URL:
Reporter RHNID:
Section: -
Language:
Workaround:

Description: As it stands, now, the code on pp 353 is as follows:

[student@serverX bin]$ cat showargs
#!/bin/bash

for ARG in "$*"; do
echo $ARG
done

[student@serverX bin]$ ./showargs 1 2 3
1 2 3

[student@serverX bin]$ cat showargs
#!/bin/bash

for ARG in "$@"; do
echo $ARG
done

[student@serverX bin]$ ./showargs 1 2 3
1
2
3

My students didn't really get what was going on here. And I don't blame them – this doesn't really prove anything.

The improvement I would suggest: pass some quoted arguments which contain spaces to showargs, e.g.:

[student@serverX bin]$ cat showargs
#!/bin/bash

for ARG in "$*"; do
echo $ARG
done

[student@serverX bin]$ ./showargs "argument 1" 2 "arg 3"
argument 1 2 arg 3

[student@serverX bin]$ cat showargs
#!/bin/bash

for ARG in "$@"; do
echo $ARG
done

[student@serverX bin]$ ./showargs "argument 1" 2 "arg 3"
argument 1
2
arg 3

I also think the two sentences on pp 352 starting with "When $* is used, all of the arguments are seen as as single word" are technically wrong, but honestly less important than a good demonstration.