This question already has an answer here:
-
how to grep one string occuring multiple times from same file
3 answers
I want to capture a string "1 row affected..". But problem is there are n no of such string present in the same file.
My concern to capture "1 row affected.." only which is present after the string " UPDATE kplustp..Service SET Service_Name = "PositionService", ServiceType = \'Z\', Hostname = " " " in log file. "1 row affected.." will be present after 3 4 sentences from " UPDATE kplustp..Service SET Service_Name = "PositionService", ServiceType = \'Z\', Hostname = " " "
I am working on solaris.
How to do?
Use ggrep with the --byte-offset option to find the byte position of the last string that will occur before the string you want to match. Then feed that into /usr/xpg4/bin/tail with the -c option to produce a truncated version of the file containing only the part after that byte location. You can then do your search.
For instance, to search file.txt to find lines containing "foo", but only after the last of any lines containing "bar":
/usr/xpg4/bin/tail -c +`ggrep --byte-offset bar file.txt | cut -d: -f1 | tail -1` file.txt | grep foo
This is somewhat more difficult on Solaris than on Linux/BSD because the default tail and grep are ancient and lack the necessary options. Knowing how to find the GNU version of grep (ggrep) and that the XPG4 (X/Open CAE Portability Guide 4) versions of some commands have useful additional options is half the trick. Both should be installed on most Solaris 10 systems.
