Reading multiple lines in bash without spawning a new subshell?

Question

I'm trying to do something like

var=0  
grep "foo" bar | while read line; do  
   var=1  
done

Unfortunately this doesn't work since the pipe causes the while to run in a subshell. Is there a better way to do this? I don't need to use "read" if there's another solution.

I've looked at Bash variable scope which is similar, but I couldn't get anything that worked from it.

This is http://mywiki.wooledge.org/BashFAQ/024 – tripleee Jul 19 '16 at 17:42 — tripleee, Jul 19 '16 at 17:42

score 25 · Accepted Answer · edited Jul 19 '16 at 17:42

25

If you really are doing something that simplistic, you don't even need the while read loop. The following would work:

VAR=0
grep "foo" bar && VAR=1
# ...

If you really do need the loop, because other things are happening in the loop, you can redirect from a <( commands ) process substitution:

VAR=0
while read line ; do
    VAR=1
    # do other stuff
done <  <(grep "foo" bar)

edited Jul 19 '16 at 17:42

that other guy

116,971
11
170
194

answered Mar 04 '10 at 00:22

Kaleb Pederson

45,767
19
102
147

2

( list )? Isn't it called process substitution? http://tldp.org/LDP/abs/html/process-sub.html – pihentagy Jun 10 '13 at 07:23
Reference: https://www.gnu.org/software/bash/manual/bash.html#Process-Substitution – Evandro Coan Jul 02 '19 at 05:39

score 3 · Answer 2 · answered Mar 04 '10 at 00:30

3

then don't use pipe ,and lose the grep

var=1
while read line
do  
   case "$line" in
    *foo* ) var=1
   esac   
done <"file"
echo "var after: $var"

answered Mar 04 '10 at 00:30

ghostdog74

327,991
56
259
343

Reading multiple lines in bash without spawning a new subshell?

2 Answers2

Linked

Related