Newbie question: Any way to improve this code?

[Bengt Richter]

On Sat, 6 Dec 2003 00:06:16 -0000, "Duncan Smith" <buzzard at urubu.freeserve.co.uk> wrote:
[...]
>
>Or you could replace the whole thing with,
>
>try:
>    f = open("firmas.txt",'r')
>    texto = f.read()
>    frases = texto.split('\n')[:-1]

If you split on something, you don't have to eliminate it,
but you may split off a trailing null string, to satisfy the
logic that you should be able to join the list with the splitter and
get back your original. Splitlines does what you want:

 >>> 'abc\ndef\n'.split('\n')
 ['abc', 'def', '']
 >>> 'abc\ndef\n'.splitlines()
 ['abc', 'def']

 >>> 'abc\ndef\n'.split('\n')
 ['abc', 'def', '']
 >>> '\n'.join('abc\ndef\n'.split('\n'))
 'abc\ndef\n'

But the result of splitlines gives you the same last line whether it ends with \n or not,
so you can't guarantee reconstruction:

 >>> '\n'.join('abc\ndef\n'.splitlines())
 'abc\ndef'

More:
 >>> 'abc\ndef\n'.splitlines()
 ['abc', 'def']
 >>> 'abc\ndef'.splitlines()
 ['abc', 'def']
 >>> 'abc\ndef\n\n'.splitlines()
 ['abc', 'def', '']
 >>> '\n'.splitlines()
 ['']
 >>> ''.splitlines()
 []
 >>> 'a'.splitlines()
 ['a']

>finally:
>    f.close()
>
I think you forgot (oh, just realized the explanation for my amazement--you're not the other Duncan ;-)
(maybe you didn't forget) that if the open doesn't succeed in the above,
f will not be (re)bound, so f.close() will be problematical either way.

I guess you could fix it (untested) by

 try:
     f = file("firmas.txt")
     try:
         frases = f.read().splitlines()
     finally:
         f.close()
 except Exception, e:
     print '%s: %s' %( e.__class__.__name__, e)
 
Regards,
Bengt Richter