_arguments bug

_arguments bug

[Oliver Kiddle]

_arguments doesn't work if the first argument doesn't exclude any other
arguments:

_foo() {
  _arguments -C '-m[m]:m'
}
compdef _foo foo
foo -<tab>

This hangs until I press Ctrl-C. If I use '(-x)-m[m]:m' instead, it will
work.

This bug was not caused by any of this mornings patches as I was having
problems last week: it was only today that I tried to trace a cause.

Oliver

Re: _arguments bug

[Bart Schaefer]

On May 22,  5:07pm, Oliver Kiddle wrote:
} Subject: _arguments bug
}
} _arguments doesn't work if the first argument doesn't exclude any other
} arguments:
} 
} _foo() {
}   _arguments -C '-m[m]:m'
} }
} compdef _foo foo
} foo -<tab>
} 
} This hangs until I press Ctrl-C.

The -m is being interpeted as an argument of _arguments, not as an argument
spec for what should be completed.  It looks like the -m option (formerly
used to initialize `ismulti') went away as of 11489.

} This bug was not caused by any of this mornings patches

Based on _complete_debug output and "cvs diff" it certainly appears to
have been caused by one of today's patches.  Try this.

Index: Completion/Base/_arguments
===================================================================
@@ -155,7 +155,7 @@
 fi
 
 subopts=()
-while [[ "$1" = -(O*|C|m*) ]]; do
+while [[ "$1" = -(O*|C) ]]; do
   case "$1" in
   -C) usecc=yes; shift ;;
   -O) subopts=( "${(@P)2}" ); shift 2 ;;


-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com