In more detail, I've been doing an assignment that asks for me to find all files under a directory in all caps, saved in a txt file with no overlapping entries, and said directory looks something like this.
website.com - - [date/date/date:date:date:date -0400] "GET /elv/HELLO/wave.gif HTTP/1.0" 200 1318
website.com - - [date/date/date:date:date:date -0400] "GET /elv/ball.gif HTTP/1.0" 200 306
I tried:
grep "/elv/" (filename.txt) | awk '{print $7}' |
awk -F/ '{print $3}' | grep ^[A-Z] | sort -u > elv.txt
I've come to the understanding what
grep "/elv/" (filename.txt)
awk '{print $7}'
grep ^[A-Z]
sort -u > elv.txt
mean, but i do not understand why the code
awk -F/ '{print $3}'
Is then necessary. Is it used to not register the rest of the $7 past the / of /elv/CAPITALISED_WORD after the CAPITALISED_WORD?
I have tried
grep "/elv/" (filename.txt) | awk '{print $7}' |
awk -F/ '{print $3}' | grep ^[A-Z] | sort -u > elv.txt
for my initial test, and the result comes in a txt file with all names that are capitalised after /elv/.
HELLO
(list of other capitalised words)
But when I tried
grep "/elv/" (filename.txt) | awk '{print $7}' | grep ^[A-Z] |
sort -u > elv.txt
The txt file would appear blank.
I would have thought that by putting in the first attempt I made, it would have given a text file that would look like this.
HELLO/wave.gif
Edited for more clarity
CodePudding user response：
You may use this awk that combines 2 grep and 2 awk commands into one:
awk '/\/elv\// && split($7, a, /\//) && a[3] ~ /^[A-Z]/ {print a[3]}' file |
sort -u
HELLO
To answer your question:
awk -F/ '{print $3}'
Prints 3rd column in a record split by /.
Details about awk command:
/\/elv\//: If an input line matches text/elv/&& split($7, a, /\//): and if 7th field can be split by/&& a[3] ~ /^[A-Z]/: and if 3rd field in split array starts with an uppercase letter{print a[3]}: Print 3rd element ofsplitarray