| 1 | #!/usr/bin/env python2
|
| 2 | from __future__ import print_function
|
| 3 |
|
| 4 | import re
|
| 5 | import unittest
|
| 6 |
|
| 7 | from lazylex import html # module under test log = html.log
|
| 8 | from doctools.util import log
|
| 9 |
|
| 10 |
|
| 11 | class RegexTest(unittest.TestCase):
|
| 12 |
|
| 13 | def testDotAll(self):
|
| 14 | # type: () -> None
|
| 15 |
|
| 16 | # Note that $ matches end of line, not end of string
|
| 17 | p1 = re.compile(r'.')
|
| 18 | print(p1.match('\n'))
|
| 19 |
|
| 20 | p2 = re.compile(r'.', re.DOTALL)
|
| 21 | print(p2.match('\n'))
|
| 22 |
|
| 23 | #p3 = re.compile(r'[.\n]', re.VERBOSE)
|
| 24 | p3 = re.compile(r'[.\n]')
|
| 25 | print(p3.match('\n'))
|
| 26 |
|
| 27 | print('Negation')
|
| 28 |
|
| 29 | p4 = re.compile(r'[^>]')
|
| 30 | print(p4.match('\n'))
|
| 31 |
|
| 32 | def testAttrRe(self):
|
| 33 | # type: () -> None
|
| 34 | _ATTR_RE = html._ATTR_RE
|
| 35 | m = _ATTR_RE.match(' empty= val')
|
| 36 | print(m.groups())
|
| 37 |
|
| 38 |
|
| 39 | class FunctionsTest(unittest.TestCase):
|
| 40 |
|
| 41 | def testToText(self):
|
| 42 | # type: () -> None
|
| 43 | t = html.ToText('<b name="&"> three < four && five </b>')
|
| 44 | self.assertEqual(' three < four && five ', t)
|
| 45 |
|
| 46 |
|
| 47 | def _MakeTagLexer(s):
|
| 48 | # type: (str) -> html.TagLexer
|
| 49 | lex = html.TagLexer(s)
|
| 50 | lex.Reset(0, len(s))
|
| 51 | return lex
|
| 52 |
|
| 53 |
|
| 54 | def _PrintTokens(lex):
|
| 55 | # type: (html.TagLexer) -> None
|
| 56 | log('')
|
| 57 | log('tag = %r', lex.GetTagName())
|
| 58 | for tok, start, end in lex.Tokens():
|
| 59 | log('%s %r', tok, lex.s[start:end])
|
| 60 |
|
| 61 |
|
| 62 | class TagLexerTest(unittest.TestCase):
|
| 63 |
|
| 64 | def testTagName_DEPRECATED(self):
|
| 65 | # type: () -> None
|
| 66 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 67 | self.assertEqual('a', lex.GetTagName())
|
| 68 |
|
| 69 | def testGetAttrRaw(self):
|
| 70 | # type: () -> None
|
| 71 | lex = _MakeTagLexer('<a>')
|
| 72 | _PrintTokens(lex)
|
| 73 | self.assertEqual(None, lex.GetAttrRaw('oops'))
|
| 74 |
|
| 75 | # <a novalue> means lex.Get('novalue') == ''
|
| 76 | # https://developer.mozilla.org/en-US/docs/Web/API/Element/hasAttribute
|
| 77 | # We are not distinguishing <a novalue=""> from <a novalue> in this API
|
| 78 | lex = _MakeTagLexer('<a novalue>')
|
| 79 | _PrintTokens(lex)
|
| 80 | self.assertEqual('', lex.GetAttrRaw('novalue'))
|
| 81 |
|
| 82 | lex = _MakeTagLexer('<a href="double quoted">')
|
| 83 | _PrintTokens(lex)
|
| 84 |
|
| 85 | self.assertEqual('double quoted', lex.GetAttrRaw('href'))
|
| 86 | self.assertEqual(None, lex.GetAttrRaw('oops'))
|
| 87 |
|
| 88 | lex = _MakeTagLexer('<a href=foo class="bar">')
|
| 89 | _PrintTokens(lex)
|
| 90 | self.assertEqual('bar', lex.GetAttrRaw('class'))
|
| 91 |
|
| 92 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 93 | _PrintTokens(lex)
|
| 94 | self.assertEqual('bar', lex.GetAttrRaw('class'))
|
| 95 |
|
| 96 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 97 | self.assertEqual('?foo=1&bar=2', lex.GetAttrRaw('href'))
|
| 98 |
|
| 99 | def testAllAttrs(self):
|
| 100 | # type: () -> None
|
| 101 | """
|
| 102 | [('key', 'value')] for all
|
| 103 | """
|
| 104 | # closed
|
| 105 | lex = _MakeTagLexer('<a href=foo class="bar" />')
|
| 106 | self.assertEqual([('href', 'foo'), ('class', 'bar')],
|
| 107 | lex.AllAttrsRaw())
|
| 108 |
|
| 109 | lex = _MakeTagLexer('<a href="?foo=1&bar=2" />')
|
| 110 | self.assertEqual([('href', '?foo=1&bar=2')], lex.AllAttrsRaw())
|
| 111 |
|
| 112 | def testEmptyMissingValues(self):
|
| 113 | # type: () -> None
|
| 114 | # equivalent to <button disabled="">
|
| 115 | lex = _MakeTagLexer('<button disabled>')
|
| 116 | all_attrs = lex.AllAttrsRaw()
|
| 117 | self.assertEqual([('disabled', '')], all_attrs)
|
| 118 |
|
| 119 | slices = lex.AllAttrsRawSlice()
|
| 120 | log('slices %s', slices)
|
| 121 |
|
| 122 | lex = _MakeTagLexer(
|
| 123 | '''<p double="" single='' empty= value missing empty2=>''')
|
| 124 | all_attrs = lex.AllAttrsRaw()
|
| 125 | self.assertEqual([
|
| 126 | ('double', ''),
|
| 127 | ('single', ''),
|
| 128 | ('empty', 'value'),
|
| 129 | ('missing', ''),
|
| 130 | ('empty2', ''),
|
| 131 | ], all_attrs)
|
| 132 | # TODO: should have
|
| 133 | log('all %s', all_attrs)
|
| 134 |
|
| 135 | slices = lex.AllAttrsRawSlice()
|
| 136 | log('slices %s', slices)
|
| 137 |
|
| 138 | def testInvalidTag(self):
|
| 139 | # type: () -> None
|
| 140 | try:
|
| 141 | lex = _MakeTagLexer('<a foo=bar !></a>')
|
| 142 | all_attrs = lex.AllAttrsRaw()
|
| 143 | except html.LexError as e:
|
| 144 | print(e)
|
| 145 | else:
|
| 146 | self.fail('Expected LexError')
|
| 147 |
|
| 148 |
|
| 149 | class LexerTest(unittest.TestCase):
|
| 150 |
|
| 151 | def testInvalid(self):
|
| 152 | # type: () -> None
|
| 153 | from data_lang.htm8_test import ValidTokenList
|
| 154 | for s in INVALID_LEX:
|
| 155 | try:
|
| 156 | tokens = ValidTokenList(s)
|
| 157 | except html.LexError as e:
|
| 158 | print(e)
|
| 159 | else:
|
| 160 | self.fail('Expected LexError %r' % s)
|
| 161 |
|
| 162 | def testValid(self):
|
| 163 | # type: () -> None
|
| 164 |
|
| 165 | from data_lang.htm8_test import Lex
|
| 166 |
|
| 167 | for s, _ in VALID_LEX:
|
| 168 | tokens = Lex(s)
|
| 169 | print()
|
| 170 |
|
| 171 |
|
| 172 | INVALID_LEX = [
|
| 173 | '< >',
|
| 174 | '<a><',
|
| 175 | '&<',
|
| 176 | '&<',
|
| 177 | # Hm > is allowed?
|
| 178 | #'a > b',
|
| 179 | 'a < b',
|
| 180 | '<!-- unfinished comment',
|
| 181 | '<? unfinished processing',
|
| 182 | '</div bad=attr> <a> <b>',
|
| 183 |
|
| 184 | # not allowed, but 3 > 4 is allowed
|
| 185 | '<a> 3 < 4 </a>',
|
| 186 | # Not a CDATA tag
|
| 187 | '<STYLEz><</STYLEz>',
|
| 188 | ]
|
| 189 |
|
| 190 | SKIP = 0
|
| 191 | UNCHANGED = 1
|
| 192 |
|
| 193 | VALID_LEX = [
|
| 194 | # TODO: convert these to XML
|
| 195 | ('<foo></foo>', UNCHANGED),
|
| 196 | ('<foo x=y></foo>', ''),
|
| 197 | #('<foo x="&"></foo>', '<foo x="&"></foo>'),
|
| 198 | ('<foo x="&"></foo>', ''),
|
| 199 |
|
| 200 | # Allowed with BadAmpersand
|
| 201 | ('<p> x & y </p>', '<p> x & y </p>'),
|
| 202 |
|
| 203 | # No ambiguity
|
| 204 | ('<img src=/ >', ''),
|
| 205 | ('<img src="/">', ''),
|
| 206 | ('<img src=foo/ >', ''),
|
| 207 | ]
|
| 208 |
|
| 209 | INVALID_PARSE = [
|
| 210 | '<a></b>',
|
| 211 | '<a>', # missing closing tag
|
| 212 | '<meta></meta>', # this is a self-closing tag
|
| 213 | ]
|
| 214 |
|
| 215 | INVALID_ATTR_LEX = [
|
| 216 | # Ambiguous, should be ""
|
| 217 | '<img src=/>',
|
| 218 | '<img src= />',
|
| 219 | '<img src=foo/>',
|
| 220 | '<img src= foo/>',
|
| 221 |
|
| 222 | # Quoting
|
| 223 | '<img src=x"y">',
|
| 224 | "<img src=j''>",
|
| 225 | ]
|
| 226 |
|
| 227 | VALID_PARSE = [
|
| 228 | ('<!DOCTYPE html>\n', ''),
|
| 229 | ('<!DOCTYPE>', ''),
|
| 230 |
|
| 231 | # empty strings
|
| 232 | ('<p x=""></p>', UNCHANGED),
|
| 233 | ("<p x=''></p>", UNCHANGED),
|
| 234 | ('<self-closing a="b" />', UNCHANGED),
|
| 235 |
|
| 236 | # We could also normalize CDATA?
|
| 237 | # Note that CDATA has an escaping problem: you need to handle it ]]> with
|
| 238 | # concatenation. It just "pushes the problem around".
|
| 239 | # So I think it's better to use ONE kind of escaping, which is <
|
| 240 | ('<script><![CDATA[ <wtf> >< ]]></script>', UNCHANGED),
|
| 241 |
|
| 242 | # allowed, but 3 < 4 is not allowed
|
| 243 | ('<a> 3 > 4 </a>', '<a> 3 > 4 </a>'),
|
| 244 | # allowed, but 3 > 4 is not allowed
|
| 245 | ('<p x="3 < 4"></p>', ''),
|
| 246 | ('<b><a href="foo">link</a></b>', UNCHANGED),
|
| 247 |
|
| 248 | # TODO: should be self-closing
|
| 249 | #('<meta><a></a>', '<meta/><a></a>'),
|
| 250 | ('<meta><a></a>', ''),
|
| 251 |
|
| 252 | # no attribute
|
| 253 | ('<button disabled></button>', ''),
|
| 254 | ('<button disabled=></button>', ''),
|
| 255 | ('<button disabled= ></button>', ''),
|
| 256 |
|
| 257 | # single quoted is pretty common
|
| 258 | ("<a href='single'></a>", ''),
|
| 259 |
|
| 260 | # Conceding to reality - I used these myself
|
| 261 | ('<a href=ble.sh></a>', ''),
|
| 262 | ('<a href=foo.html></a>', ''),
|
| 263 | ('<foo x="&"></foo>', ''),
|
| 264 |
|
| 265 | # caps
|
| 266 | ('<foo></FOO>', ''),
|
| 267 | ('<Foo></fOO>', ''),
|
| 268 |
|
| 269 | # capital VOID tag
|
| 270 | ('<META><a></a>', ''),
|
| 271 | ('<script><</script>', ''),
|
| 272 | # matching
|
| 273 | ('<SCRipt><</SCRipt>', ''),
|
| 274 | ('<SCRIPT><</SCRIPT>', ''),
|
| 275 | ('<STYLE><</STYLE>', ''),
|
| 276 | #'<SCRipt><</script>',
|
| 277 |
|
| 278 | # Regression test from blog
|
| 279 | ('<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>',
|
| 280 | '')
|
| 281 |
|
| 282 | # Note: Python HTMLParser.py does DYNAMIC compilation of regex with re.I
|
| 283 | # flag to handle this! Gah I want something faster.
|
| 284 | #'<script><</SCRIPT>',
|
| 285 |
|
| 286 | # TODO: Test <svg> and <math> ?
|
| 287 | ]
|
| 288 |
|
| 289 | VALID_XML = [
|
| 290 | '<meta></meta>',
|
| 291 | ]
|
| 292 |
|
| 293 | INVALID_TAG_LEX = [
|
| 294 | # not allowed, but 3 < 4 is allowed
|
| 295 | '<p x="3 > 4"></p>',
|
| 296 | # same thing
|
| 297 | '<a href=">"></a>',
|
| 298 | '<a foo=bar !></a>', # bad attr
|
| 299 | ]
|
| 300 |
|
| 301 |
|
| 302 | class ValidateTest(unittest.TestCase):
|
| 303 |
|
| 304 | def testInvalid(self):
|
| 305 | # type: () -> None
|
| 306 | counters = html.Counters()
|
| 307 | for s in INVALID_LEX + INVALID_TAG_LEX + INVALID_ATTR_LEX:
|
| 308 | try:
|
| 309 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 310 | except html.LexError as e:
|
| 311 | print(e)
|
| 312 | else:
|
| 313 | self.fail('Expected LexError %r' % s)
|
| 314 |
|
| 315 | for s in INVALID_PARSE:
|
| 316 | try:
|
| 317 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 318 | except html.ParseError as e:
|
| 319 | print(e)
|
| 320 | else:
|
| 321 | self.fail('Expected ParseError')
|
| 322 |
|
| 323 | def testValid(self):
|
| 324 | # type: () -> None
|
| 325 | counters = html.Counters()
|
| 326 | for s, _ in VALID_PARSE:
|
| 327 | print('HTML5 %r' % s)
|
| 328 | html.Validate(s, html.BALANCED_TAGS, counters)
|
| 329 | #print('HTML5 attrs %r' % counters.debug_attrs)
|
| 330 |
|
| 331 | def testValidXml(self):
|
| 332 | # type: () -> None
|
| 333 | counters = html.Counters()
|
| 334 | for s in VALID_XML:
|
| 335 | print('XML %r' % s)
|
| 336 | html.Validate(s, html.BALANCED_TAGS | html.NO_SPECIAL_TAGS,
|
| 337 | counters)
|
| 338 | #print('XML attrs %r' % counters.debug_attrs)
|
| 339 |
|
| 340 |
|
| 341 | class XmlTest(unittest.TestCase):
|
| 342 |
|
| 343 | def testValid(self):
|
| 344 | # type: () -> None
|
| 345 | counters = html.Counters()
|
| 346 | for h, expected_xml in VALID_LEX + VALID_PARSE:
|
| 347 | actual = html.ToXml(h)
|
| 348 | if expected_xml == UNCHANGED: # Unchanged
|
| 349 | self.assertEqual(h, actual)
|
| 350 | elif expected_xml == '': # Skip
|
| 351 | pass
|
| 352 | else:
|
| 353 | self.assertEqual(expected_xml, actual)
|
| 354 |
|
| 355 |
|
| 356 | if __name__ == '__main__':
|
| 357 | unittest.main()
|