I have a quirky situation whereby I'm using SED to selectively comment out a line in a crontab job (on Solaris, I know but it's connected to the Linux function I'm working with).

What's happening is this.

Remove hash symbol
Restore hash symbol
The problem I'm running into is that the script can sometimes prepend an extra hash # symbol if run more than once (I have a lockfile that I poll for to discourage this, but that's not perfect).

I wonder how I can modify that sed statement to remove any/all leading hash marks up to the first other character, in this case it's a 0 (zero) which is a crontab file.

Thanks!

Try this replacing your substitute command with this:

No luck, I still ended up with repeat hash symbols on the reverse and they were not corrected with the asterisk.

Thanks, tho.

s/^#0/0/g

or maybe

s/^##/#/g

To turn ## into #

The first will work (I tried it); however, because of the way the script runs, a repeat run will add another #.

The second option will work if you assume that there's no other problem and only 2 hashes exist.

I think a better solution might be (which I'm figuring out) to tell 'sed' to replace 1 or more instances of a # that begins at ^ up to the next alphanumeric character, which in this case is always 0 (zero).

OR

Use 'sed' to simply search for the matching line, delete it and replace with another full line that has the hash removed already. This might be more efficient/reliable.

Suggestions welcomed. This is one of those good learning experiences... :-)