UNIX for Dummies Questions & Answers

Help regarding effective usage of diff for comparing files

Hi All,

I have few doubts regarding best usage of diff command. I also have some questions with out put of diff command.


File1:
ABC
DEF

File2:
ABC
DEFAA

Code:

diff file1 file2
2c2 
<DEF
-----
>DEFAA

QUESTION:
What does this 2c2 signifies. I also see some times it as 2d5, i have googled regarding this but couldn't found any information regarding this. If some one can please tell me what it signifies, that will be of great help, preferably if some link to study this.

COMING TO MY MAIN PROBLEM:

Problem statement:

I want to dump non common lines when two files are diffed to new files. In case of above example two files should be generated
i.e. file1_new and file2_new that contains only "DEF" and "DEFAA" only (lines which are not common).

I also want to maintain the line number (original) of these files to be dumped in new files.


Thanks in advance,
Sarbjit

Sorry I don't think that the "diff" command is suitable. The output of "diff" gives lines in context and will therefore output some lines which are not different. It also does not number the lines.

The requirement for the line number eliminates every unix utility I can think of.
We'd certainly need to be sure that these are unix format text files and know a lot more about the data structure. Programs to process differences are really only effective on sorted files which do not contain large numbers of identical data.
In this case sorting the data would appear to be out of the question because you want to maintain the order of the data.