You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

htmlentitydefs.py 19KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257
  1. """HTML character entity references."""
  2. entitydefs = {
  3. 'AElig': '\306', # latin capital letter AE = latin capital ligature AE, U+00C6 ISOlat1
  4. 'Aacute': '\301', # latin capital letter A with acute, U+00C1 ISOlat1
  5. 'Acirc': '\302', # latin capital letter A with circumflex, U+00C2 ISOlat1
  6. 'Agrave': '\300', # latin capital letter A with grave = latin capital letter A grave, U+00C0 ISOlat1
  7. 'Alpha': 'Α', # greek capital letter alpha, U+0391
  8. 'Aring': '\305', # latin capital letter A with ring above = latin capital letter A ring, U+00C5 ISOlat1
  9. 'Atilde': '\303', # latin capital letter A with tilde, U+00C3 ISOlat1
  10. 'Auml': '\304', # latin capital letter A with diaeresis, U+00C4 ISOlat1
  11. 'Beta': 'Β', # greek capital letter beta, U+0392
  12. 'Ccedil': '\307', # latin capital letter C with cedilla, U+00C7 ISOlat1
  13. 'Chi': 'Χ', # greek capital letter chi, U+03A7
  14. 'Dagger': '‡', # double dagger, U+2021 ISOpub
  15. 'Delta': 'Δ', # greek capital letter delta, U+0394 ISOgrk3
  16. 'ETH': '\320', # latin capital letter ETH, U+00D0 ISOlat1
  17. 'Eacute': '\311', # latin capital letter E with acute, U+00C9 ISOlat1
  18. 'Ecirc': '\312', # latin capital letter E with circumflex, U+00CA ISOlat1
  19. 'Egrave': '\310', # latin capital letter E with grave, U+00C8 ISOlat1
  20. 'Epsilon': 'Ε', # greek capital letter epsilon, U+0395
  21. 'Eta': 'Η', # greek capital letter eta, U+0397
  22. 'Euml': '\313', # latin capital letter E with diaeresis, U+00CB ISOlat1
  23. 'Gamma': 'Γ', # greek capital letter gamma, U+0393 ISOgrk3
  24. 'Iacute': '\315', # latin capital letter I with acute, U+00CD ISOlat1
  25. 'Icirc': '\316', # latin capital letter I with circumflex, U+00CE ISOlat1
  26. 'Igrave': '\314', # latin capital letter I with grave, U+00CC ISOlat1
  27. 'Iota': 'Ι', # greek capital letter iota, U+0399
  28. 'Iuml': '\317', # latin capital letter I with diaeresis, U+00CF ISOlat1
  29. 'Kappa': 'Κ', # greek capital letter kappa, U+039A
  30. 'Lambda': 'Λ', # greek capital letter lambda, U+039B ISOgrk3
  31. 'Mu': 'Μ', # greek capital letter mu, U+039C
  32. 'Ntilde': '\321', # latin capital letter N with tilde, U+00D1 ISOlat1
  33. 'Nu': 'Ν', # greek capital letter nu, U+039D
  34. 'OElig': 'Œ', # latin capital ligature OE, U+0152 ISOlat2
  35. 'Oacute': '\323', # latin capital letter O with acute, U+00D3 ISOlat1
  36. 'Ocirc': '\324', # latin capital letter O with circumflex, U+00D4 ISOlat1
  37. 'Ograve': '\322', # latin capital letter O with grave, U+00D2 ISOlat1
  38. 'Omega': 'Ω', # greek capital letter omega, U+03A9 ISOgrk3
  39. 'Omicron': 'Ο', # greek capital letter omicron, U+039F
  40. 'Oslash': '\330', # latin capital letter O with stroke = latin capital letter O slash, U+00D8 ISOlat1
  41. 'Otilde': '\325', # latin capital letter O with tilde, U+00D5 ISOlat1
  42. 'Ouml': '\326', # latin capital letter O with diaeresis, U+00D6 ISOlat1
  43. 'Phi': 'Φ', # greek capital letter phi, U+03A6 ISOgrk3
  44. 'Pi': 'Π', # greek capital letter pi, U+03A0 ISOgrk3
  45. 'Prime': '″', # double prime = seconds = inches, U+2033 ISOtech
  46. 'Psi': 'Ψ', # greek capital letter psi, U+03A8 ISOgrk3
  47. 'Rho': 'Ρ', # greek capital letter rho, U+03A1
  48. 'Scaron': 'Š', # latin capital letter S with caron, U+0160 ISOlat2
  49. 'Sigma': 'Σ', # greek capital letter sigma, U+03A3 ISOgrk3
  50. 'THORN': '\336', # latin capital letter THORN, U+00DE ISOlat1
  51. 'Tau': 'Τ', # greek capital letter tau, U+03A4
  52. 'Theta': 'Θ', # greek capital letter theta, U+0398 ISOgrk3
  53. 'Uacute': '\332', # latin capital letter U with acute, U+00DA ISOlat1
  54. 'Ucirc': '\333', # latin capital letter U with circumflex, U+00DB ISOlat1
  55. 'Ugrave': '\331', # latin capital letter U with grave, U+00D9 ISOlat1
  56. 'Upsilon': 'Υ', # greek capital letter upsilon, U+03A5 ISOgrk3
  57. 'Uuml': '\334', # latin capital letter U with diaeresis, U+00DC ISOlat1
  58. 'Xi': 'Ξ', # greek capital letter xi, U+039E ISOgrk3
  59. 'Yacute': '\335', # latin capital letter Y with acute, U+00DD ISOlat1
  60. 'Yuml': 'Ÿ', # latin capital letter Y with diaeresis, U+0178 ISOlat2
  61. 'Zeta': 'Ζ', # greek capital letter zeta, U+0396
  62. 'aacute': '\341', # latin small letter a with acute, U+00E1 ISOlat1
  63. 'acirc': '\342', # latin small letter a with circumflex, U+00E2 ISOlat1
  64. 'acute': '\264', # acute accent = spacing acute, U+00B4 ISOdia
  65. 'aelig': '\346', # latin small letter ae = latin small ligature ae, U+00E6 ISOlat1
  66. 'agrave': '\340', # latin small letter a with grave = latin small letter a grave, U+00E0 ISOlat1
  67. 'alefsym': 'ℵ', # alef symbol = first transfinite cardinal, U+2135 NEW
  68. 'alpha': 'α', # greek small letter alpha, U+03B1 ISOgrk3
  69. 'amp': '\46', # ampersand, U+0026 ISOnum
  70. 'and': '∧', # logical and = wedge, U+2227 ISOtech
  71. 'ang': '∠', # angle, U+2220 ISOamso
  72. 'aring': '\345', # latin small letter a with ring above = latin small letter a ring, U+00E5 ISOlat1
  73. 'asymp': '≈', # almost equal to = asymptotic to, U+2248 ISOamsr
  74. 'atilde': '\343', # latin small letter a with tilde, U+00E3 ISOlat1
  75. 'auml': '\344', # latin small letter a with diaeresis, U+00E4 ISOlat1
  76. 'bdquo': '„', # double low-9 quotation mark, U+201E NEW
  77. 'beta': 'β', # greek small letter beta, U+03B2 ISOgrk3
  78. 'brvbar': '\246', # broken bar = broken vertical bar, U+00A6 ISOnum
  79. 'bull': '•', # bullet = black small circle, U+2022 ISOpub
  80. 'cap': '∩', # intersection = cap, U+2229 ISOtech
  81. 'ccedil': '\347', # latin small letter c with cedilla, U+00E7 ISOlat1
  82. 'cedil': '\270', # cedilla = spacing cedilla, U+00B8 ISOdia
  83. 'cent': '\242', # cent sign, U+00A2 ISOnum
  84. 'chi': 'χ', # greek small letter chi, U+03C7 ISOgrk3
  85. 'circ': 'ˆ', # modifier letter circumflex accent, U+02C6 ISOpub
  86. 'clubs': '♣', # black club suit = shamrock, U+2663 ISOpub
  87. 'cong': '≅', # approximately equal to, U+2245 ISOtech
  88. 'copy': '\251', # copyright sign, U+00A9 ISOnum
  89. 'crarr': '↵', # downwards arrow with corner leftwards = carriage return, U+21B5 NEW
  90. 'cup': '∪', # union = cup, U+222A ISOtech
  91. 'curren': '\244', # currency sign, U+00A4 ISOnum
  92. 'dArr': '⇓', # downwards double arrow, U+21D3 ISOamsa
  93. 'dagger': '†', # dagger, U+2020 ISOpub
  94. 'darr': '↓', # downwards arrow, U+2193 ISOnum
  95. 'deg': '\260', # degree sign, U+00B0 ISOnum
  96. 'delta': 'δ', # greek small letter delta, U+03B4 ISOgrk3
  97. 'diams': '♦', # black diamond suit, U+2666 ISOpub
  98. 'divide': '\367', # division sign, U+00F7 ISOnum
  99. 'eacute': '\351', # latin small letter e with acute, U+00E9 ISOlat1
  100. 'ecirc': '\352', # latin small letter e with circumflex, U+00EA ISOlat1
  101. 'egrave': '\350', # latin small letter e with grave, U+00E8 ISOlat1
  102. 'empty': '∅', # empty set = null set = diameter, U+2205 ISOamso
  103. 'emsp': ' ', # em space, U+2003 ISOpub
  104. 'ensp': ' ', # en space, U+2002 ISOpub
  105. 'epsilon': 'ε', # greek small letter epsilon, U+03B5 ISOgrk3
  106. 'equiv': '≡', # identical to, U+2261 ISOtech
  107. 'eta': 'η', # greek small letter eta, U+03B7 ISOgrk3
  108. 'eth': '\360', # latin small letter eth, U+00F0 ISOlat1
  109. 'euml': '\353', # latin small letter e with diaeresis, U+00EB ISOlat1
  110. 'euro': '€', # euro sign, U+20AC NEW
  111. 'exist': '∃', # there exists, U+2203 ISOtech
  112. 'fnof': 'ƒ', # latin small f with hook = function = florin, U+0192 ISOtech
  113. 'forall': '∀', # for all, U+2200 ISOtech
  114. 'frac12': '\275', # vulgar fraction one half = fraction one half, U+00BD ISOnum
  115. 'frac14': '\274', # vulgar fraction one quarter = fraction one quarter, U+00BC ISOnum
  116. 'frac34': '\276', # vulgar fraction three quarters = fraction three quarters, U+00BE ISOnum
  117. 'frasl': '⁄', # fraction slash, U+2044 NEW
  118. 'gamma': 'γ', # greek small letter gamma, U+03B3 ISOgrk3
  119. 'ge': '≥', # greater-than or equal to, U+2265 ISOtech
  120. 'gt': '\76', # greater-than sign, U+003E ISOnum
  121. 'hArr': '⇔', # left right double arrow, U+21D4 ISOamsa
  122. 'harr': '↔', # left right arrow, U+2194 ISOamsa
  123. 'hearts': '♥', # black heart suit = valentine, U+2665 ISOpub
  124. 'hellip': '…', # horizontal ellipsis = three dot leader, U+2026 ISOpub
  125. 'iacute': '\355', # latin small letter i with acute, U+00ED ISOlat1
  126. 'icirc': '\356', # latin small letter i with circumflex, U+00EE ISOlat1
  127. 'iexcl': '\241', # inverted exclamation mark, U+00A1 ISOnum
  128. 'igrave': '\354', # latin small letter i with grave, U+00EC ISOlat1
  129. 'image': 'ℑ', # blackletter capital I = imaginary part, U+2111 ISOamso
  130. 'infin': '∞', # infinity, U+221E ISOtech
  131. 'int': '∫', # integral, U+222B ISOtech
  132. 'iota': 'ι', # greek small letter iota, U+03B9 ISOgrk3
  133. 'iquest': '\277', # inverted question mark = turned question mark, U+00BF ISOnum
  134. 'isin': '∈', # element of, U+2208 ISOtech
  135. 'iuml': '\357', # latin small letter i with diaeresis, U+00EF ISOlat1
  136. 'kappa': 'κ', # greek small letter kappa, U+03BA ISOgrk3
  137. 'lArr': '⇐', # leftwards double arrow, U+21D0 ISOtech
  138. 'lambda': 'λ', # greek small letter lambda, U+03BB ISOgrk3
  139. 'lang': '〈', # left-pointing angle bracket = bra, U+2329 ISOtech
  140. 'laquo': '\253', # left-pointing double angle quotation mark = left pointing guillemet, U+00AB ISOnum
  141. 'larr': '←', # leftwards arrow, U+2190 ISOnum
  142. 'lceil': '⌈', # left ceiling = apl upstile, U+2308 ISOamsc
  143. 'ldquo': '“', # left double quotation mark, U+201C ISOnum
  144. 'le': '≤', # less-than or equal to, U+2264 ISOtech
  145. 'lfloor': '⌊', # left floor = apl downstile, U+230A ISOamsc
  146. 'lowast': '∗', # asterisk operator, U+2217 ISOtech
  147. 'loz': '◊', # lozenge, U+25CA ISOpub
  148. 'lrm': '‎', # left-to-right mark, U+200E NEW RFC 2070
  149. 'lsaquo': '‹', # single left-pointing angle quotation mark, U+2039 ISO proposed
  150. 'lsquo': '‘', # left single quotation mark, U+2018 ISOnum
  151. 'lt': '\74', # less-than sign, U+003C ISOnum
  152. 'macr': '\257', # macron = spacing macron = overline = APL overbar, U+00AF ISOdia
  153. 'mdash': '—', # em dash, U+2014 ISOpub
  154. 'micro': '\265', # micro sign, U+00B5 ISOnum
  155. 'middot': '\267', # middle dot = Georgian comma = Greek middle dot, U+00B7 ISOnum
  156. 'minus': '−', # minus sign, U+2212 ISOtech
  157. 'mu': 'μ', # greek small letter mu, U+03BC ISOgrk3
  158. 'nabla': '∇', # nabla = backward difference, U+2207 ISOtech
  159. 'nbsp': '\240', # no-break space = non-breaking space, U+00A0 ISOnum
  160. 'ndash': '–', # en dash, U+2013 ISOpub
  161. 'ne': '≠', # not equal to, U+2260 ISOtech
  162. 'ni': '∋', # contains as member, U+220B ISOtech
  163. 'not': '\254', # not sign, U+00AC ISOnum
  164. 'notin': '∉', # not an element of, U+2209 ISOtech
  165. 'nsub': '⊄', # not a subset of, U+2284 ISOamsn
  166. 'ntilde': '\361', # latin small letter n with tilde, U+00F1 ISOlat1
  167. 'nu': 'ν', # greek small letter nu, U+03BD ISOgrk3
  168. 'oacute': '\363', # latin small letter o with acute, U+00F3 ISOlat1
  169. 'ocirc': '\364', # latin small letter o with circumflex, U+00F4 ISOlat1
  170. 'oelig': 'œ', # latin small ligature oe, U+0153 ISOlat2
  171. 'ograve': '\362', # latin small letter o with grave, U+00F2 ISOlat1
  172. 'oline': '‾', # overline = spacing overscore, U+203E NEW
  173. 'omega': 'ω', # greek small letter omega, U+03C9 ISOgrk3
  174. 'omicron': 'ο', # greek small letter omicron, U+03BF NEW
  175. 'oplus': '⊕', # circled plus = direct sum, U+2295 ISOamsb
  176. 'or': '∨', # logical or = vee, U+2228 ISOtech
  177. 'ordf': '\252', # feminine ordinal indicator, U+00AA ISOnum
  178. 'ordm': '\272', # masculine ordinal indicator, U+00BA ISOnum
  179. 'oslash': '\370', # latin small letter o with stroke, = latin small letter o slash, U+00F8 ISOlat1
  180. 'otilde': '\365', # latin small letter o with tilde, U+00F5 ISOlat1
  181. 'otimes': '⊗', # circled times = vector product, U+2297 ISOamsb
  182. 'ouml': '\366', # latin small letter o with diaeresis, U+00F6 ISOlat1
  183. 'para': '\266', # pilcrow sign = paragraph sign, U+00B6 ISOnum
  184. 'part': '∂', # partial differential, U+2202 ISOtech
  185. 'permil': '‰', # per mille sign, U+2030 ISOtech
  186. 'perp': '⊥', # up tack = orthogonal to = perpendicular, U+22A5 ISOtech
  187. 'phi': 'φ', # greek small letter phi, U+03C6 ISOgrk3
  188. 'pi': 'π', # greek small letter pi, U+03C0 ISOgrk3
  189. 'piv': 'ϖ', # greek pi symbol, U+03D6 ISOgrk3
  190. 'plusmn': '\261', # plus-minus sign = plus-or-minus sign, U+00B1 ISOnum
  191. 'pound': '\243', # pound sign, U+00A3 ISOnum
  192. 'prime': '′', # prime = minutes = feet, U+2032 ISOtech
  193. 'prod': '∏', # n-ary product = product sign, U+220F ISOamsb
  194. 'prop': '∝', # proportional to, U+221D ISOtech
  195. 'psi': 'ψ', # greek small letter psi, U+03C8 ISOgrk3
  196. 'quot': '\42', # quotation mark = APL quote, U+0022 ISOnum
  197. 'rArr': '⇒', # rightwards double arrow, U+21D2 ISOtech
  198. 'radic': '√', # square root = radical sign, U+221A ISOtech
  199. 'rang': '〉', # right-pointing angle bracket = ket, U+232A ISOtech
  200. 'raquo': '\273', # right-pointing double angle quotation mark = right pointing guillemet, U+00BB ISOnum
  201. 'rarr': '→', # rightwards arrow, U+2192 ISOnum
  202. 'rceil': '⌉', # right ceiling, U+2309 ISOamsc
  203. 'rdquo': '”', # right double quotation mark, U+201D ISOnum
  204. 'real': 'ℜ', # blackletter capital R = real part symbol, U+211C ISOamso
  205. 'reg': '\256', # registered sign = registered trade mark sign, U+00AE ISOnum
  206. 'rfloor': '⌋', # right floor, U+230B ISOamsc
  207. 'rho': 'ρ', # greek small letter rho, U+03C1 ISOgrk3
  208. 'rlm': '‏', # right-to-left mark, U+200F NEW RFC 2070
  209. 'rsaquo': '›', # single right-pointing angle quotation mark, U+203A ISO proposed
  210. 'rsquo': '’', # right single quotation mark, U+2019 ISOnum
  211. 'sbquo': '‚', # single low-9 quotation mark, U+201A NEW
  212. 'scaron': 'š', # latin small letter s with caron, U+0161 ISOlat2
  213. 'sdot': '⋅', # dot operator, U+22C5 ISOamsb
  214. 'sect': '\247', # section sign, U+00A7 ISOnum
  215. 'shy': '\255', # soft hyphen = discretionary hyphen, U+00AD ISOnum
  216. 'sigma': 'σ', # greek small letter sigma, U+03C3 ISOgrk3
  217. 'sigmaf': 'ς', # greek small letter final sigma, U+03C2 ISOgrk3
  218. 'sim': '∼', # tilde operator = varies with = similar to, U+223C ISOtech
  219. 'spades': '♠', # black spade suit, U+2660 ISOpub
  220. 'sub': '⊂', # subset of, U+2282 ISOtech
  221. 'sube': '⊆', # subset of or equal to, U+2286 ISOtech
  222. 'sum': '∑', # n-ary sumation, U+2211 ISOamsb
  223. 'sup': '⊃', # superset of, U+2283 ISOtech
  224. 'sup1': '\271', # superscript one = superscript digit one, U+00B9 ISOnum
  225. 'sup2': '\262', # superscript two = superscript digit two = squared, U+00B2 ISOnum
  226. 'sup3': '\263', # superscript three = superscript digit three = cubed, U+00B3 ISOnum
  227. 'supe': '⊇', # superset of or equal to, U+2287 ISOtech
  228. 'szlig': '\337', # latin small letter sharp s = ess-zed, U+00DF ISOlat1
  229. 'tau': 'τ', # greek small letter tau, U+03C4 ISOgrk3
  230. 'there4': '∴', # therefore, U+2234 ISOtech
  231. 'theta': 'θ', # greek small letter theta, U+03B8 ISOgrk3
  232. 'thetasym': 'ϑ', # greek small letter theta symbol, U+03D1 NEW
  233. 'thinsp': ' ', # thin space, U+2009 ISOpub
  234. 'thorn': '\376', # latin small letter thorn with, U+00FE ISOlat1
  235. 'tilde': '˜', # small tilde, U+02DC ISOdia
  236. 'times': '\327', # multiplication sign, U+00D7 ISOnum
  237. 'trade': '™', # trade mark sign, U+2122 ISOnum
  238. 'uArr': '⇑', # upwards double arrow, U+21D1 ISOamsa
  239. 'uacute': '\372', # latin small letter u with acute, U+00FA ISOlat1
  240. 'uarr': '↑', # upwards arrow, U+2191 ISOnum
  241. 'ucirc': '\373', # latin small letter u with circumflex, U+00FB ISOlat1
  242. 'ugrave': '\371', # latin small letter u with grave, U+00F9 ISOlat1
  243. 'uml': '\250', # diaeresis = spacing diaeresis, U+00A8 ISOdia
  244. 'upsih': 'ϒ', # greek upsilon with hook symbol, U+03D2 NEW
  245. 'upsilon': 'υ', # greek small letter upsilon, U+03C5 ISOgrk3
  246. 'uuml': '\374', # latin small letter u with diaeresis, U+00FC ISOlat1
  247. 'weierp': '℘', # script capital P = power set = Weierstrass p, U+2118 ISOamso
  248. 'xi': 'ξ', # greek small letter xi, U+03BE ISOgrk3
  249. 'yacute': '\375', # latin small letter y with acute, U+00FD ISOlat1
  250. 'yen': '\245', # yen sign = yuan sign, U+00A5 ISOnum
  251. 'yuml': '\377', # latin small letter y with diaeresis, U+00FF ISOlat1
  252. 'zeta': 'ζ', # greek small letter zeta, U+03B6 ISOgrk3
  253. 'zwj': '‍', # zero width joiner, U+200D NEW RFC 2070
  254. 'zwnj': '‌', # zero width non-joiner, U+200C NEW RFC 2070
  255. }