python re.sub with variable

Question

Input text:

Ell és la víctima que expia els nostres pecats, i no tan sols els nostres, sinó els del món sencer.  

Expected output:

Ell és la víctima que expia els nostres pecats, i no tan sols els nostres, sinó els del món sencer.

Known facts: unichr(233)=é

for now i have

re.sub('&#([^;]*);', r'unichr(int(\1))', inputtext, flags=re.UNICODE)

and of course is not working, don't know how to pass function on \1

Any idea?

Answer 1

Use a lambda function:

re.sub('&#([^;]*);', lambda match: unichr(int(match.group(1))), t, flags=re.UNICODE)

Answer 2

Fortunately for you, re.sub accepts a function as an argument as well. The function will recieve a "MatchObject" -- From there, you can get the matched groups by match.group(1), match.group(2), etc. etc. The return value of the function will be the string to replace the matched group in the input text.

def fn(match):
  return unichr(int(match.group(1)))

re.sub('&#([^;]*);', fn, inputtext, flags=re.UNICODE)

If you really want, you can inline this and use a lambda -- But I think lambda makes it harder to read in this case¹.

---

By the way, depending on your python version, there are better ways to un-escape html (as it will also handle the special escape sequences like '&':

Python2.x

>>> import HTMLParser
>>> s = 'Ell és la víctima que expia els nostres pecats, i no tan sols els nostres, sinó els del món sencer.'
>>> print HTMLParser.HTMLParser().unescape(s)
Ell és la víctima que expia els nostres pecats, i no tan sols els nostres, sinó els del món sencer.

Python3.x

>>> import html
>>> html.unescape(s)

reference

^{1especially if you give fn a more sensible name ;-)}