| 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 doctools.util import log
|
| 8 |
|
| 9 |
|
| 10 | class FunctionsTest(unittest.TestCase):
|
| 11 |
|
| 12 | def testToText(self):
|
| 13 | # type: () -> None
|
| 14 | t = html.ToText('<b name="&"> three < four && five </b>')
|
| 15 | self.assertEqual(' three < four && five ', t)
|
| 16 |
|
| 17 |
|
| 18 | def _MakeTagLexer(s):
|
| 19 | # type: (str) -> html.TagLexer
|
| 20 | lex = html.TagLexer(s)
|
| 21 | lex.Reset(0, len(s))
|
| 22 | return lex
|
| 23 |
|
| 24 |
|
| 25 | def _PrintTokens(lex):
|
| 26 | # type: (html.TagLexer) -> None
|
| 27 | log('')
|
| 28 | log('tag = %r', lex.GetTagName())
|
| 29 | for tok, start, end in lex.Tokens():
|
| 30 | log('%s %r', tok, lex.s[start:end])
|
| 31 |
|
| 32 |
|
| 33 | class TagLexerTest(unittest.TestCase):
|
| 34 |
|
| 35 | def testTagName_DEPRECATED(self):
|
| 36 | # type: () -> None
|
| 37 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 38 | self.assertEqual('a', lex.GetTagName())
|
| 39 |
|
| 40 | def testGetAttrRaw(self):
|
| 41 | # type: () -> None
|
| 42 | lex = _MakeTagLexer('<a>')
|
| 43 | _PrintTokens(lex)
|
| 44 | self.assertEqual(None, lex.GetAttrRaw('oops'))
|
| 45 |
|
| 46 | # <a novalue> means lex.Get('novalue') == ''
|
| 47 | # https://developer.mozilla.org/en-US/docs/Web/API/Element/hasAttribute
|
| 48 | # We are not distinguishing <a novalue=""> from <a novalue> in this API
|
| 49 | lex = _MakeTagLexer('<a novalue>')
|
| 50 | _PrintTokens(lex)
|
| 51 | self.assertEqual('', lex.GetAttrRaw('novalue'))
|
| 52 |
|
| 53 | lex = _MakeTagLexer('<a href="double quoted">')
|
| 54 | _PrintTokens(lex)
|
| 55 |
|
| 56 | self.assertEqual('double quoted', lex.GetAttrRaw('href'))
|
| 57 | self.assertEqual(None, lex.GetAttrRaw('oops'))
|
| 58 |
|
| 59 | lex = _MakeTagLexer('<a href=foo class="bar">')
|
| 60 | _PrintTokens(lex)
|
| 61 | self.assertEqual('bar', lex.GetAttrRaw('class'))
|
| 62 |
|
| 63 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 64 | _PrintTokens(lex)
|
| 65 | self.assertEqual('bar', lex.GetAttrRaw('class'))
|
| 66 |
|
| 67 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 68 | self.assertEqual('?foo=1&bar=2', lex.GetAttrRaw('href'))
|
| 69 |
|
| 70 | def testAllAttrs(self):
|
| 71 | # type: () -> None
|
| 72 | """
|
| 73 | [('key', 'value')] for all
|
| 74 | """
|
| 75 | # closed
|
| 76 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 77 | self.assertEqual([('href', 'foo'), ('class', 'bar')],
|
| 78 | lex.AllAttrsRaw())
|
| 79 |
|
| 80 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 81 | self.assertEqual([('href', '?foo=1&bar=2')], lex.AllAttrsRaw())
|
| 82 |
|
| 83 | def testEmptyMissingValues(self):
|
| 84 | # type: () -> None
|
| 85 | # equivalent to <button disabled="">
|
| 86 | lex = _MakeTagLexer('<button disabled>')
|
| 87 | all_attrs = lex.AllAttrsRaw()
|
| 88 | self.assertEqual([('disabled', '')], all_attrs)
|
| 89 |
|
| 90 | slices = lex.AllAttrsRawSlice()
|
| 91 | log('slices %s', slices)
|
| 92 |
|
| 93 | lex = _MakeTagLexer(
|
| 94 | '''<p double="" single='' empty= value missing empty2=>''')
|
| 95 | all_attrs = lex.AllAttrsRaw()
|
| 96 | self.assertEqual([
|
| 97 | ('double', ''),
|
| 98 | ('single', ''),
|
| 99 | ('empty', 'value'),
|
| 100 | ('missing', ''),
|
| 101 | ('empty2', ''),
|
| 102 | ], all_attrs)
|
| 103 | # TODO: should have
|
| 104 | log('all %s', all_attrs)
|
| 105 |
|
| 106 | slices = lex.AllAttrsRawSlice()
|
| 107 | log('slices %s', slices)
|
| 108 |
|
| 109 | def testInvalidTag(self):
|
| 110 | # type: () -> None
|
| 111 | try:
|
| 112 | lex = _MakeTagLexer('<a foo=bar !></a>')
|
| 113 | all_attrs = lex.AllAttrsRaw()
|
| 114 | except html.LexError as e:
|
| 115 | print(e)
|
| 116 | else:
|
| 117 | self.fail('Expected LexError')
|
| 118 |
|
| 119 |
|
| 120 | def _MakeAttrValueLexer(s):
|
| 121 | # type: (str) -> html.AttrValueLexer
|
| 122 | lex = html.AttrValueLexer(s)
|
| 123 | lex.Reset(0, len(s))
|
| 124 | return lex
|
| 125 |
|
| 126 |
|
| 127 | class AttrValueLexerTest(unittest.TestCase):
|
| 128 |
|
| 129 | def testGood(self):
|
| 130 | # type: () -> None
|
| 131 | lex = _MakeAttrValueLexer('?foo=42&bar=99')
|
| 132 | n = lex.NumTokens()
|
| 133 | self.assertEqual(3, n)
|
| 134 |
|
| 135 |
|
| 136 | class LexerTest(unittest.TestCase):
|
| 137 |
|
| 138 | def testInvalid(self):
|
| 139 | # type: () -> None
|
| 140 | from data_lang.htm8_test import ValidTokenList
|
| 141 | for s in INVALID_LEX:
|
| 142 | try:
|
| 143 | tokens = ValidTokenList(s)
|
| 144 | except html.LexError as e:
|
| 145 | print(e)
|
| 146 | else:
|
| 147 | self.fail('Expected LexError %r' % s)
|
| 148 |
|
| 149 | def testValid(self):
|
| 150 | # type: () -> None
|
| 151 |
|
| 152 | from data_lang.htm8_test import Lex
|
| 153 |
|
| 154 | for s, _ in VALID_LEX:
|
| 155 | tokens = Lex(s)
|
| 156 | print()
|
| 157 |
|
| 158 |
|
| 159 | INVALID_LEX = [
|
| 160 | '< >',
|
| 161 | '<a><',
|
| 162 | '&<',
|
| 163 | '&<',
|
| 164 | # Hm > is allowed?
|
| 165 | #'a > b',
|
| 166 | 'a < b',
|
| 167 | '<!-- unfinished comment',
|
| 168 | '<? unfinished processing',
|
| 169 | '</div bad=attr> <a> <b>',
|
| 170 |
|
| 171 | # not allowed, but 3 > 4 is allowed
|
| 172 | '<a> 3 < 4 </a>',
|
| 173 | # Not a CDATA tag
|
| 174 | '<STYLEz><</STYLEz>',
|
| 175 | ]
|
| 176 |
|
| 177 | SKIP = 0
|
| 178 | UNCHANGED = 1
|
| 179 |
|
| 180 | VALID_LEX = [
|
| 181 | # TODO: convert these to XML
|
| 182 | ('<foo></foo>', UNCHANGED),
|
| 183 | ('<foo x=y></foo>', ''),
|
| 184 | #('<foo x="&"></foo>', '<foo x="&"></foo>'),
|
| 185 | ('<foo x="&"></foo>', ''),
|
| 186 |
|
| 187 | # Allowed with BadAmpersand
|
| 188 | ('<p> x & y </p>', '<p> x & y </p>'),
|
| 189 |
|
| 190 | # No ambiguity
|
| 191 | ('<img src=/ >', ''),
|
| 192 | ('<img src="/">', ''),
|
| 193 | ('<img src=foo/ >', ''),
|
| 194 | ]
|
| 195 |
|
| 196 | INVALID_PARSE = [
|
| 197 | '<a></b>',
|
| 198 | '<a>', # missing closing tag
|
| 199 | '<meta></meta>', # this is a self-closing tag
|
| 200 | ]
|
| 201 |
|
| 202 | INVALID_ATTR_LEX = [
|
| 203 | # Ambiguous, should be ""
|
| 204 | '<img src=/>',
|
| 205 | '<img src= />',
|
| 206 | '<img src=foo/>',
|
| 207 | '<img src= foo/>',
|
| 208 | ]
|
| 209 |
|
| 210 | VALID_PARSE = [
|
| 211 | ('<!DOCTYPE html>\n', ''),
|
| 212 | ('<!DOCTYPE>', ''),
|
| 213 |
|
| 214 | # empty strings
|
| 215 | ('<p x=""></p>', UNCHANGED),
|
| 216 | ("<p x=''></p>", UNCHANGED),
|
| 217 | ('<self-closing a="b" />', UNCHANGED),
|
| 218 |
|
| 219 | # We could also normalize CDATA?
|
| 220 | # Note that CDATA has an escaping problem: you need to handle it ]]> with
|
| 221 | # concatenation. It just "pushes the problem around".
|
| 222 | # So I think it's better to use ONE kind of escaping, which is <
|
| 223 | ('<script><![CDATA[ <wtf> >< ]]></script>', UNCHANGED),
|
| 224 |
|
| 225 | # allowed, but 3 < 4 is not allowed
|
| 226 | ('<a> 3 > 4 </a>', '<a> 3 > 4 </a>'),
|
| 227 | # allowed, but 3 > 4 is not allowed
|
| 228 | ('<p x="3 < 4"></p>', ''),
|
| 229 | ('<b><a href="foo">link</a></b>', UNCHANGED),
|
| 230 |
|
| 231 | # TODO: should be self-closing
|
| 232 | #('<meta><a></a>', '<meta/><a></a>'),
|
| 233 | ('<meta><a></a>', ''),
|
| 234 |
|
| 235 | # no attribute
|
| 236 | ('<button disabled></button>', ''),
|
| 237 | ('<button disabled=></button>', ''),
|
| 238 | ('<button disabled= ></button>', ''),
|
| 239 |
|
| 240 | # single quoted is pretty common
|
| 241 | ("<a href='single'></a>", ''),
|
| 242 |
|
| 243 | # Conceding to reality - I used these myself
|
| 244 | ('<a href=ble.sh></a>', ''),
|
| 245 | ('<a href=foo.html></a>', ''),
|
| 246 | ('<foo x="&"></foo>', ''),
|
| 247 |
|
| 248 | # caps
|
| 249 | ('<foo></FOO>', ''),
|
| 250 | ('<Foo></fOO>', ''),
|
| 251 |
|
| 252 | # capital VOID tag
|
| 253 | ('<META><a></a>', ''),
|
| 254 | ('<script><</script>', ''),
|
| 255 | # matching
|
| 256 | ('<SCRipt><</SCRipt>', ''),
|
| 257 | ('<SCRIPT><</SCRIPT>', ''),
|
| 258 | ('<STYLE><</STYLE>', ''),
|
| 259 | #'<SCRipt><</script>',
|
| 260 |
|
| 261 | # Regression test from blog
|
| 262 | ('<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>',
|
| 263 | '')
|
| 264 |
|
| 265 | # Note: Python HTMLParser.py does DYNAMIC compilation of regex with re.I
|
| 266 | # flag to handle this! Gah I want something faster.
|
| 267 | #'<script><</SCRIPT>',
|
| 268 |
|
| 269 | # TODO: Test <svg> and <math> ?
|
| 270 | ]
|
| 271 |
|
| 272 | VALID_XML = [
|
| 273 | '<meta></meta>',
|
| 274 | ]
|
| 275 |
|
| 276 | INVALID_TAG_LEX = [
|
| 277 | # not allowed, but 3 < 4 is allowed
|
| 278 | '<p x="3 > 4"></p>',
|
| 279 | # same thing
|
| 280 | '<a href=">"></a>',
|
| 281 | '<a foo=bar !></a>', # bad attr
|
| 282 | ]
|
| 283 |
|
| 284 |
|
| 285 | class ValidateTest(unittest.TestCase):
|
| 286 |
|
| 287 | def testInvalidOld(self):
|
| 288 | # type: () -> None
|
| 289 | counters = html.Counters()
|
| 290 | for s in INVALID_LEX + INVALID_TAG_LEX:
|
| 291 | try:
|
| 292 | html.ValidateOld(s, html.BALANCED_TAGS, counters)
|
| 293 | except html.LexError as e:
|
| 294 | print(e)
|
| 295 | else:
|
| 296 | self.fail('Expected LexError %r' % s)
|
| 297 |
|
| 298 | for s in INVALID_PARSE:
|
| 299 | try:
|
| 300 | html.ValidateOld(s, html.BALANCED_TAGS, counters)
|
| 301 | except html.ParseError as e:
|
| 302 | print(e)
|
| 303 | else:
|
| 304 | self.fail('Expected ParseError')
|
| 305 |
|
| 306 | def testValidOld(self):
|
| 307 | # type: () -> None
|
| 308 | counters = html.Counters()
|
| 309 | for s, _ in VALID_PARSE:
|
| 310 | html.ValidateOld(s, html.BALANCED_TAGS, counters)
|
| 311 | print('HTML5 %r' % s)
|
| 312 | #print('HTML5 attrs %r' % counters.debug_attrs)
|
| 313 |
|
| 314 | def testValidXmlOld(self):
|
| 315 | # type: () -> None
|
| 316 | counters = html.Counters()
|
| 317 | for s in VALID_XML:
|
| 318 | html.ValidateOld(s, html.BALANCED_TAGS | html.NO_SPECIAL_TAGS,
|
| 319 | counters)
|
| 320 | print('XML %r' % s)
|
| 321 | #print('XML attrs %r' % counters.debug_attrs)
|
| 322 |
|
| 323 | def testInvalid(self):
|
| 324 | # type: () -> None
|
| 325 | counters = html.Counters()
|
| 326 | for s in INVALID_LEX + INVALID_TAG_LEX + INVALID_ATTR_LEX:
|
| 327 | try:
|
| 328 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 329 | except html.LexError as e:
|
| 330 | print(e)
|
| 331 | else:
|
| 332 | self.fail('Expected LexError %r' % s)
|
| 333 |
|
| 334 | for s in INVALID_PARSE:
|
| 335 | try:
|
| 336 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 337 | except html.ParseError as e:
|
| 338 | print(e)
|
| 339 | else:
|
| 340 | self.fail('Expected ParseError')
|
| 341 |
|
| 342 | def testValid(self):
|
| 343 | # type: () -> None
|
| 344 | counters = html.Counters()
|
| 345 | for s, _ in VALID_PARSE:
|
| 346 | print('HTML5 %r' % s)
|
| 347 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 348 | #print('HTML5 attrs %r' % counters.debug_attrs)
|
| 349 |
|
| 350 | def testValidXml(self):
|
| 351 | # type: () -> None
|
| 352 | counters = html.Counters()
|
| 353 | for s in VALID_XML:
|
| 354 | print('XML %r' % s)
|
| 355 | html.Validate(s, html.BALANCED_TAGS | html.NO_SPECIAL_TAGS,
|
| 356 | counters)
|
| 357 | #print('XML attrs %r' % counters.debug_attrs)
|
| 358 |
|
| 359 |
|
| 360 | class XmlTest(unittest.TestCase):
|
| 361 |
|
| 362 | def testValid(self):
|
| 363 | # type: () -> None
|
| 364 | counters = html.Counters()
|
| 365 | for h, expected_xml in VALID_LEX + VALID_PARSE:
|
| 366 | actual = html.ToXml(h)
|
| 367 | if expected_xml == UNCHANGED: # Unchanged
|
| 368 | self.assertEqual(h, actual)
|
| 369 | elif expected_xml == '': # Skip
|
| 370 | pass
|
| 371 | else:
|
| 372 | self.assertEqual(expected_xml, actual)
|
| 373 |
|
| 374 |
|
| 375 | if __name__ == '__main__':
|
| 376 | unittest.main()
|