Shell Programming and Scripting

I have a number of strings that I want to remove:

Code:

<b>Task: 100</b>
<b>Task: 1100 </b>
<b>Task: 2200 </b>

But the numbers in them can vary from 4, 5 8, digits in length. But, no alpha chars.
I tried this:

Code:

sed '/Task:/,//d' $file > tmpfile ; mv tmpfile $file

But it removed additional text on the line that I need.

How do I use a wildcard in sed?

Code:

sed 's/<b>Task:[*]</b>//' > tmpfile ; mv tmpfile $file

One way:

Code:

sed '/<b>Task:/s/ [0-9]*//' file

Thanks

Hi Everyone,

I am trying to do this:

sed -i 's/exclude=kernel* iptables selinux* ganglia* torque* *el5_5*/exclude=kernel* iptables selinux* ganglia* torque*/g' /etc/yum.conf

But I know the * in the command is messing me up. Basically I am trying to remove the *el5_5* from the line in /etc/yum.conf

I also tried this:

sed -i 's/exclude=*/exclude=kernel\* iptables selinux\* \*el5_5\* ganglia\* torque\*/g' /etc/yum.conf

In this command I think it is the second half of the replace statement that is messed up. Additionally I thought the /* would negate the meaning and match it as a regular character not a special character.

Thanks
John

Code:

$ ruby -ne 'print $_.gsub(/<b>.*?<\/b>/,"")' file

* is a metacharacter, use \ to escape it.