|
| 1 | +# A parser for SGML, using the derived class as static DTD. |
| 2 | + |
| 3 | +# XXX This only supports those SGML features used by HTML. |
| 4 | + |
| 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). |
| 9 | + |
| 10 | + |
| 11 | +import regex |
| 12 | +import string |
| 13 | + |
| 14 | + |
| 15 | +# Regular expressions used for parsing |
| 16 | + |
| 17 | +incomplete = regex.compile( \ |
| 18 | + '<!-?\|</[a-zA-Z][a-zA-Z0-9]*[ \t\n]*\|</?\|' + \ |
| 19 | + '&#[a-zA-Z0-9]*\|&[a-zA-Z][a-zA-Z0-9]*\|&') |
| 20 | +entityref = regex.compile('&[a-zA-Z][a-zA-Z0-9]*[;.]') |
| 21 | +charref = regex.compile('&#[a-zA-Z0-9]+;') |
| 22 | +starttagopen = regex.compile('<[a-zA-Z]') |
| 23 | +endtag = regex.compile('</[a-zA-Z][a-zA-Z0-9]*[ \t\n]*>') |
| 24 | +commentopen = regex.compile('<!--') |
| 25 | + |
| 26 | + |
| 27 | +# SGML parser base class -- find tags and call handler functions. |
| 28 | +# Usage: p = SGMLParser(); p.feed(data); ...; p.close(). |
| 29 | +# The dtd is defined by deriving a class which defines methods |
| 30 | +# with special names to handle tags: start_foo and end_foo to handle |
| 31 | +# <foo> and </foo>, respectively, or do_foo to handle <foo> by itself. |
| 32 | +# (Tags are converted to lower case for this purpose.) The data |
| 33 | +# between tags is passed to the parser by calling self.handle_data() |
| 34 | +# with some data as argument (the data may be split up in arbutrary |
| 35 | +# chunks). Entity references are passed by calling |
| 36 | +# self.handle_entityref() with the entity reference as argument. |
| 37 | + |
| 38 | +class SGMLParser: |
| 39 | + |
| 40 | + # Interface -- initialize and reset this instance |
| 41 | + def __init__(self): |
| 42 | + self.reset() |
| 43 | + |
| 44 | + # Interface -- reset this instance. Loses all unprocessed data |
| 45 | + def reset(self): |
| 46 | + self.rawdata = '' |
| 47 | + self.stack = [] |
| 48 | + self.nomoretags = 0 |
| 49 | + self.literal = 0 |
| 50 | + |
| 51 | + # For derived classes only -- enter literal mode (CDATA) till EOF |
| 52 | + def setnomoretags(self): |
| 53 | + self.nomoretags = self.literal = 1 |
| 54 | + |
| 55 | + # For derived classes only -- enter literal mode (CDATA) |
| 56 | + def setliteral(self, *args): |
| 57 | + self.literal = 1 |
| 58 | + |
| 59 | + # Interface -- feed some data to the parser. Call this as |
| 60 | + # often as you want, with as little or as much text as you |
| 61 | + # want (may include '\n'). (This just saves the text, all the |
| 62 | + # processing is done by process() or close().) |
| 63 | + def feed(self, data): |
| 64 | + self.rawdata = self.rawdata + data |
| 65 | + self.goahead(0) |
| 66 | + |
| 67 | + # Interface -- handle the remaining data |
| 68 | + def close(self): |
| 69 | + self.goahead(1) |
| 70 | + |
| 71 | + # Internal -- handle data as far as reasonable. May leave state |
| 72 | + # and data to be processed by a subsequent call. If 'end' is |
| 73 | + # true, force handling all data as if followed by EOF marker. |
| 74 | + def goahead(self, end): |
| 75 | + rawdata = self.rawdata |
| 76 | + i = 0 |
| 77 | + n = len(rawdata) |
| 78 | + while i < n: |
| 79 | + if self.nomoretags: |
| 80 | + self.handle_data(rawdata[i:n]) |
| 81 | + i = n |
| 82 | + break |
| 83 | + j = incomplete.search(rawdata, i) |
| 84 | + if j < 0: j = n |
| 85 | + if i < j: self.handle_data(rawdata[i:j]) |
| 86 | + i = j |
| 87 | + if i == n: break |
| 88 | + if rawdata[i] == '<': |
| 89 | + if starttagopen.match(rawdata, i) >= 0: |
| 90 | + if self.literal: |
| 91 | + self.handle_data(rawdata[i]) |
| 92 | + i = i+1 |
| 93 | + continue |
| 94 | + k = self.parse_starttag(i) |
| 95 | + if k < 0: break |
| 96 | + i = i + k |
| 97 | + continue |
| 98 | + k = endtag.match(rawdata, i) |
| 99 | + if k >= 0: |
| 100 | + j = i+k |
| 101 | + self.parse_endtag(rawdata[i:j]) |
| 102 | + i = j |
| 103 | + self.literal = 0 |
| 104 | + continue |
| 105 | + if commentopen.match(rawdata, i) >= 0: |
| 106 | + if self.literal: |
| 107 | + self.handle_data(rawdata[i]) |
| 108 | + i = i+1 |
| 109 | + continue |
| 110 | + k = self.parse_comment(i) |
| 111 | + if k < 0: break |
| 112 | + i = i+k |
| 113 | + continue |
| 114 | + elif rawdata[i] == '&': |
| 115 | + k = charref.match(rawdata, i) |
| 116 | + if k >= 0: |
| 117 | + j = i+k |
| 118 | + self.handle_charref(rawdata[i+2:j-1]) |
| 119 | + i = j |
| 120 | + continue |
| 121 | + k = entityref.match(rawdata, i) |
| 122 | + if k >= 0: |
| 123 | + j = i+k |
| 124 | + self.handle_entityref(rawdata[i+1:j-1]) |
| 125 | + i = j |
| 126 | + continue |
| 127 | + else: |
| 128 | + raise RuntimeError, 'neither < nor & ??' |
| 129 | + # We get here only if incomplete matches but |
| 130 | + # nothing else |
| 131 | + k = incomplete.match(rawdata, i) |
| 132 | + if k < 0: raise RuntimeError, 'no incomplete match ??' |
| 133 | + j = i+k |
| 134 | + if j == n: break # Really incomplete |
| 135 | + self.handle_data(rawdata[i:j]) |
| 136 | + i = j |
| 137 | + # end while |
| 138 | + if end and i < n: |
| 139 | + self.handle_data(rawdata[i:n]) |
| 140 | + i = n |
| 141 | + self.rawdata = rawdata[i:] |
| 142 | + # XXX if end: check for empty stack |
| 143 | + |
| 144 | + # Internal -- parse comment, return length or -1 if not ternimated |
| 145 | + def parse_comment(self, i): |
| 146 | + rawdata = self.rawdata |
| 147 | + if rawdata[i:i+4] <> '<!--': |
| 148 | + raise RuntimeError, 'unexpected call to handle_comment' |
| 149 | + try: |
| 150 | + j = string.index(rawdata, '--', i+4) |
| 151 | + except string.index_error: |
| 152 | + return -1 |
| 153 | + self.handle_comment(rawdata[i+4: j]) |
| 154 | + j = j+2 |
| 155 | + n = len(rawdata) |
| 156 | + while j < n and rawdata[j] in ' \t\n': j = j+1 |
| 157 | + if j == n: return -1 # Wait for final '>' |
| 158 | + if rawdata[j] == '>': |
| 159 | + j = j+1 |
| 160 | + else: |
| 161 | + print '*** comment not terminated with >' |
| 162 | + print repr(rawdata[j-5:j]), '*!*', repr(rawdata[j:j+5]) |
| 163 | + return j-i |
| 164 | + |
| 165 | + # Internal -- handle starttag, return length or -1 if not terminated |
| 166 | + def parse_starttag(self, i): |
| 167 | + rawdata = self.rawdata |
| 168 | + try: |
| 169 | + j = string.index(rawdata, '>', i) |
| 170 | + except string.index_error: |
| 171 | + return -1 |
| 172 | + # Now parse the data between i+1 and j into a tag and attrs |
| 173 | + attrs = [] |
| 174 | + tagfind = regex.compile('[a-zA-Z][a-zA-Z0-9]*') |
| 175 | + attrfind = regex.compile( \ |
| 176 | + '[ \t\n]+\([a-zA-Z][a-zA-Z0-9]*\)' + \ |
| 177 | + '\([ \t\n]*=[ \t\n]*' + \ |
| 178 | + '\(\'[^\']*\';\|"[^"]*"\|[-a-zA-Z0-9./:+*%?!()_#]+\)\)?') |
| 179 | + k = tagfind.match(rawdata, i+1) |
| 180 | + if k < 0: |
| 181 | + raise RuntimeError, 'unexpected call to parse_starttag' |
| 182 | + k = i+1+k |
| 183 | + tag = string.lower(rawdata[i+1:k]) |
| 184 | + while k < j: |
| 185 | + l = attrfind.match(rawdata, k) |
| 186 | + if l < 0: break |
| 187 | + regs = attrfind.regs |
| 188 | + a1, b1 = regs[1] |
| 189 | + a2, b2 = regs[2] |
| 190 | + a3, b3 = regs[3] |
| 191 | + attrname = rawdata[a1:b1] |
| 192 | + if '=' in rawdata[k:k+l]: |
| 193 | + attrvalue = rawdata[a3:b3] |
| 194 | + if attrvalue[:1] == '\'' == attrvalue[-1:] or \ |
| 195 | + attrvalue[:1] == '"' == attrvalue[-1:]: |
| 196 | + attrvalue = attrvalue[1:-1] |
| 197 | + else: |
| 198 | + attrvalue = '' |
| 199 | + attrs.append(string.lower(attrname), attrvalue) |
| 200 | + k = k + l |
| 201 | + j = j+1 |
| 202 | + try: |
| 203 | + method = getattr(self, 'start_' + tag) |
| 204 | + except AttributeError: |
| 205 | + try: |
| 206 | + method = getattr(self, 'do_' + tag) |
| 207 | + except AttributeError: |
| 208 | + self.unknown_starttag(tag, attrs) |
| 209 | + return j-i |
| 210 | + method(attrs) |
| 211 | + return j-i |
| 212 | + self.stack.append(tag) |
| 213 | + method(attrs) |
| 214 | + return j-i |
| 215 | + |
| 216 | + # Internal -- parse endtag |
| 217 | + def parse_endtag(self, data): |
| 218 | + if data[:2] <> '</' or data[-1:] <> '>': |
| 219 | + raise RuntimeError, 'unexpected call to parse_endtag' |
| 220 | + tag = string.lower(string.strip(data[2:-1])) |
| 221 | + try: |
| 222 | + method = getattr(self, 'end_' + tag) |
| 223 | + except AttributeError: |
| 224 | + self.unknown_endtag(tag) |
| 225 | + return |
| 226 | + if self.stack and self.stack[-1] == tag: |
| 227 | + del self.stack[-1] |
| 228 | + else: |
| 229 | + print '*** Unbalanced </' + tag + '>' |
| 230 | + print '*** Stack:', self.stack |
| 231 | + found = None |
| 232 | + for i in range(len(self.stack)): |
| 233 | + if self.stack[i] == tag: found = i |
| 234 | + if found <> None: |
| 235 | + del self.stack[found:] |
| 236 | + method() |
| 237 | + |
| 238 | + # Example -- handle character reference, no need to override |
| 239 | + def handle_charref(self, name): |
| 240 | + try: |
| 241 | + n = string.atoi(name) |
| 242 | + except string.atoi_error: |
| 243 | + self.unknown_charref(name) |
| 244 | + return |
| 245 | + if not 0 <= n <= 255: |
| 246 | + self.unknown_charref(name) |
| 247 | + return |
| 248 | + self.handle_data(chr(n)) |
| 249 | + |
| 250 | + # Definition of entities -- derived classes may override |
| 251 | + entitydefs = \ |
| 252 | + {'lt': '<', 'gt': '>', 'amp': '&', 'quot': '"', 'apos': '\''} |
| 253 | + |
| 254 | + # Example -- handle entity reference, no need to override |
| 255 | + def handle_entityref(self, name): |
| 256 | + table = self.__class__.entitydefs |
| 257 | + name = string.lower(name) |
| 258 | + if table.has_key(name): |
| 259 | + self.handle_data(table[name]) |
| 260 | + else: |
| 261 | + self.unknown_entityref(name) |
| 262 | + return |
| 263 | + |
| 264 | + # Example -- handle data, should be overridden |
| 265 | + def handle_data(self, data): |
| 266 | + pass |
| 267 | + |
| 268 | + # Example -- handle comment, could be overridden |
| 269 | + def handle_comment(self, data): |
| 270 | + pass |
| 271 | + |
| 272 | + # To be overridden -- handlers for unknown objects |
| 273 | + def unknown_starttag(self, tag, attrs): pass |
| 274 | + def unknown_endtag(self, tag): pass |
| 275 | + def unknown_charref(self, ref): pass |
| 276 | + def unknown_entityref(self, ref): pass |
| 277 | + |
| 278 | + |
| 279 | +class TestSGML(SGMLParser): |
| 280 | + |
| 281 | + def handle_data(self, data): |
| 282 | + r = repr(data) |
| 283 | + if len(r) > 72: |
| 284 | + r = r[:35] + '...' + r[-35:] |
| 285 | + print 'data:', r |
| 286 | + |
| 287 | + def handle_comment(self, data): |
| 288 | + r = repr(data) |
| 289 | + if len(r) > 68: |
| 290 | + r = r[:32] + '...' + r[-32:] |
| 291 | + print 'comment:', r |
| 292 | + |
| 293 | + def unknown_starttag(self, tag, attrs): |
| 294 | + print 'start tag: <' + tag, |
| 295 | + for name, value in attrs: |
| 296 | + print name + '=' + '"' + value + '"', |
| 297 | + print '>' |
| 298 | + |
| 299 | + def unknown_endtag(self, tag): |
| 300 | + print 'end tag: </' + tag + '>' |
| 301 | + |
| 302 | + def unknown_entityref(self, ref): |
| 303 | + print '*** unknown entity ref: &' + ref + ';' |
| 304 | + |
| 305 | + def unknown_charref(self, ref): |
| 306 | + print '*** unknown char ref: &#' + ref + ';' |
| 307 | + |
| 308 | + |
| 309 | +def test(): |
| 310 | + file = 'test.html' |
| 311 | + f = open(file, 'r') |
| 312 | + x = TestSGML() |
| 313 | + while 1: |
| 314 | + line = f.readline() |
| 315 | + if not line: |
| 316 | + x.close() |
| 317 | + break |
| 318 | + x.feed(line) |
| 319 | + |
| 320 | + |
| 321 | +#test() |
0 commit comments