#6 `subbash/prompt`: GBranch: simplify the command

I was looking for a command to get the current branch name and I found this (src):

git rev-parse --abbrev-ref HEAD

Would it not be better than the current method?

# Current method
git status --porcelain=2 -b 2>/dev/null | tr '\n' ':' | \
awk 'match($0,/# branch.head [^ :]+/) {print substr($0,RSTART+14,RLENGTH-14)}'

If anything, it should faster (I’ve removed the branch name from the output):

$ time git status --porcelain=2 -b 2>/dev/null | tr '\n' ':' | \
awk 'match($0,/# branch.head [^ :]+/) {print substr($0,RSTART+14,RLENGTH-14)}'

real    0m0.008s
user    0m0.007s
sys 0m0.007s

$ time git rev-parse --abbrev-ref HEAD

real    0m0.003s
user    0m0.003s
sys 0m0.000s

Currently the awk is using a variable.
So a better test would be:

~ -> cd projects/dotfiles/
dotfiles + [M] -> GStatus="$(git status --porcelain=2 -b 2>/dev/null | tr '\n' ':')"
dotfiles [M] -> time echo ${GStatus} | awk 'match($0,/# branch.head [^ :]+/) {print substr($0,RSTART+14,RLENGTH-14)}'
master

real    0m0.007s
user    0m0.007s
sys     0m0.001s
dotfiles [M] -> time git rev-parse --abbrev-ref HEAD
master

real    0m0.005s
user    0m0.000s
sys     0m0.005s
dotfiles [M] ->

I lean a bit more towards keeping the code base uniform as the rest of the git checking also uses the git status (and it seems like minimal difference in time to me.)

No problem. I just gave a suggestion which can be rejected—as always. :)