Le 30/10/2013 17:32, "Jörg Knappen" a écrit :
The data did not only contain latin-1 type mangling for the
non-existent Windows characters, but also sequences with the raw
C1 control characters for all of latin-1. So I had to do them, too.
The data weren't consistent at all, not even in their errors.
--Jörg Knappen
Your question helped me dust off and repair a non working python snippet
I wrote for a similar problem. I was stuck with the mixing of
windows-1252 and latin1 controls (linked with a chinese characters). I
write it below for reference.
The python snippet below does not need sed, defines a function
(unscramble(S)) which works on strings. The extension to files should be
easy.
Frédéric Grosshans
def Step1Filter(S):
for c in S :
#works character/character because of the cp1252/latin1 ambiguity
try :
yield c.encode('cp1252')
except UnicodeEncodeError :
yield c.encode('latin1')
#Useful where cp1252 is undefined (81, 8D, 8F, 90, 9D)
def unscramble(S):
return b''.join(c for c in Step1Filter(S)).decode('utf8')
PS: If anyone is interested in a licence, I consider this simple enough
to be in the public domain an uncopyrightable.
