/apps/lib/html.py

https://bitbucket.org/resplin/byteflow · Python · 15 lines · 10 code · 4 blank · 1 comment · 2 complexity · a96393838ae5ae7a96c40beb206aa7a1 MD5 · raw file 

    

  1. import htmlentitydefs
    2. 

  2. import re
    3. 

  3. 

  4. pattern = re.compile("&(\w+?);")
    5. 

  5. 

  6. def descape_entity(m, defs=htmlentitydefs.name2codepoint):
    7. 

  7.     # callback: translate one entity to its ISO Latin value
    8. 

  8.     try:
    9. 

  9.         return unichr(defs[m.group(1)])
    10. 

  10.     except KeyError:
    11. 

  11.         return m.group(0) # use as is
    12. 

  12. 

  13. def descape(string):
    14. 

  14.     return pattern.sub(descape_entity, string)
    15.