Error in title to filename function

[GoogleCodeExporter]

I get the following error when trying to create a song with many umlauts:

File "../share/exposong/lib/exposong/glob.py", line 47, in title_to_filename
    elif ret[len(ret)-1].isalnum() and title[i] not in "'\",.?":
IndexError: string index out of range

Patch is appended.

Original issue reported on code.google.com by s.mehrbrodt on 4 Mar 2010 at 11:12

Attachments:

• titletofilename.patch

[GoogleCodeExporter]

One problem that I see right away is that "new not in ..." is comparing the 
whole
string, but it was just a character before. I'm not sure I see how this works 
though.
How does "not new.endswith('_')" work better for foreign characters? Is that how
python sees utf8 extended characters?

I think the problem is here is that the first character is not alphanumeric. So 
just
saying:
elif len(ret) > 0 and ret[len(ret)-1].isalnum()...
would probably do the trick.

Is it easy to translate the ü and similar characters to ascii? The openlyrics 
said
just to use the foreign characters, but I'm not comfortable with it, so I think 
it
would be good to translate the characters if possible.

Original comment by bradleelandis on 5 Mar 2010 at 3:10

• Added labels: Type-Defect, Milestone-Release0.6.1, Priority-High

[GoogleCodeExporter]

You are right with your first statement, that's a typo and it should be "i not 
in ...".

"not new.endswith('_')" is doing the same thing like before, that has nothing 
to do
with umlauts, it looks just if there was a '_' before, and if so, it doesn't 
append
another '_'.
So umlauts are not translated, but replaced with '_' as before and if you've 
got more
umlauts in a row, there will be only one '_'.

You are right, the problem appears if the first letter is non-alphanumerical. 
So your
solution would work also, but it's a bit inconvenient to loop a string with 
"for i in
range(len(str))" if you can do it with "for i in str"

Original comment by s.mehrbrodt on 6 Mar 2010 at 1:41

[GoogleCodeExporter]

I've commited this to svn, I haven't found an easy way to translate umlauts to 
ascii,
but this is ok for me that they are converted to '_'.

Original comment by s.mehrbrodt on 7 Mar 2010 at 4:24

• Changed state: Fixed