[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: iconv bug in 2.2.3
From: |
Dana Lacoste |
Subject: |
Re: iconv bug in 2.2.3 |
Date: |
Tue, 6 Nov 2001 08:31:24 -0800 |
I see in the digest that it stripped the 8 bit characters.
oh well. I've updated the methodology to show what I did.
Note that the 'bad' characters showed up in both Linux and
Win2k (VMWare under Linux :) so the bug is real, it's just
not readily apparent below :)
--
Methodology :
1 - Create a test file :
cd /tmp
vi test1
^v169 - copyright symbol
^v174 - registered trademark
2 - run the iconv program to convert the symbols :
iconv -f ibm850 -t unicodelittle < test1 > output1
3 - look at the results
cat output1
(insert registered trademark symbol here) - copyright symbol
(insert two left angle brackets symbol here) - registered trademark
3a- Insert Shocked Response
WHAT?
4 - try again with a different/compatible code page to see
what happens:
iconv -f cp1250 -t unicodelittle < test1 > output2
(IF I remember correctly [and web searches seem to back me up]
CP1250 and IBM850 are the same)
5 - look at the results (they're correct this time)
cat output2
- copyright symbol
- registered trademark
6 - Draw conclusions
It looks like there's something wrong with IBM850.