Mercurial Repositories > eric / file comparison
comparison: eric6/ThirdParty/Pygments/pygments/util.py
eric6/ThirdParty/Pygments/pygments/util.py
- changeset 7701
- 25f42e208e08
- parent 7547
- 21b0534faebc
- child 7983
- 54c5cfbb1e29
equal
deleted
inserted
replaced
| 7700:a3cf077a8db3 | 7701:25f42e208e08 |
|---|---|
| 3 pygments.util | 3 pygments.util |
| 4 ~~~~~~~~~~~~~ | 4 ~~~~~~~~~~~~~ |
| 5 | 5 |
| 6 Utility functions. | 6 Utility functions. |
| 7 | 7 |
| 8 :copyright: Copyright 2006-2019 by the Pygments team, see AUTHORS. | 8 :copyright: Copyright 2006-2020 by the Pygments team, see AUTHORS. |
| 9 :license: BSD, see LICENSE for details. | 9 :license: BSD, see LICENSE for details. |
| 10 """ | 10 """ |
| 11 | 11 |
| 12 import re | 12 import re |
| 13 import sys | 13 import sys |
| 203 rv = tag_re.search(text[:1000]) is not None | 203 rv = tag_re.search(text[:1000]) is not None |
| 204 _looks_like_xml_cache[key] = rv | 204 _looks_like_xml_cache[key] = rv |
| 205 return rv | 205 return rv |
| 206 | 206 |
| 207 | 207 |
| 208 # Python narrow build compatibility | 208 def surrogatepair(c): |
| 209 | 209 """Given a unicode character code with length greater than 16 bits, |
| 210 def _surrogatepair(c): | 210 return the two 16 bit surrogate pair. |
| 211 # Given a unicode character code | 211 """ |
| 212 # with length greater than 16 bits, | |
| 213 # return the two 16 bit surrogate pair. | |
| 214 # From example D28 of: | 212 # From example D28 of: |
| 215 # http://www.unicode.org/book/ch03.pdf | 213 # http://www.unicode.org/book/ch03.pdf |
| 216 return (0xd7c0 + (c >> 10), (0xdc00 + (c & 0x3ff))) | 214 return (0xd7c0 + (c >> 10), (0xdc00 + (c & 0x3ff))) |
| 217 | |
| 218 | |
| 219 def unirange(a, b): | |
| 220 """Returns a regular expression string to match the given non-BMP range.""" | |
| 221 if b < a: | |
| 222 raise ValueError("Bad character range") | |
| 223 if a < 0x10000 or b < 0x10000: | |
| 224 raise ValueError("unirange is only defined for non-BMP ranges") | |
| 225 | |
| 226 if sys.maxunicode > 0xffff: | |
| 227 # wide build | |
| 228 return u'[%s-%s]' % (chr(a), chr(b)) | |
| 229 else: | |
| 230 # narrow build stores surrogates, and the 're' module handles them | |
| 231 # (incorrectly) as characters. Since there is still ordering among | |
| 232 # these characters, expand the range to one that it understands. Some | |
| 233 # background in http://bugs.python.org/issue3665 and | |
| 234 # http://bugs.python.org/issue12749 | |
| 235 # | |
| 236 # Additionally, the lower constants are using chr rather than | |
| 237 # literals because jython [which uses the wide path] can't load this | |
| 238 # file if they are literals. | |
| 239 ah, al = _surrogatepair(a) | |
| 240 bh, bl = _surrogatepair(b) | |
| 241 if ah == bh: | |
| 242 return u'(?:%s[%s-%s])' % (chr(ah), chr(al), chr(bl)) | |
| 243 else: | |
| 244 buf = [] | |
| 245 buf.append(u'%s[%s-%s]' % (chr(ah), chr(al), | |
| 246 ah == bh and chr(bl) or chr(0xdfff))) | |
| 247 if ah - bh > 1: | |
| 248 buf.append(u'[%s-%s][%s-%s]' % | |
| 249 chr(ah+1), chr(bh-1), chr(0xdc00), chr(0xdfff)) | |
| 250 if ah != bh: | |
| 251 buf.append(u'%s[%s-%s]' % | |
| 252 (chr(bh), chr(0xdc00), chr(bl))) | |
| 253 | |
| 254 return u'(?:' + u'|'.join(buf) + u')' | |
| 255 | 215 |
| 256 | 216 |
| 257 def format_lines(var_name, seq, raw=False, indent_level=0): | 217 def format_lines(var_name, seq, raw=False, indent_level=0): |
| 258 """Formats a sequence of strings for output.""" | 218 """Formats a sequence of strings for output.""" |
| 259 lines = [] | 219 lines = [] |
