qof_book_merge first release

[Derek Atkins]

Hi,

Sorry for taking so long to get to this patch.

I was just looking at it to commit it to CVS but I'm a little
concerned about making GncAddress a first-class QOF type.  In
particular, I'm a bit concerned about how it will try to get saved in
the XML file (or SQL, for that matter).  Have you actually tried
saving a GnuCash data file with actual Customer information with this
patch?  I would think that it would save each GncAddress object twice,
once in the parent entity and once when it iterates over the QOF
types.

Also, I don't think you can use qof_instance_version_cmp() on the
GncAddress as the GncAddress is NOT a QofInstance.  That will surely
cause a crash (or at least strange behavior).

Currently the address "object" is really just a "convenience object"
that lives under a Customer, Vendor, or Employee, but has no
standalone meaning outside of those objects.  An Address has no GUID
of its own, it was an abstraction so I didn't need the same code in
four places.  Granted, this was written well before QOF was started,
so perhaps we might need to reexplore that design decision.

I guess my main concern is the definition of the QofObject around an
address.  Are you sure you need it?  If so, why?

-derek

Neil Williams <linux at codehelp.co.uk> writes:

> qof_book_merge - release 1.
>
> Includes
> qof_book_merge.c
> qof_book_merge.h
> test-book-merge.c test routine
> New Account Hierarchy druid 
> Sundry adjustments to QOF support.
> Tweaks to several Makefile.am files to support new files.
> Tweaks to window-main.c to support new menu item
> Changes to druid-hierarchy.c to support the merge druid.
>
> Replaces the patch of 27th August.
> 18 files to be patched
> 6 files to be created. 285Kb in total.

-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord at MIT.EDU                        PGP key available