A Python parser for MediaWiki wikicode https://mwparserfromhell.readthedocs.io/
25개 이상의 토픽을 선택하실 수 없습니다. Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 

182 lines
6.6 KiB

  1. # -*- coding: utf-8 -*-
  2. #
  3. # Copyright (C) 2012-2013 Ben Kurtovic <ben.kurtovic@verizon.net>
  4. #
  5. # Permission is hereby granted, free of charge, to any person obtaining a copy
  6. # of this software and associated documentation files (the "Software"), to deal
  7. # in the Software without restriction, including without limitation the rights
  8. # to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  9. # copies of the Software, and to permit persons to whom the Software is
  10. # furnished to do so, subject to the following conditions:
  11. #
  12. # The above copyright notice and this permission notice shall be included in
  13. # all copies or substantial portions of the Software.
  14. #
  15. # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  16. # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  17. # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  18. # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  19. # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  20. # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  21. # SOFTWARE.
  22. from __future__ import unicode_literals
  23. from . import Node
  24. from ..compat import htmlentities, py3k, str
  25. __all__ = ["HTMLEntity"]
  26. class HTMLEntity(Node):
  27. """Represents an HTML entity, like ``&nbsp;``, either named or unnamed."""
  28. def __init__(self, value, named=None, hexadecimal=False, hex_char="x"):
  29. super(HTMLEntity, self).__init__()
  30. self._value = value
  31. if named is None: # Try to guess whether or not the entity is named
  32. try:
  33. int(value)
  34. self._named = False
  35. self._hexadecimal = False
  36. except ValueError:
  37. try:
  38. int(value, 16)
  39. self._named = False
  40. self._hexadecimal = True
  41. except ValueError:
  42. self._named = True
  43. self._hexadecimal = False
  44. else:
  45. self._named = named
  46. self._hexadecimal = hexadecimal
  47. self._hex_char = hex_char
  48. def __unicode__(self):
  49. if self.named:
  50. return "&{0};".format(self.value)
  51. if self.hexadecimal:
  52. return "&#{0}{1};".format(self.hex_char, self.value)
  53. return "&#{0};".format(self.value)
  54. def __strip__(self, normalize, collapse):
  55. if normalize:
  56. return self.normalize()
  57. return self
  58. if not py3k:
  59. @staticmethod
  60. def _unichr(value):
  61. """Implement builtin unichr() with support for non-BMP code points.
  62. On wide Python builds, this functions like the normal unichr(). On
  63. narrow builds, this returns the value's encoded surrogate pair.
  64. """
  65. try:
  66. return unichr(value)
  67. except ValueError:
  68. # Test whether we're on the wide or narrow Python build. Check
  69. # the length of a non-BMP code point
  70. # (U+1F64A, SPEAK-NO-EVIL MONKEY):
  71. if len("\U0001F64A") == 2:
  72. # Ensure this is within the range we can encode:
  73. if value > 0x10FFFF:
  74. raise ValueError("unichr() arg not in range(0x110000)")
  75. code = value - 0x10000
  76. if value < 0: # Invalid code point
  77. raise
  78. lead = 0xD800 + (code >> 10)
  79. trail = 0xDC00 + (code % (1 << 10))
  80. return unichr(lead) + unichr(trail)
  81. raise
  82. @property
  83. def value(self):
  84. """The string value of the HTML entity."""
  85. return self._value
  86. @property
  87. def named(self):
  88. """Whether the entity is a string name for a codepoint or an integer.
  89. For example, ``&Sigma;``, ``&#931;``, and ``&#x3a3;`` refer to the same
  90. character, but only the first is "named", while the others are integer
  91. representations of the codepoint.
  92. """
  93. return self._named
  94. @property
  95. def hexadecimal(self):
  96. """If unnamed, this is whether the value is hexadecimal or decimal."""
  97. return self._hexadecimal
  98. @property
  99. def hex_char(self):
  100. """If the value is hexadecimal, this is the letter denoting that.
  101. For example, the hex_char of ``"&#x1234;"`` is ``"x"``, whereas the
  102. hex_char of ``"&#X1234;"`` is ``"X"``. Lowercase and uppercase ``x``
  103. are the only values supported.
  104. """
  105. return self._hex_char
  106. @value.setter
  107. def value(self, newval):
  108. newval = str(newval)
  109. try:
  110. int(newval)
  111. except ValueError:
  112. try:
  113. int(newval, 16)
  114. except ValueError:
  115. if newval not in htmlentities.entitydefs:
  116. raise ValueError("entity value is not a valid name")
  117. self._named = True
  118. self._hexadecimal = False
  119. else:
  120. if int(newval, 16) < 0 or int(newval, 16) > 0x10FFFF:
  121. raise ValueError("entity value is not in range(0x110000)")
  122. self._named = False
  123. self._hexadecimal = True
  124. else:
  125. test = int(newval, 16 if self.hexadecimal else 10)
  126. if test < 0 or test > 0x10FFFF:
  127. raise ValueError("entity value is not in range(0x110000)")
  128. self._named = False
  129. self._value = newval
  130. @named.setter
  131. def named(self, newval):
  132. newval = bool(newval)
  133. if newval and self.value not in htmlentities.entitydefs:
  134. raise ValueError("entity value is not a valid name")
  135. if not newval:
  136. try:
  137. int(self.value, 16)
  138. except ValueError:
  139. err = "current entity value is not a valid Unicode codepoint"
  140. raise ValueError(err)
  141. self._named = newval
  142. @hexadecimal.setter
  143. def hexadecimal(self, newval):
  144. newval = bool(newval)
  145. if newval and self.named:
  146. raise ValueError("a named entity cannot be hexadecimal")
  147. self._hexadecimal = newval
  148. @hex_char.setter
  149. def hex_char(self, newval):
  150. newval = str(newval)
  151. if newval not in ("x", "X"):
  152. raise ValueError(newval)
  153. self._hex_char = newval
  154. def normalize(self):
  155. """Return the unicode character represented by the HTML entity."""
  156. chrfunc = chr if py3k else HTMLEntity._unichr
  157. if self.named:
  158. return chrfunc(htmlentities.name2codepoint[self.value])
  159. if self.hexadecimal:
  160. return chrfunc(int(self.value, 16))
  161. return chrfunc(int(self.value))