an attempt at shell scripting in haskell
Shell lets you do fun stuff like
ls -la | grep foo | wc -l (I know, I know, grep -c, but it's hard to think of good random pipelines on the spot!)
but sometimes makes you do awful stuff like
for x in $(seq 10); do echo $x; done
or
if [ $# -lt 5 ]; then echo "oh no!"; fi
or
x=$((x + 1))
and the list goes on. Even worse, functions, data structures, modules, and other important programming niceties often don't quite match expectations. The typical solution to this problem is to switch to a more "programm-y" scripting language, like Python. But there, the beautiful pipelines turn into
p1 = subprocess.Popen(['ls', '-la'], stdout=subprocess.PIPE)
p2 = subprocess.Popen(['grep', 'foo'], stdin=p1.stdout, stdout=subprocess.PIPE)
p3 = subprocess.Popen(['wc', '-l'], stdin=p2.stdout)which is nice and all, but it's no ls -la | grep foo | wc -l
Enter Haskell. I have a feeling that with Haskell's strengths in domain specific languages, it may be possible to make something that looks a lot more like a shell pipeline, but still lets you quickly escape hatch into "proper" programming.
The best I have done so far is:
runPipeline $ initPipeline `pipe` qProc "ls -la" `pipe` qProc "grep foo" `pipe` qProc "wc"or the best SIGILED-UP attempt so far:
(>%) $ "ls -la" $|| "grep foo" $| "wc"which feels like a pretty decent start.