#!/usr/bin/env python
"""This is a toy example not a serious conformance checker. In
particular, it only reports parse errors when reading the document; it
does not report any of the other (many) possible types of conformance
errors that may exist in a HTML5 document"""
import sys
import urllib2
import cgi
import html5lib
htmlTemplate = u"""
%(title)s
%(title)s
%(body)s
"""
def parseDocument(document):
"""Parse the document and return a list of errors and a parse tree"""
p = html5lib.HTMLParser()
tree = p.parse(document)
return p.errors, cgi.escape(tree.printTree(), True)
def getDocument(uri):
if uri.startswith("http://") or uri.startswith("https://"):
#Why is string conversion necessary here?
document = "".join(urllib2.urlopen(uri).readlines())[:-1]
#print "<--!%s-->"%(document,)
else:
raise ValueError, "Unrecognised URI type"
return document
def writeValid(uri, treeStr):
bodyText = """%s is valid HTML5!
Parse Tree:
%s
"""%(uri, treeStr)
writeOutput(htmlTemplate%{"title":"Validation Results", "body":bodyText})
def writeInvalid(uri, treeStr, errors):
errList=[]
for pos, message in errors:
errList.append("Line %i Col %i"%pos + " " + message)
errStr = "
\n".join(errList)
bodyText = """%s is not valid HTML5
Errors:
%s
Parse Tree:
%s
"""%(uri, errStr, treeStr)
writeOutput(htmlTemplate%{"title":"Validation Results", "body":bodyText})
def writeErr(uri):
bodyText = "Failed to load URI %s
"%(uri,)
writeOutput(htmlTemplate%{"title":"Error", "body":bodyText})
def writeOutput(s):
print s.encode('utf-8')
print "Content-type: text/html"
print ""
try:
form = cgi.FieldStorage()
uri = form.getvalue("uri")
document = getDocument(uri)
except:
writeErr(uri)
sys.exit(1)
errors, tree = parseDocument(document)
if errors:
writeInvalid(uri, tree, errors)
else:
writeValid(uri, tree)