cannot delete certain files with sed
 

Is this a bug??

On certain files I can use this command correctly:
sed 1d file1

However on other files (extracted from logs), I cannot remove the first line... sed 1d somefile seems to be performing nothing. This typical file, when viewed by vi, has a lot of "@" which disappear when i scroll downwards... What are these "@" ??? This file was extracted from a parsed logfile.


I need to delete the first line of this file since an output of a "diff file1 file2" will pre-append some string which i don't need.

thanks for your help, ganninu.

sed 1d filename > newfile
will delete the first line of a text file.

If the file has a binary format (some log files do) then it will not work the way you want.

A binary file is one where the logical record is NOT delimited by a newline.

will give you the number of newlines in the file. However, binary files can have ANY character embedded in the logical record, even newlines -- which are ignored.

so what do you suggest? :)

I do not know much about the file you are working with


What are you trying to do with the file? There are lots of ways to do things.

The idea is this. I have a big growing logfile on a server. I'm grabbing the last 100 lines of this logfile and then I send it to a remote server. This remote server will have already a previous copy of this extracted file. Then I do a "diff" between these two versions. The problem is that 'diff -n file1 file2" returns me a reduntant line at the head of the file and i want to remove it. So, to cut it short, I'm using sed d1 file2 to remove the first line (generated by diff) of file2...

ganninu.

You're using
and the first line of tmpfile always has a bunch of garbage in it?

For this file with garbage in it (using the filenames in the example I gave above) execute these commands (you can use vim or vi)
vi tmpfile

:set nu
G

You should now be at the last line of the file.
You should see a 100 over in the left column - meaning your file has 100 lines. What do you see? Do you see 100?

Plus, what does the garbage first line look like?

erm... There is no garbage... everything is text... The first line i was talking abt was that one generated by DIFF - "d1 26" ... is there some special argument given to 'diff' to get rid of that???


...likewise, is there another fool-proof method to delete some lines out of a file?

thanks, ganninu