Ben Pfaff [Sat, 6 May 2006 21:13:08 +0000 (21:13 +0000)]
Make FILTER without any further keywords turn off filtering, with a
warning, for compatibility. Change errors from cascading failures to
ordinary failures. Check for command terminator.
Ben Pfaff [Sat, 6 May 2006 20:27:23 +0000 (20:27 +0000)]
Continue reforming procedure execution. In this phase, remove PROCESS
IF, which was deprecated anyway and can be easily simulated with
TEMPORARY followed by SELECT IF.
Ben Pfaff [Sat, 6 May 2006 17:46:52 +0000 (17:46 +0000)]
Continue reforming procedure execution. In this phase, get rid of the
output code for SPLIT FILE groups in procedure.c, which really
shouldn't be doing any output. Move it into the individual procedures
instead. This also adds some flexibility.
Ben Pfaff [Sat, 6 May 2006 05:56:59 +0000 (05:56 +0000)]
Continue reforming procedure execution. Change internal_procedure()
so that it calls open_active_file() and close_active_file(), which
isolates most of the actual procedure functionality.
Ben Pfaff [Fri, 5 May 2006 04:53:12 +0000 (04:53 +0000)]
Continue reforming procedure execution. In this phase, move
procedure.c and procedure.h from src to src/data. Update makefiles
and #includes accordingly.
Ben Pfaff [Thu, 4 May 2006 06:19:02 +0000 (06:19 +0000)]
Continue reforming procedure execution. In this phase, get rid of
many global variables, consolidating procedure execution in
procedure.c. Encapsulate transformations in new "struct trns_chain".
Also, change implementation of N OF CASES, FILTER, and PROCESS IF from
special cases to transformations.
Ben Pfaff [Mon, 1 May 2006 22:14:53 +0000 (22:14 +0000)]
Prohibit LAG following TEMPORARY. This both matches SPSS behavior and
fixes a bug: we saved the cases for LAG before TEMPORARY but allowed
access to variables created afterward anyhow (which could cause a
segfault).
Ben Pfaff [Thu, 27 Apr 2006 02:51:12 +0000 (02:51 +0000)]
Continue reforming procedure execution. In this phase, assert that
add_transformation() is not called during procedure execution. Thanks
to John Darrington for the suggestion.
Ben Pfaff [Wed, 26 Apr 2006 20:54:07 +0000 (20:54 +0000)]
Improve the way we handle the various parsing "states". Until now
we've hard-coded the state transitions in the command definition file,
but that's error-prone and, worse, it's redundant--we can figure out
what state we're in anyhow. We can cleanly handle INPUT PROGRAM and
FILE TYPE with a nested command-processing loop.
Ben Pfaff [Tue, 25 Apr 2006 18:03:03 +0000 (18:03 +0000)]
Continue reforming error message support. In this phase, drop actual
message printing from core code, substituting a callback, and add the
callback to each UI. Also, move verbose_msg() into its own module.
Ben Pfaff [Tue, 25 Apr 2006 16:41:33 +0000 (16:41 +0000)]
* main.c: (terminate) Mark static and NO_RETURN. If called
recursively, which can only happen via signal, just exit instead of
trying to gracefully shut down.
Ben Pfaff [Tue, 25 Apr 2006 00:41:12 +0000 (00:41 +0000)]
Continue reforming error message support. In this phase, rename
all the message functions and types to start with "msg", except
for the ones that will be moving to other modules anyway.
Ben Pfaff [Mon, 24 Apr 2006 05:09:59 +0000 (05:09 +0000)]
Continue reforming error message support. In this phase, get rid of
message "titles" and put the message text in `struct error'. Now
`struct error' encapsulates a message more properly.
Ben Pfaff [Mon, 17 Apr 2006 01:37:51 +0000 (01:37 +0000)]
We don't really support anything but Unix-like environments well, so
we might as well de-obfuscate by writing directory and path separators
explicitly.
Ben Pfaff [Sun, 16 Apr 2006 23:35:19 +0000 (23:35 +0000)]
* filename.c: (fn_tilde_expand) Rewrite for cleaner code. Also, now
it only tilde-expands file names, not paths.
(fn_search_path) Tilde-expand one directory at a time.
Ben Pfaff [Sun, 16 Apr 2006 23:08:02 +0000 (23:08 +0000)]
Continue reforming error message support. In this phase, we get rid
of VM() and the other msg() support for "verbosity", replacing it by a
new function verbose_msg().
Ben Pfaff [Sun, 16 Apr 2006 18:55:09 +0000 (18:55 +0000)]
Start reforming error message support. In this phase, we get rid of
"installation errors" and change all uses of msg() in the output
drivers to uses of error() or error_at_line().
Ben Pfaff [Sun, 16 Apr 2006 02:15:06 +0000 (02:15 +0000)]
Complain about missing prerequisites in a group at the end of
configuration, not piecemeal. This should make building PSPP less
frustrating. Thanks to John Darrington for the suggestion.