I need to find in a given file (text) all lines containing the word 'bla' AND all lines containing the word 'blue' and so on...
The following command is not good because it gives the lines containing all words:
cat text | grep bla | grep blue

Thanx
8-)

http://www.geocities.com/cohvi

cat <file> | grep -i -E "xls|jpg|bmp|doc|etc"

cat is unnecessary. You can do it this way

grep -E "one|two|three" file.txt

With cat you must read the entire file first and then direct the output to the grep. If you have a small file this is fine, but if you have a very large file, say 1mb or more, like as in a log file, it will have a delay and take time away from your other application processes since 50% of all time slices are reserved for the OS and root, at least on HPUX that is the way it is.

__________________
My brain is your brain

root@catty:[130] >uname -a
SunOS catty 5.8 Generic_108528-13 sun4u sparc SUNW,Ultra-2
root@catty:[131] >grep -E "test" *
grep: illegal option -- E
Usage: grep -hblcnsviw pattern file . . .
root@catty:[132] >

Champ, this is a forum you can add commentary. You might need to download a newer version of grep for your sparc. Most modern versions of grep support -E.

Of course, your version of UNIX must support the -E option. If you don't have it you must use either:

egrep or

grep -e text -e "text" file.out

__________________
My brain is your brain

On SunOS, /usr/xpg4/bin/grep is a modern grep that supports -E.