I need a script to recursivly search and edit files from a directory start point with the second line being blank and have that line removed. All other lines that are blank in the file must be left intact.
so something like

if line 2 = null then
remove line 2

Does anyone know how this would be accomplished?

Assuming the blank lines contain no spaces, the file names contain no spaces, and all the files are plain text:-
will create a new file with the same name as the old file, but with "_new" on the end and all the blank lines removed, from all the files in the current directory.

You should be able to build on this to accomplish what you want.

Here's an example I was putzing with

Whoops! Missed the second line thing...

A variation on homey's script:-

Yah I had a similar one to your first one but that gets rid of all instances of blank lines in the whole file instead of just the first instance if its on the second line.

Thats great thanks, never used sed before.

For some dumb reason, the general problem on a sed command to delete a blank second line took me a while.

jschiwal@linux:~/sedwork> cat -n test1
1 line 1
2
3 line 3
4 line 4
5
6 line 5
7 line 6
8
jschiwal@linux:~/sedwork> cat -n test2
1 line 1
2 Line 2
3 line 3
4 line 4
5
6 line 5
7 line 6
8
jschiwal@linux:~/sedwork> sed '/^$/{2d}' test1
line 1
line 3
line 4

line 5
line 6

jschiwal@linux:~/sedwork> sed '/^$/{2d}' test2
line 1
Line 2
line 3
line 4

line 5
line 6

The test for line 2 is nested inside the blank line test.
So this one liner will do what you want:
for file in *; do sed -i '/^$/{2d}' $file; done


Of course, you might want to use double quotes around "$file" if a file-name could contain white space characters.

The problem with the first 2 suggestions is they seemed to be removing the 2nd line even if it wasn't blank. maybe I was missing something.
Now onto the last example that works good accept it won't span directiores so.. I put this into a a file fix.sh
sed -i '/^$/{2d}' $1

then I run the following

find -name "*" -exec /home/blah/fix.sh {} \;

That way it will recurse directories.

Strange on my test machine running redhat I can run sed with the -i flag but on my production machine -i is an invalid option? what is the -i?
It said invalid option --i
? isn't -i the flag to tell it to overwrite the file and -ie would be tomake and archive of the file with an e at the end?

Also if a line looks like a blank line and it doesn't have a space in it what is it? The test files I was using are not the same as the real files I have to edit. When I run the script on a tes file with a blank line it works but when I take the real file sed doesn't seem to recognize the line as a blank line. but when I tell sed to just delete the line it can delete the line? very strange.

The -i option is to edit the file in place, and the -e option is just telling sed to execute the following commands. You don't realkly need it if you are just giving sed one command, but you can string commands together with it.
It could be that the lines are not actually blank, but have a space or tab in them.

Try:-
to check for spaces as well.

EDIT: I should point out that if the "e" follows the "i", as in "-ie", then sed will use the "e" as a suffix to back up the file.

p.s. You seem to have a pretty good grasp of sed now, how about looking at "man sed" to see all the options?