rcexpandparam bug

[Peter Stephenson <pws@ifh.de>]

I've localised the rcexpandparam bug; unfortunately it's in the 
paramsubst() code where I don't know what anything means.

In the long `if (isarr)' chunk, right at the start there is a test
for no array value being set, with plan9 (rcexpandparam) not in
operation.  If that is true, something mysterious happens anyway and
non-null (whatever `n' is) is passed back.  If plan9 is set, there's a
chunk later on which handles that:  lots and lots of mysterious things
happen, until at the end there is another `if plan9' test to see if it
hasn't been unset:  this two line chunk then returns null.  That's
causing the problem:  this gets passed back by stringsubst() to
prefork(), causing the latter to abort all further processing.  The
result is that in things like `=ls $*', the =ls doesn't get expanded
if $* was null and rcexpandparam was set.  I have no idea what to do
about it.

-- 
Peter Stephenson <pws@ifh.de>       Tel: +49 33762 77366
WWW:  http://www.ifh.de/~pws/       Fax: +49 33762 77413
Deutsches Elektronen-Synchrotron --- Institut fuer Hochenergiephysik Zeuthen
DESY-IfH, Platanenallee 6, 15738 Zeuthen, Germany.