'''HTML parser, built from standard lib's sgmllib.

Tag names are normalised to upper case, the usual HTML fashion.
'''

from sgmllib import SGMLParser
from xml.dom import core
from xml.dom.builder import Builder
import string

class HtmlBuilder(SGMLParser, Builder):
	from htmlentitydefs import entitydefs
	
	def __init__(self):
		SGMLParser.__init__(self)
		Builder.__init__(self)

		self.empties = [
			'META', 'BASE', 'LINK', 
			'HR', 'BR',
			'IMG', 'PARAM',
			'INPUT', 'OPTION', 'ISINDEX'
		]
		list = ('OL', 'UL', 'DL')
		heading = ('H1', 'H2', 'H3', 'H4', 'H5', 'H6')
		blocks = ('P', 'ADDRESS', 'BLOCKQUOTE', 'FORM', 'TABLE', 'PRE') + \
			heading # + list
		self.infer_ends = {
			'P': blocks,

			'LI': ('LI',),
			'DT': ('DT',),
			'DD': ('DT', 'DD'),

			'TR': ('TR',), 
			'TH': ('TH', 'TD', 'TR'),
			'TD': ('TH', 'TD', 'TR'),
		}

	
	def unknown_starttag(self, tag, attrs):
		tag = string.upper(tag)
		#print 'starting', tag
		attributes = {}
		for k, v in attrs:
			attributes[string.upper(k)] = v

		#print self.stack
		while self.stack:
			if self.infer_ends.has_key(self.stack[-1]): 
				if tag in self.infer_ends[self.stack[-1]]:
					#print tag, 'ending', self.stack[-1]
					Builder.endElement(self, self.stack[-1])
					del self.stack[-1]
					#print self.stack
				else:
					break
			else:
				break
		#print self.stack

		Builder.startElement(self, tag, attributes)
		if not tag in self.empties:
			self.stack.append(tag)
		else:
			Builder.endElement(self, tag)


	def unknown_endtag(self, tag):
		tag = string.upper(tag)
		#print 'ending', tag

		while self.stack:
			if tag in self.empties:
				continue
			start_tag = self.stack[-1]
			del self.stack[-1]
			Builder.endElement(self, start_tag)
			if start_tag == tag:
				break

	def handle_data(self, s):
		#print `s`
		Builder.text(self, s)


# Test.
if __name__ == '__main__':
	import sys
	b = HtmlBuilder()
	b.feed(open(sys.argv[1]).read())
	b.close()
#	print b.document
#	print b.document.documentElement

	from writer import HtmlLineariser
	w = HtmlLineariser()
	print w.linearise(b.document)