UNIX for Advanced & Expert Users

I need compare two files and bdiff cut my file and use diff to compare but Because of the segmenting of the files, bdiff does not necessarily find a smallest sufficient set of file differences.

there are another utiliy?

thanks

Okay. I do not understand. bdiff is for really big files and gives the same output as does diff. So why use diff?

Tell us what you really want to do. Not what you think is how to do it?

its simple. I want to get the difference of two files.
diff have limitations about the size of the file.
bdiff cut my file in small files and use diff to compare. so for example:

two big files with numbers:
File a file b
1 3
2 6
3 7
5 8
6 9
7 1

Big diff cut the files like these

File a1 File b1
1 3
2 6
3 7

File a2 File b2
5 8
6 9
7 1

and use diff to compare a1 with b1 and a2 with b2
and the result is
1
2
6
7
5
8
9



like the man bdiff says : "The output of bdiff is exactly that of diff, with line numbers adjusted to account for the segmenting of the files
(that is, to make it look as if the files had been processed
whole). Note: Because of the segmenting of the files, bdiff
does not necessarily find a smallest sufficient set of file
differences."

http://www.csee.usf.edu/cgi-bin/man-cgi?bdiff+1

i hope that you understand my problem.

I still don't know what your goal is. Unless you are trying to use diff output to automate an ed script, it will not make any difference.

However, use split to cut large files down. Assume your largefile limit starts at 2 GB.
2147483647 bytes try something like this.

Code:

#!/bin/ksh

split -n 2147483646 ./file1  /tmp/file1
split -n 2147483646 ./file2  /tmp/file2
> ./difflog
find /tmp -name 'file1*' | \
while read smallfilename
do
    let pos=${#smallfilename}
    let pos=$pos-2
    last_two_letters=`expr substr $smallfilename $pos 2`
    diff $smallfilename "/tmp/file2""$last_two_letters"
done > ./difflog

This runs diff on the largest file chunk possible.