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