1
"""A parser for HTML and XHTML."""
3
# This file is based on sgmllib.py, but the API is slightly different.
5
# XXX There should be a way to distinguish between PCDATA (parsed
6
# character data -- the normal case), RCDATA (replaceable character
7
# data -- only char and entity references and end tags are special)
8
# and CDATA (character data -- only end tags are special).
14
# Regular expressions used for parsing
16
interesting_normal = re.compile('[&<]')
17
interesting_cdata = re.compile(r'<(/|\Z)')
18
incomplete = re.compile('&[a-zA-Z#]')
20
entityref = re.compile('&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]')
21
charref = re.compile('&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]')
23
starttagopen = re.compile('<[a-zA-Z]')
24
piclose = re.compile('>')
25
commentclose = re.compile(r'--\s*>')
26
tagfind = re.compile('[a-zA-Z][-.a-zA-Z0-9:_]*')
27
attrfind = re.compile(
28
r'\s*([a-zA-Z_][-.:a-zA-Z_0-9]*)(\s*=\s*'
29
r'(\'[^\']*\'|"[^"]*"|[-a-zA-Z0-9./,:;+*%?!&$\(\)_#=~@]*))?')
31
locatestarttagend = re.compile(r"""
32
<[a-zA-Z][-.a-zA-Z0-9:_]* # tag name
33
(?:\s+ # whitespace before attribute name
34
(?:[a-zA-Z_][-.:a-zA-Z0-9_]* # attribute name
35
(?:\s*=\s* # value indicator
36
(?:'[^']*' # LITA-enclosed value
37
|\"[^\"]*\" # LIT-enclosed value
38
|[^'\">\s]+ # bare value
43
\s* # trailing whitespace
45
endendtag = re.compile('>')
46
endtagfind = re.compile('</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>')
49
class HTMLParseError(Exception):
50
"""Exception raised for all parse errors."""
52
def __init__(self, msg, position=(None, None)):
55
self.lineno = position[0]
56
self.offset = position[1]
60
if self.lineno is not None:
61
result = result + ", at line %d" % self.lineno
62
if self.offset is not None:
63
result = result + ", column %d" % (self.offset + 1)
67
class HTMLParser(markupbase.ParserBase):
68
"""Find tags and other markup and call handler functions.
76
Start tags are handled by calling self.handle_starttag() or
77
self.handle_startendtag(); end tags by self.handle_endtag(). The
78
data between tags is passed from the parser to the derived class
79
by calling self.handle_data() with the data as argument (the data
80
may be split up in arbitrary chunks). Entity references are
81
passed by calling self.handle_entityref() with the entity
82
reference as the argument. Numeric character references are
83
passed to self.handle_charref() with the string containing the
84
reference as the argument.
87
CDATA_CONTENT_ELEMENTS = ("script", "style")
91
"""Initialize and reset this instance."""
95
"""Reset this instance. Loses all unprocessed data."""
98
self.interesting = interesting_normal
99
markupbase.ParserBase.reset(self)
101
def feed(self, data):
102
"""Feed data to the parser.
104
Call this as often as you want, with as little or as much text
105
as you want (may include '\n').
107
self.rawdata = self.rawdata + data
111
"""Handle any buffered data."""
114
def error(self, message):
115
raise HTMLParseError(message, self.getpos())
117
__starttag_text = None
119
def get_starttag_text(self):
120
"""Return full source of start tag: '<...>'."""
121
return self.__starttag_text
123
def set_cdata_mode(self):
124
self.interesting = interesting_cdata
126
def clear_cdata_mode(self):
127
self.interesting = interesting_normal
129
# Internal -- handle data as far as reasonable. May leave state
130
# and data to be processed by a subsequent call. If 'end' is
131
# true, force handling all data as if followed by EOF marker.
132
def goahead(self, end):
133
rawdata = self.rawdata
137
match = self.interesting.search(rawdata, i) # < or &
142
if i < j: self.handle_data(rawdata[i:j])
143
i = self.updatepos(i, j)
145
startswith = rawdata.startswith
146
if startswith('<', i):
147
if starttagopen.match(rawdata, i): # < + letter
148
k = self.parse_starttag(i)
149
elif startswith("</", i):
150
k = self.parse_endtag(i)
151
elif startswith("<!--", i):
152
k = self.parse_comment(i)
153
elif startswith("<?", i):
155
elif startswith("<!", i):
156
k = self.parse_declaration(i)
158
self.handle_data("<")
164
self.error("EOF in middle of construct")
166
i = self.updatepos(i, k)
167
elif startswith("&#", i):
168
match = charref.match(rawdata, i)
170
name = match.group()[2:-1]
171
self.handle_charref(name)
173
if not startswith(';', k-1):
175
i = self.updatepos(i, k)
179
elif startswith('&', i):
180
match = entityref.match(rawdata, i)
182
name = match.group(1)
183
self.handle_entityref(name)
185
if not startswith(';', k-1):
187
i = self.updatepos(i, k)
189
match = incomplete.match(rawdata, i)
191
# match.group() will contain at least 2 chars
192
if end and match.group() == rawdata[i:]:
193
self.error("EOF in middle of entity or char ref")
197
# not the end of the buffer, and can't be confused
198
# with some other construct
199
self.handle_data("&")
200
i = self.updatepos(i, i + 1)
204
assert 0, "interesting.search() lied"
207
self.handle_data(rawdata[i:n])
208
i = self.updatepos(i, n)
209
self.rawdata = rawdata[i:]
211
# Internal -- parse processing instr, return end or -1 if not terminated
212
def parse_pi(self, i):
213
rawdata = self.rawdata
214
assert rawdata[i:i+2] == '<?', 'unexpected call to parse_pi()'
215
match = piclose.search(rawdata, i+2) # >
219
self.handle_pi(rawdata[i+2: j])
223
# Internal -- handle starttag, return end or -1 if not terminated
224
def parse_starttag(self, i):
225
self.__starttag_text = None
226
endpos = self.check_for_whole_start_tag(i)
229
rawdata = self.rawdata
230
self.__starttag_text = rawdata[i:endpos]
232
# Now parse the data between i+1 and j into a tag and attrs
234
match = tagfind.match(rawdata, i+1)
235
assert match, 'unexpected call to parse_starttag()'
237
self.lasttag = tag = rawdata[i+1:k].lower()
240
m = attrfind.match(rawdata, k)
243
attrname, rest, attrvalue = m.group(1, 2, 3)
246
elif attrvalue[:1] == '\'' == attrvalue[-1:] or \
247
attrvalue[:1] == '"' == attrvalue[-1:]:
248
attrvalue = attrvalue[1:-1]
249
attrvalue = self.unescape(attrvalue)
250
attrs.append((attrname.lower(), attrvalue))
253
end = rawdata[k:endpos].strip()
254
if end not in (">", "/>"):
255
lineno, offset = self.getpos()
256
if "\n" in self.__starttag_text:
257
lineno = lineno + self.__starttag_text.count("\n")
258
offset = len(self.__starttag_text) \
259
- self.__starttag_text.rfind("\n")
261
offset = offset + len(self.__starttag_text)
262
self.error("junk characters in start tag: %r"
263
% (rawdata[k:endpos][:20],))
264
if end.endswith('/>'):
265
# XHTML-style empty tag: <span attr="value" />
266
self.handle_startendtag(tag, attrs)
268
self.handle_starttag(tag, attrs)
269
if tag in self.CDATA_CONTENT_ELEMENTS:
270
self.set_cdata_mode()
273
# Internal -- check to see if we have a complete starttag; return end
274
# or -1 if incomplete.
275
def check_for_whole_start_tag(self, i):
276
rawdata = self.rawdata
277
m = locatestarttagend.match(rawdata, i)
280
next = rawdata[j:j+1]
284
if rawdata.startswith("/>", j):
286
if rawdata.startswith("/", j):
290
self.updatepos(i, j + 1)
291
self.error("malformed empty start tag")
295
if next in ("abcdefghijklmnopqrstuvwxyz=/"
296
"ABCDEFGHIJKLMNOPQRSTUVWXYZ"):
297
# end of input in or before attribute value, or we have the
298
# '/' from a '/>' ending
301
self.error("malformed start tag")
302
raise AssertionError("we should not get here!")
304
# Internal -- parse endtag, return end or -1 if incomplete
305
def parse_endtag(self, i):
306
rawdata = self.rawdata
307
assert rawdata[i:i+2] == "</", "unexpected call to parse_endtag"
308
match = endendtag.search(rawdata, i+1) # >
312
match = endtagfind.match(rawdata, i) # </ + tag + >
314
self.error("bad end tag: %r" % (rawdata[i:j],))
316
self.handle_endtag(tag.lower())
317
self.clear_cdata_mode()
320
# Overridable -- finish processing of start+end tag: <tag.../>
321
def handle_startendtag(self, tag, attrs):
322
self.handle_starttag(tag, attrs)
323
self.handle_endtag(tag)
325
# Overridable -- handle start tag
326
def handle_starttag(self, tag, attrs):
329
# Overridable -- handle end tag
330
def handle_endtag(self, tag):
333
# Overridable -- handle character reference
334
def handle_charref(self, name):
337
# Overridable -- handle entity reference
338
def handle_entityref(self, name):
341
# Overridable -- handle data
342
def handle_data(self, data):
345
# Overridable -- handle comment
346
def handle_comment(self, data):
349
# Overridable -- handle declaration
350
def handle_decl(self, decl):
353
# Overridable -- handle processing instruction
354
def handle_pi(self, data):
357
def unknown_decl(self, data):
358
self.error("unknown declaration: %r" % (data,))
360
# Internal -- helper to remove special character quoting
362
def unescape(self, s):
365
def replaceEntities(s):
369
if s[0] in ['x','X']:
375
# Cannot use name2codepoint directly, because HTMLParser supports apos,
376
# which is not part of HTML 4
377
import htmlentitydefs
378
if HTMLParser.entitydefs is None:
379
entitydefs = HTMLParser.entitydefs = {'apos':u"'"}
380
for k, v in htmlentitydefs.name2codepoint.iteritems():
381
entitydefs[k] = unichr(v)
383
return self.entitydefs[s]
387
return re.sub(r"&(#?[xX]?(?:[0-9a-fA-F]+|\w{1,8}));", replaceEntities, s)