Public SourceMage Discussion List

[Jason Flatt <jason AT flattfamily.com>]

On Sunday 22 August 2004 9:31 am, Andrew wrote:
>
> There is a config_query_options which you can use for that. However you
> can mess with them as much as you want after persistent_add'ing them
> if I understand things right.
>
> Variables marked persistent are saved and then cleared between spells
> files (and thus spells).

I'm not sure at what point in this thread I should be sending this, but I've 
been modifying a spell and using the config_query_option as specified on the 
Wiki (config_query_option OPTS "Use X ?" y "--with-x" "--without-x"), but I 
have been running into some difficulty.

I'm not sure if I'm understanding its usage wrong, or if I'm just plain using 
it wrong, or if there's a bug in it.  It seems that the variable I'm using 
isn't being cleared when using the -r switch of cast, and that is making it 
difficult for my to use it.

Here is what I have tried in the CONFIGURE file:
config_query_option CONF_OPTS "Compile in support for zlib?" y "HAVE_ZLIB" 
"NO_ZLIB"  &&
config_query_option CONF_OPTS "Compile in support for gettext?" y 
"HAVE_GETTEXT" "NO_GETTEXT"  &&
config_query_option CONF_OPTS "Compile in support for IP v6?" y "HAVE_IPV6" 
"NO_IPV6"

The first pass through, with the defaults results in this:
CONF_OPTS='HAVE_ZLIB HAVE_GETTEXT HAVE_IPV6'

The second time through, without a -r, results in the same settings without 
being re-asked for the values (as it should).

The third time through, with a -r and choosing different options, results in 
the following:
CONF_OPTS='HAVE_GETTEXT HAVE_IPV6 HAVE_ZLIB NO_GETTEXT NO_IPV6'

I went through a fourth time like the second (no -r), and got the same 
results 
as the third time through.  That aspect seems to be operating correctly.


Believing that CONF_OPTS was somehow defective, I chose to try a different 
track and use unique, custom variables for each option, like this:
config_query_option ZLIB_OPT "Compile in support for zlib?" y "HAVE_ZLIB" 
"NO_ZLIB"  &&
config_query_option GETTEXT_OPT "Compile in support for gettext?" y 
"HAVE_GETTEXT" "NO_GETTEXT"  &&
config_query_option IPV6_OPT "Compile in support for IP v6?" y "HAVE_IPV6" 
"NO_IPV6"

But I got the same behavior except that the variables were split out, i.e.:
ZLIB_OPT='HAVE_ZLIB'
GETTEXT_OPT='HAVE_GETTEXT'
IPV6_OPT='HAVE_IPV6'

and:
ZLIB_OPT='HAVE_ZLIB'
GETTEXT_OPT='HAVE_GETTEXT NO_GETTEXT'
IPV6_OPT='HAVE_IPV6 NO_IPV6'


Also, as I was having this odd behavior, I added the NO_ZLIB to the last 
variable for the testing and verifying process.  Ideally, I would like to 
leave that last option blank, like this:
config_query_option CONF_OPTS "Compile in support for zlib?" y "HAVE_ZLIB" "" 
 
&&
config_query_option CONF_OPTS "Compile in support for gettext?" y 
"HAVE_GETTEXT" ""  &&
config_query_option CONF_OPTS "Compile in support for IP v6?" y "HAVE_IPV6" ""

I have no idea yet if that's possible or not, but it is desired.  :^)


I have tried it both using persistent_add and not using it, with no 
difference 
in results.  It was my understanding that using a -r would clear the 
persistenceness of the variables, but that does not seem to be the case.

So, which is it?  Am I misunderstanding the usage, just plain using it wrong 
or is there a bug?

On that machine, I have Test Sorcery 1.11.0-rc1 and the latest Test grimoire 
as of about 11:00 AM EST (GMT -0400, IIRC).


-- 
Jason Flatt (jason @ flattfamily . com)
Father of five (http://www.flattfamily.com/)
Linux user (http://www.sourcemage.org/)
IRC Nick: Oadae  Channels: #sourcemage, #lvlug  Server: irc.freenode.net

-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCS/IT d(-) s+:- a>++++ C++$(+++) UL++++ P(+) L+++ E- W++ N+@ o? K? w-- O? 
M-- V PS- PE+ Y+ PGP@ t 5+ X R tv--() b+@ DI+@ D+++ G e h---- r+++ y++++
------END GEEK CODE BLOCK------