bash get first word last line

Question

I got a bash script and I started to write a changelog to keep track of changes.

I added in the script a function that echoes "Script version --> " But I have to manually edit it.

So I would need a function to get the first word until space of the last line and store it as $CURRENTVERSION

Changelog syntax is:

v-0.3.2 Added x, y.

v-0.3.2.1 Fixed y.

v-0.4 Added z.

Briefly I would need to store v-0.4 as $CURRENTVERSION

score 6 · Accepted Answer · answered Nov 12 '13 at 17:37

6

Using awk:

awk '{w=$1} END{print w}' file

OR using sed:

sed -n '$s/^\([^ ]*\).*$/\1/p' file

answered Nov 12 '13 at 17:37

anubhava

LOL sed is way too complex, is there a wiki somewere? (I don't like too much to use what i don't know/understand pretty good) Tanks to the awk example you told me i would use this: currentver=(`awk 'END{print}' changelog.txt`) Cause i need to keep $1 empty, and then currentver[0] would be what i need; Is this a good way? – Wyatt_LandWalker Nov 12 '13 at 17:45
Yes awk is far more simple here. – anubhava Nov 12 '13 at 17:45
wouldn't `-r` with `sed` be better? it's won't require escaping brackets. – jkshah Nov 12 '13 at 17:49
Also `awk` will process entire file, isn't it? what if change_log is too long! just hypothetical query. I would have preferred `sed` or `tail` with `grep` – jkshah Nov 12 '13 at 17:50
1

I think you would need to measure performance to determine if the entire file was being processed to reach the end. In some sense, all line-oriented programs need to read from the beginning of the file to reach the end; binary search coupled with `seek()` could be used to start later, if you can determine that doing so would not miss any earlier lines required by the program. – chepner Nov 12 '13 at 19:12

score 3 · Answer 2 · answered Nov 12 '13 at 17:48

3

If you don't like awk or sed, you can do this:

tail -n1 file | cut -d' ' -f1

But really, awk and sed are better tools for the job...

answered Nov 12 '13 at 17:48

Jonathan Hall

YEAH i tink i would use `awk 'END{print w}' changelog.txt | cut -d' ' -f1` – Wyatt_LandWalker Nov 12 '13 at 17:51
The only time I wouldn't use awk or sed for this, would be if I was on an embedded system where they weren't installed, for instance. – Jonathan Hall Nov 12 '13 at 17:55

score 2 · Answer 3 · answered Nov 12 '13 at 18:00

2

bash:

read CURRENTVERSION rest < <(tail -1 changelog)

answered Nov 12 '13 at 18:00

glenn jackman

score 0 · Answer 4 · answered Aug 05 '21 at 10:08

0

CURRENTVERSION=$( perl -anle 'print $F[0] if eof' changelog )
echo $CURRENTVERSION 
v-0.4

answered Aug 05 '21 at 10:08

Kjetil S.

4 Answers4