I would like to turn this string:

a = '\\a'

into this one

b = '\a'

It doesn't seem like there is an obvious way to do this with replace?

To be more precise, I want to change the escaping of the backslash to escaping the character a.

The character '\a' is the ASCII BEL character, chr(7).

To do the conversion in Python 2:

from __future__ import print_function
a = '\\a'
c = a.decode('string-escape')
print(repr(a), repr(c))

output

'\\a' '\x07'

And for future reference, in Python 3:

a = '\\a'
b = bytes(a, encoding='ascii')
c = b.decode('unicode-escape')
print(repr(a), repr(c))

This gives identical output to the above snippet.

In Python 3, if you were working with bytes objects you'd do something like this:

a = b'\\a'
c = bytes(a.decode('unicode-escape'), 'ascii')
print(repr(a), repr(c))

output

b'\\a' b'\x07'

As Antti Haapala mentions, this simple strategy for Python 3 won't work if the source string contains unicode characters too. In tha case, please see his answer for a more robust solution.