Grep and catch the last result

DoME69 · 06-30-2015, 02:07 PM

I have huge files and i want to grep only the last line that grep find.

I don't want to use tail since count line is change also from file to file and I don't want grep will search the all file.

for example file contain:

Quote:

... (1000000 lines)
bla bla bla
la la la
OPS hello
--- (5000000 lines)
No yes no
kuku OPS kuku
opala OPS
... (90000 lines)

grep to find OPS will give the last result
"opala OPS"

I don't want grep will work the all file since it can be Giga or more.
only search from the end of the file until he catch the "OPS" word.

Thanks

rknichols · 06-30-2015, 02:17 PM

You could use tac (think, "cat" spelled backwards) to list the lines in reverse order and then print the first match. Something like

Code:

tac somefile | grep -m 1 OPS

It's still going to be a bit slow. Reading a file backward is not as efficient as reading in the forward direction (unless someone has built "backward readahead" into the kernel while I wasn't looking).

netnix99 · 06-30-2015, 02:19 PM

This may not be exactly what you are looking for, but it does work...

Code:

tac filename | grep -m 1 "OPS"

tac is the opposite of cat, and it starts at the end of the file. grep -m 1 "target" searches for the target and stops after the first match his found.

DoME69 · 06-30-2015, 02:24 PM

Thanks.