69,6 @@ def __init__(self, strict = False, tree=simpletree.TreeBuilder):
             "afterFrameset": AfterFramesetPhase(self, self.tree),
             "trailingEnd": TrailingEndPhase(self, self.tree)
         }
-        self.phase = self.phases["initial"]
-        # We only seem to have InBodyPhase testcases where the following is
-        # relevant ... need others too
-        self.lastPhase = None
 
     def parse(self, stream, innerHTML=False):
         """Parse a HTML document into a well-formed tree
@@ -84,6 +80,12 @@ def parse(self, stream, innerHTML=False):
         """
 
         self.tree.reset()
+        self.errors = []
+
+        self.phase = self.phases["initial"]
+        # We only seem to have InBodyPhase testcases where the following is
+        # relevant ... need others too
+        self.lastPhase = None
 
         # We don't actually support innerHTML yet but this should allow
         # assertations
diff --git a/src/treebuilders/etree.py b/src/treebuilders/etree.py
@@ -151,13 +151,13 @@ def serializeElement(element, indent=0):
 
     return "\n".join(rv)
 
-def write(element):
+def tostring(element):
     """Serialize an element and its child nodes to a string"""
     rv = []
     finalText = None
     def serializeElement(element):
         if element.tag is DocumentType:
-            rv.append("<!DOCTYPE %s>\n"%(element.text,))
+            rv.append("<!DOCTYPE %s>"%(element.text,))
         elif element.tag is Document:
             if element.text:
                 rv.append(element.text)
@@ -168,8 +168,9 @@ def serializeElement(element):
                 serializeElement(child)
 
         elif element.tag is Comment:
-            rv.append("<!-- %s -->\n"%(element.text,))
+            rv.append("<!--%s-->"%(element.text,))
         else:
+            #This is assumed to be an ordinary element
             if not element.attrib:
                 rv.append("<%s>"%(element.tag,))
             else:
@@ -182,10 +183,10 @@ def serializeElement(element):
             for child in element.getchildren():
                 serializeElement(child)
 
-            rv.append("</%s>\n"%(element.tag,))
+            rv.append("</%s>"%(element.tag,))
 
         if element.tail:
-            rv.append(element.tail + "\n")
+            rv.append(element.tail)
 
     serializeElement(element)
 
