| 1 | #!/usr/bin/env python2
|
| 2 | from __future__ import print_function
|
| 3 |
|
| 4 | import unittest
|
| 5 |
|
| 6 | from lazylex import html # module under test log = html.log
|
| 7 | from typing import List
|
| 8 | from typing import Tuple
|
| 9 |
|
| 10 | log = html.log
|
| 11 |
|
| 12 | with open('lazylex/testdata.html') as f:
|
| 13 | TEST_HTML = f.read()
|
| 14 |
|
| 15 |
|
| 16 | class RegexTest(unittest.TestCase):
|
| 17 |
|
| 18 | def testDotAll(self):
|
| 19 | # type: () -> None
|
| 20 | import re
|
| 21 |
|
| 22 | # Note that $ matches end of line, not end of string
|
| 23 | p1 = re.compile(r'.')
|
| 24 | print(p1.match('\n'))
|
| 25 |
|
| 26 | p2 = re.compile(r'.', re.DOTALL)
|
| 27 | print(p2.match('\n'))
|
| 28 |
|
| 29 | #p3 = re.compile(r'[.\n]', re.VERBOSE)
|
| 30 | p3 = re.compile(r'[.\n]')
|
| 31 | print(p3.match('\n'))
|
| 32 |
|
| 33 | print('Negation')
|
| 34 |
|
| 35 | p4 = re.compile(r'[^>]')
|
| 36 | print(p4.match('\n'))
|
| 37 |
|
| 38 | def testAttrRe(self):
|
| 39 | # type: () -> None
|
| 40 | _ATTR_RE = html._ATTR_RE
|
| 41 | m = _ATTR_RE.match(' empty= val')
|
| 42 | print(m.groups())
|
| 43 |
|
| 44 |
|
| 45 | class FunctionsTest(unittest.TestCase):
|
| 46 |
|
| 47 | def testFindLineNum(self):
|
| 48 | # type: () -> None
|
| 49 | s = 'foo\n' * 3
|
| 50 | for pos in [1, 5, 10, 50]: # out of bounds
|
| 51 | line_num = html.FindLineNum(s, pos)
|
| 52 | print(line_num)
|
| 53 |
|
| 54 | def testToText(self):
|
| 55 | # type: () -> None
|
| 56 | t = html.ToText('<b name="&"> three < four && five </b>')
|
| 57 | self.assertEqual(' three < four && five ', t)
|
| 58 |
|
| 59 |
|
| 60 | def _MakeTagLexer(s):
|
| 61 | # type: (str) -> html.TagLexer
|
| 62 | lex = html.TagLexer(s)
|
| 63 | lex.Reset(0, len(s))
|
| 64 | return lex
|
| 65 |
|
| 66 |
|
| 67 | def _PrintTokens(lex):
|
| 68 | # type: (html.TagLexer) -> None
|
| 69 | log('')
|
| 70 | log('tag = %r', lex.TagName())
|
| 71 | for tok, start, end in lex.Tokens():
|
| 72 | log('%s %r', tok, lex.s[start:end])
|
| 73 |
|
| 74 |
|
| 75 | class TagLexerTest(unittest.TestCase):
|
| 76 |
|
| 77 | def testTagLexer(self):
|
| 78 | # type: () -> None
|
| 79 | # Invalid!
|
| 80 | #lex = _MakeTagLexer('< >')
|
| 81 | #print(lex.Tag())
|
| 82 |
|
| 83 | lex = _MakeTagLexer('<a>')
|
| 84 | _PrintTokens(lex)
|
| 85 |
|
| 86 | lex = _MakeTagLexer('<a novalue>')
|
| 87 | _PrintTokens(lex)
|
| 88 |
|
| 89 | # Note: we could have a different HasAttr() method
|
| 90 | # <a novalue> means lex.Get('novalue') == ''
|
| 91 | # https://developer.mozilla.org/en-US/docs/Web/API/Element/hasAttribute
|
| 92 | self.assertEqual('', lex.GetAttrRaw('novalue'))
|
| 93 |
|
| 94 | lex = _MakeTagLexer('<a href="double quoted">')
|
| 95 | _PrintTokens(lex)
|
| 96 |
|
| 97 | self.assertEqual('double quoted', lex.GetAttrRaw('href'))
|
| 98 | self.assertEqual(None, lex.GetAttrRaw('oops'))
|
| 99 |
|
| 100 | lex = _MakeTagLexer('<a href=foo class="bar">')
|
| 101 | _PrintTokens(lex)
|
| 102 |
|
| 103 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 104 | _PrintTokens(lex)
|
| 105 |
|
| 106 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 107 | self.assertEqual('?foo=1&bar=2', lex.GetAttrRaw('href'))
|
| 108 |
|
| 109 | def testTagName(self):
|
| 110 | # type: () -> None
|
| 111 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 112 | self.assertEqual('a', lex.TagName())
|
| 113 |
|
| 114 | def testAllAttrs(self):
|
| 115 | # type: () -> None
|
| 116 | """
|
| 117 | [('key', 'value')] for all
|
| 118 | """
|
| 119 | # closed
|
| 120 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 121 | self.assertEqual([('href', 'foo'), ('class', 'bar')],
|
| 122 | lex.AllAttrsRaw())
|
| 123 |
|
| 124 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 125 | self.assertEqual([('href', '?foo=1&bar=2')], lex.AllAttrsRaw())
|
| 126 |
|
| 127 | def testEmptyMissingValues(self):
|
| 128 | # type: () -> None
|
| 129 | # equivalent to <button disabled="">
|
| 130 | lex = _MakeTagLexer('<button disabled>')
|
| 131 | all_attrs = lex.AllAttrsRaw()
|
| 132 | self.assertEqual([('disabled', '')], all_attrs)
|
| 133 |
|
| 134 | slices = lex.AllAttrsRawSlice()
|
| 135 | log('slices %s', slices)
|
| 136 |
|
| 137 | lex = _MakeTagLexer(
|
| 138 | '''<p double="" single='' empty= value missing empty2=>''')
|
| 139 | all_attrs = lex.AllAttrsRaw()
|
| 140 | self.assertEqual([
|
| 141 | ('double', ''),
|
| 142 | ('single', ''),
|
| 143 | ('empty', 'value'),
|
| 144 | ('missing', ''),
|
| 145 | ('empty2', ''),
|
| 146 | ], all_attrs)
|
| 147 | # TODO: should have
|
| 148 | log('all %s', all_attrs)
|
| 149 |
|
| 150 | slices = lex.AllAttrsRawSlice()
|
| 151 | log('slices %s', slices)
|
| 152 |
|
| 153 | def testInvalidTag(self):
|
| 154 | # type: () -> None
|
| 155 | try:
|
| 156 | lex = _MakeTagLexer('<a foo=bar !></a>')
|
| 157 | all_attrs = lex.AllAttrsRaw()
|
| 158 | except html.LexError as e:
|
| 159 | print(e)
|
| 160 | else:
|
| 161 | self.fail('Expected LexError')
|
| 162 |
|
| 163 |
|
| 164 | def _MakeAttrValueLexer(s):
|
| 165 | # type: (str) -> html.AttrValueLexer
|
| 166 | lex = html.AttrValueLexer(s)
|
| 167 | lex.Reset(0, len(s))
|
| 168 | return lex
|
| 169 |
|
| 170 |
|
| 171 | class AttrValueLexerTest(unittest.TestCase):
|
| 172 |
|
| 173 | def testGood(self):
|
| 174 | # type: () -> None
|
| 175 | lex = _MakeAttrValueLexer('?foo=42&bar=99')
|
| 176 | n = lex.NumTokens()
|
| 177 | self.assertEqual(3, n)
|
| 178 |
|
| 179 |
|
| 180 | def Lex(h, no_special_tags=False):
|
| 181 | # type: (str, bool) -> List[Tuple[int, int]]
|
| 182 | print(repr(h))
|
| 183 | tokens = html.ValidTokenList(h, no_special_tags=no_special_tags)
|
| 184 | start_pos = 0
|
| 185 | for tok_id, end_pos in tokens:
|
| 186 | frag = h[start_pos:end_pos]
|
| 187 | log('%d %s %r', end_pos, html.TokenName(tok_id), frag)
|
| 188 | start_pos = end_pos
|
| 189 | return tokens
|
| 190 |
|
| 191 |
|
| 192 | class LexerTest(unittest.TestCase):
|
| 193 |
|
| 194 | # IndexLinker in devtools/make_help.py
|
| 195 | # <pre> sections in doc/html_help.py
|
| 196 | # TocExtractor in devtools/cmark.py
|
| 197 |
|
| 198 | def testPstrip(self):
|
| 199 | # type: () -> None
|
| 200 | """Remove anything like this.
|
| 201 |
|
| 202 | <p><pstrip> </pstrip></p>
|
| 203 | """
|
| 204 | pass
|
| 205 |
|
| 206 | def testCommentParse(self):
|
| 207 | # type: () -> None
|
| 208 | n = len(TEST_HTML)
|
| 209 | tokens = Lex(TEST_HTML)
|
| 210 |
|
| 211 | def testCommentParse2(self):
|
| 212 | # type: () -> None
|
| 213 |
|
| 214 | Tok = html.Tok
|
| 215 | h = '''
|
| 216 | hi <!-- line 1
|
| 217 | line 2 --><br/>'''
|
| 218 | tokens = Lex(h)
|
| 219 |
|
| 220 | self.assertEqual(
|
| 221 | [
|
| 222 | (Tok.RawData, 12),
|
| 223 | (Tok.Comment, 50), # <? err ?>
|
| 224 | (Tok.StartEndTag, 55),
|
| 225 | (Tok.EndOfStream, 55),
|
| 226 | ],
|
| 227 | tokens)
|
| 228 |
|
| 229 | def testProcessingInstruction(self):
|
| 230 | # type: () -> None
|
| 231 | # <?xml ?> header
|
| 232 | Tok = html.Tok
|
| 233 | h = 'hi <? err ?>'
|
| 234 | tokens = Lex(h)
|
| 235 |
|
| 236 | self.assertEqual(
|
| 237 | [
|
| 238 | (Tok.RawData, 3),
|
| 239 | (Tok.Processing, 12), # <? err ?>
|
| 240 | (Tok.EndOfStream, 12),
|
| 241 | ],
|
| 242 | tokens)
|
| 243 |
|
| 244 | def testScriptStyle(self):
|
| 245 | # type: () -> None
|
| 246 | Tok = html.Tok
|
| 247 | h = '''
|
| 248 | hi <script src=""> if (x < 1 && y > 2 ) { console.log(""); }
|
| 249 | </script>
|
| 250 | '''
|
| 251 | tokens = Lex(h)
|
| 252 |
|
| 253 | expected = [
|
| 254 | (Tok.RawData, 12),
|
| 255 | (Tok.StartTag, 27), # <script>
|
| 256 | (Tok.HtmlCData, 78), # JavaScript code is HTML CData
|
| 257 | (Tok.EndTag, 87), # </script>
|
| 258 | (Tok.RawData, 96), # \n
|
| 259 | (Tok.EndOfStream, 96), # \n
|
| 260 | ]
|
| 261 | self.assertEqual(expected, tokens)
|
| 262 |
|
| 263 | # Test case matching
|
| 264 | tokens = Lex(h.replace('script', 'scrIPT'))
|
| 265 | self.assertEqual(expected, tokens)
|
| 266 |
|
| 267 | def testScriptStyleXml(self):
|
| 268 | # type: () -> None
|
| 269 | Tok = html.Tok
|
| 270 | h = 'hi <script src=""> < </script>'
|
| 271 | # XML mode
|
| 272 | tokens = Lex(h, no_special_tags=True)
|
| 273 |
|
| 274 | self.assertEqual(
|
| 275 | [
|
| 276 | (Tok.RawData, 3),
|
| 277 | (Tok.StartTag, 18), # <script>
|
| 278 | (Tok.RawData, 19), # space
|
| 279 | (Tok.CharEntity, 23), # </script>
|
| 280 | (Tok.RawData, 24), # \n
|
| 281 | (Tok.EndTag, 33), # \n
|
| 282 | (Tok.EndOfStream, 33), # \n
|
| 283 | ],
|
| 284 | tokens)
|
| 285 |
|
| 286 | def testCData(self):
|
| 287 | # type: () -> None
|
| 288 | Tok = html.Tok
|
| 289 |
|
| 290 | # from
|
| 291 | # /home/andy/src/languages/Python-3.11.5/Lib/test/xmltestdata/c14n-20/inC14N4.xml
|
| 292 | h = '<compute><![CDATA[value>"0" && value<"10" ?"valid":"error"]]></compute>'
|
| 293 | tokens = Lex(h)
|
| 294 |
|
| 295 | self.assertEqual([
|
| 296 | (Tok.StartTag, 9),
|
| 297 | (Tok.CData, 61),
|
| 298 | (Tok.EndTag, 71),
|
| 299 | (Tok.EndOfStream, 71),
|
| 300 | ], tokens)
|
| 301 |
|
| 302 | def testEntity(self):
|
| 303 | # type: () -> None
|
| 304 | Tok = html.Tok
|
| 305 |
|
| 306 | # from
|
| 307 | # /home/andy/src/Python-3.12.4/Lib/test/xmltestdata/c14n-20/inC14N5.xml
|
| 308 | h = '&ent1;, &ent2;!'
|
| 309 |
|
| 310 | tokens = Lex(h)
|
| 311 |
|
| 312 | self.assertEqual([
|
| 313 | (Tok.CharEntity, 6),
|
| 314 | (Tok.RawData, 8),
|
| 315 | (Tok.CharEntity, 14),
|
| 316 | (Tok.RawData, 15),
|
| 317 | (Tok.EndOfStream, 15),
|
| 318 | ], tokens)
|
| 319 |
|
| 320 | def testStartTag(self):
|
| 321 | # type: () -> None
|
| 322 | Tok = html.Tok
|
| 323 |
|
| 324 | h = '<a>hi</a>'
|
| 325 | tokens = Lex(h)
|
| 326 |
|
| 327 | self.assertEqual([
|
| 328 | (Tok.StartTag, 3),
|
| 329 | (Tok.RawData, 5),
|
| 330 | (Tok.EndTag, 9),
|
| 331 | (Tok.EndOfStream, 9),
|
| 332 | ], tokens)
|
| 333 |
|
| 334 | # Make sure we don't consume too much
|
| 335 | h = '<a><source>1.7</source></a>'
|
| 336 |
|
| 337 | tokens = Lex(h)
|
| 338 |
|
| 339 | self.assertEqual([
|
| 340 | (Tok.StartTag, 3),
|
| 341 | (Tok.StartTag, 11),
|
| 342 | (Tok.RawData, 14),
|
| 343 | (Tok.EndTag, 23),
|
| 344 | (Tok.EndTag, 27),
|
| 345 | (Tok.EndOfStream, 27),
|
| 346 | ], tokens)
|
| 347 |
|
| 348 | return
|
| 349 |
|
| 350 | h = '''
|
| 351 | <configuration>
|
| 352 | <source>1.7</source>
|
| 353 | </configuration>'''
|
| 354 |
|
| 355 | tokens = Lex(h)
|
| 356 |
|
| 357 | self.assertEqual([
|
| 358 | (Tok.RawData, 9),
|
| 359 | (Tok.StartTag, 24),
|
| 360 | (Tok.RawData, 9),
|
| 361 | (Tok.EndOfStream, 9),
|
| 362 | ], tokens)
|
| 363 |
|
| 364 | def testBad(self):
|
| 365 | # type: () -> None
|
| 366 | Tok = html.Tok
|
| 367 |
|
| 368 | h = '&'
|
| 369 | tokens = Lex(h)
|
| 370 |
|
| 371 | self.assertEqual([
|
| 372 | (Tok.BadAmpersand, 1),
|
| 373 | (Tok.EndOfStream, 1),
|
| 374 | ], tokens)
|
| 375 |
|
| 376 | h = '>'
|
| 377 | tokens = Lex(h)
|
| 378 |
|
| 379 | self.assertEqual([
|
| 380 | (Tok.BadGreaterThan, 1),
|
| 381 | (Tok.EndOfStream, 1),
|
| 382 | ], tokens)
|
| 383 |
|
| 384 | def testInvalid(self):
|
| 385 | # type: () -> None
|
| 386 | Tok = html.Tok
|
| 387 |
|
| 388 | for s in INVALID_LEX:
|
| 389 | try:
|
| 390 | tokens = html.ValidTokenList(s)
|
| 391 | except html.LexError as e:
|
| 392 | print(e)
|
| 393 | else:
|
| 394 | self.fail('Expected LexError %r' % s)
|
| 395 |
|
| 396 | def testValid(self):
|
| 397 | # type: () -> None
|
| 398 | for s, _ in VALID_LEX:
|
| 399 | tokens = Lex(s)
|
| 400 | print()
|
| 401 |
|
| 402 |
|
| 403 | INVALID_LEX = [
|
| 404 | '<a><',
|
| 405 | '&<',
|
| 406 | '&<',
|
| 407 | # Hm > is allowed?
|
| 408 | #'a > b',
|
| 409 | 'a < b',
|
| 410 | '<!-- unfinished comment',
|
| 411 | '<? unfinished processing',
|
| 412 | '</div bad=attr> <a> <b>',
|
| 413 |
|
| 414 | # not allowed, but 3 > 4 is allowed
|
| 415 | '<a> 3 < 4 </a>',
|
| 416 | # Not a CDATA tag
|
| 417 | '<STYLEz><</STYLEz>',
|
| 418 | ]
|
| 419 |
|
| 420 | SKIP = 0
|
| 421 | UNCHANGED = 1
|
| 422 |
|
| 423 | VALID_LEX = [
|
| 424 | # TODO: convert these to XML
|
| 425 | ('<foo></foo>', UNCHANGED),
|
| 426 | ('<foo x=y></foo>', ''),
|
| 427 | #('<foo x="&"></foo>', '<foo x="&"></foo>'),
|
| 428 | ('<foo x="&"></foo>', ''),
|
| 429 |
|
| 430 | # Allowed with BadAmpersand
|
| 431 | ('<p> x & y </p>', '<p> x & y </p>'),
|
| 432 | ]
|
| 433 |
|
| 434 | INVALID_PARSE = [
|
| 435 | '<a></b>',
|
| 436 | '<a>', # missing closing tag
|
| 437 | '<meta></meta>', # this is a self-closing tag
|
| 438 | ]
|
| 439 |
|
| 440 | VALID_PARSE = [
|
| 441 | ('<!DOCTYPE html>\n', ''),
|
| 442 | ('<!DOCTYPE>', ''),
|
| 443 |
|
| 444 | # empty strings
|
| 445 | ('<p x=""></p>', UNCHANGED),
|
| 446 | ("<p x=''></p>", UNCHANGED),
|
| 447 | ('<self-closing a="b" />', UNCHANGED),
|
| 448 |
|
| 449 | # We could also normalize CDATA?
|
| 450 | # Note that CDATA has an escaping problem: you need to handle it ]]> with
|
| 451 | # concatenation. It just "pushes the problem around".
|
| 452 | # So I think it's better to use ONE kind of escaping, which is <
|
| 453 | ('<script><![CDATA[ <wtf> >< ]]></script>', UNCHANGED),
|
| 454 |
|
| 455 | # allowed, but 3 < 4 is not allowed
|
| 456 | ('<a> 3 > 4 </a>', '<a> 3 > 4 </a>'),
|
| 457 | # allowed, but 3 > 4 is not allowed
|
| 458 | ('<p x="3 < 4"></p>', ''),
|
| 459 | ('<b><a href="foo">link</a></b>', UNCHANGED),
|
| 460 |
|
| 461 | # TODO: should be self-closing
|
| 462 | #('<meta><a></a>', '<meta/><a></a>'),
|
| 463 | ('<meta><a></a>', ''),
|
| 464 |
|
| 465 | # no attribute
|
| 466 | ('<button disabled></button>', ''),
|
| 467 | ('<button disabled=></button>', ''),
|
| 468 | ('<button disabled= ></button>', ''),
|
| 469 |
|
| 470 | # single quoted is pretty common
|
| 471 | ("<a href='single'></a>", ''),
|
| 472 |
|
| 473 | # Conceding to reality - I used these myself
|
| 474 | ('<a href=ble.sh></a>', ''),
|
| 475 | ('<a href=foo.html></a>', ''),
|
| 476 | ('<foo x="&"></foo>', ''),
|
| 477 |
|
| 478 | # caps
|
| 479 | ('<foo></FOO>', ''),
|
| 480 | ('<Foo></fOO>', ''),
|
| 481 |
|
| 482 | # capital VOID tag
|
| 483 | ('<META><a></a>', ''),
|
| 484 | ('<script><</script>', ''),
|
| 485 | # matching
|
| 486 | ('<SCRipt><</SCRipt>', ''),
|
| 487 | ('<SCRIPT><</SCRIPT>', ''),
|
| 488 | ('<STYLE><</STYLE>', ''),
|
| 489 | #'<SCRipt><</script>',
|
| 490 |
|
| 491 | # Note: Python HTMLParser.py does DYNAMIC compilation of regex with re.I
|
| 492 | # flag to handle this! Gah I want something faster.
|
| 493 | #'<script><</SCRIPT>',
|
| 494 |
|
| 495 | # TODO: Test <svg> and <math> ?
|
| 496 | ]
|
| 497 |
|
| 498 | VALID_XML = [
|
| 499 | '<meta></meta>',
|
| 500 | ]
|
| 501 |
|
| 502 | INVALID_TAG_LEX = [
|
| 503 | # not allowed, but 3 < 4 is allowed
|
| 504 | '<p x="3 > 4"></p>',
|
| 505 | # same thing
|
| 506 | '<a href=">"></a>',
|
| 507 | '<a foo=bar !></a>', # bad attr
|
| 508 | ]
|
| 509 |
|
| 510 |
|
| 511 | class ValidateTest(unittest.TestCase):
|
| 512 |
|
| 513 | def testInvalid(self):
|
| 514 | # type: () -> None
|
| 515 | counters = html.Counters()
|
| 516 | for s in INVALID_LEX + INVALID_TAG_LEX:
|
| 517 | try:
|
| 518 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 519 | except html.LexError as e:
|
| 520 | print(e)
|
| 521 | else:
|
| 522 | self.fail('Expected LexError %r' % s)
|
| 523 |
|
| 524 | for s in INVALID_PARSE:
|
| 525 | try:
|
| 526 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 527 | except html.ParseError as e:
|
| 528 | print(e)
|
| 529 | else:
|
| 530 | self.fail('Expected ParseError')
|
| 531 |
|
| 532 | def testValid(self):
|
| 533 | # type: () -> None
|
| 534 | counters = html.Counters()
|
| 535 | for s, _ in VALID_PARSE:
|
| 536 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 537 | print('HTML5 %r' % s)
|
| 538 | print('HTML5 attrs %r' % counters.debug_attrs)
|
| 539 |
|
| 540 | def testValidXml(self):
|
| 541 | # type: () -> None
|
| 542 | counters = html.Counters()
|
| 543 | for s in VALID_XML:
|
| 544 | html.Validate(s, html.BALANCED_TAGS | html.NO_SPECIAL_TAGS,
|
| 545 | counters)
|
| 546 | print('XML %r' % s)
|
| 547 | print('XML attrs %r' % counters.debug_attrs)
|
| 548 |
|
| 549 |
|
| 550 | class XmlTest(unittest.TestCase):
|
| 551 |
|
| 552 | def testValid(self):
|
| 553 | # type: () -> None
|
| 554 | counters = html.Counters()
|
| 555 | for h, expected_xml in VALID_LEX + VALID_PARSE:
|
| 556 | actual = html.ToXml(h)
|
| 557 | if expected_xml == UNCHANGED: # Unchanged
|
| 558 | self.assertEqual(h, actual)
|
| 559 | elif expected_xml == '': # Skip
|
| 560 | pass
|
| 561 | else:
|
| 562 | self.assertEqual(expected_xml, actual)
|
| 563 |
|
| 564 |
|
| 565 | if __name__ == '__main__':
|
| 566 | unittest.main()
|