diff -y + UTF-8 = irregular columns

[Erik Karlsson]

Hello.

While in an UTF-8 enviroment, comparing one or two UTF-8 files the
column widths seem to be miscalculated. It seems as if each byte of
multi-byte characters are counted as separate characters, giving a
slight skew on lines with such characters.

Steps to reproduce:
 1. diff -y test test2

Where test and test2 are attached files. test is UTF-8 and test2 not,
although the problem appears when both files are in UTF-8 as well.

Regards,
Erik

diffver
Description: Text document

test
Description: Text document

test2
Description: Text document