1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
|
# -*- coding: utf-8 -*-
from HTMLParser import HTMLParser, HTMLParseError
from htmlentitydefs import name2codepoint
import sys
import re
import cgi
# Normalization code, adapted from
# https://github.com/karlcow/markdown-testsuite/
significant_attrs = ["alt", "href", "src", "title"]
whitespace_re = re.compile('\s+')
class MyHTMLParser(HTMLParser):
def __init__(self):
HTMLParser.__init__(self)
self.last = "starttag"
self.in_pre = False
self.output = u""
self.last_tag = ""
def handle_data(self, data):
after_tag = self.last == "endtag" or self.last == "starttag"
after_block_tag = after_tag and self.is_block_tag(self.last_tag)
if after_tag and self.last_tag == "br":
data = data.lstrip('\n')
data = whitespace_re.sub(' ', data)
if after_block_tag and not self.in_pre:
if self.last == "starttag":
data = data.lstrip()
elif self.last == "endtag":
data = data.strip()
self.output += data
self.last = "data"
def handle_endtag(self, tag):
if tag == "pre":
self.in_pre = False
elif self.is_block_tag(tag):
self.output = self.output.rstrip()
self.output += "</" + tag + ">"
self.last_tag = tag
self.last = "endtag"
def handle_starttag(self, tag, attrs):
if tag == "pre":
self.in_pre = True
if self.is_block_tag(tag):
self.output = self.output.rstrip()
self.output += "<" + tag
# For now we don't strip out 'extra' attributes, because of
# raw HTML test cases.
# attrs = filter(lambda attr: attr[0] in significant_attrs, attrs)
if attrs:
attrs.sort()
for (k,v) in attrs:
self.output += " " + k
if v != None:
self.output += ("=" + '"' + cgi.escape(v,quote=True) + '"')
self.output += ">"
self.last_tag = tag
self.last = "starttag"
def handle_startendtag(self, tag, attrs):
"""Ignore closing tag for self-closing """
self.handle_starttag(tag, attrs)
self.last_tag = tag
self.last = "endtag"
def handle_comment(self, data):
self.output += '<!--' + data + '-->'
self.last = "comment"
def handle_decl(self, data):
self.output += '<!' + data + '>'
self.last = "decl"
def unknown_decl(self, data):
self.output += '<!' + data + '>'
self.last = "decl"
def handle_pi(self,data):
self.output += '<?' + data + '>'
self.last = "pi"
def handle_entityref(self, name):
try:
c = unichr(name2codepoint[name])
except KeyError:
c = None
self.output_char(c, '&' + name + ';')
self.last = "ref"
def handle_charref(self, name):
try:
if name.startswith("x"):
c = unichr(int(name[1:], 16))
else:
c = unichr(int(name))
except ValueError:
c = None
self.output_char(c, '&' + name + ';')
self.last = "ref"
# Helpers.
def output_char(self, c, fallback):
if c == u'<':
self.output += "<"
elif c == u'>':
self.output += ">"
elif c == u'&':
self.output += "&"
elif c == u'"':
self.output += """
elif c == None:
self.output += fallback
else:
self.output += c
def is_block_tag(self,tag):
return (tag in ['article', 'header', 'aside', 'hgroup', 'blockquote',
'hr', 'iframe', 'body', 'li', 'map', 'button', 'object', 'canvas',
'ol', 'caption', 'output', 'col', 'p', 'colgroup', 'pre', 'dd',
'progress', 'div', 'section', 'dl', 'table', 'td', 'dt',
'tbody', 'embed', 'textarea', 'fieldset', 'tfoot', 'figcaption',
'th', 'figure', 'thead', 'footer', 'tr', 'form', 'ul',
'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'video', 'script', 'style'])
def normalize_html(html):
r"""
Return normalized form of HTML which ignores insignificant output
differences:
Multiple inner whitespaces are collapsed to a single space (except
in pre tags):
>>> normalize_html("<p>a \t b</p>")
u'<p>a b</p>'
>>> normalize_html("<p>a \t\nb</p>")
u'<p>a b</p>'
* Whitespace surrounding block-level tags is removed.
>>> normalize_html("<p>a b</p>")
u'<p>a b</p>'
>>> normalize_html(" <p>a b</p>")
u'<p>a b</p>'
>>> normalize_html("<p>a b</p> ")
u'<p>a b</p>'
>>> normalize_html("\n\t<p>\n\t\ta b\t\t</p>\n\t")
u'<p>a b</p>'
>>> normalize_html("<i>a b</i> ")
u'<i>a b</i> '
* Self-closing tags are converted to open tags.
>>> normalize_html("<br />")
u'<br>'
* Attributes are sorted and lowercased.
>>> normalize_html('<a title="bar" HREF="foo">x</a>')
u'<a href="foo" title="bar">x</a>'
* References are converted to unicode, except that '<', '>', '&', and
'"' are rendered using entities.
>>> normalize_html("∀&><"")
u'\u2200&><"'
"""
html_chunk_re = re.compile("(\<!\[CDATA\[.*?\]\]\>|\<[^>]*\>|[^<]+)")
try:
parser = MyHTMLParser()
# We work around HTMLParser's limitations parsing CDATA
# by breaking the input into chunks and passing CDATA chunks
# through verbatim.
for chunk in re.finditer(html_chunk_re, html):
if chunk.group(0)[:8] == "<