>>> len('a\\nb')
4
>>> len('a\\nb'.decode('string_escape'))
3
>>>
Or for unicode strings
>>> len(u'\N{euro sign}\\nB')
4
>>> len(u'\N{euro sign}\\nB'.encode('utf-8').decode('string_escape').decode('utf-8'))
3
This compares to naive approach to decode character escape by writing
your own scanner in pure Python. For example:
def decode(s):
output = []
iterator = iter(s)
for c in iterator:
if c == '\\':
...enter your state machine and decode...
else:
output.append(c)
return ''.join(output)
or
def decode(s):
return s\
.replace('\\n','\n')\
.replace('\\t','\t')\
...and so on for the few escapes supported...
The navie approaches are expected to be much slower.
