[Open Office] Announcement : KDE address book support in OpenOffice.org (m680 code line)

[Éric Bischoff]

Hi all,


(I am not subscribed to this list, so answers in private, please)

I have first alpha code for what I hope will become a new subproject of 
KDE/OpenOffice.org integration project (http://kde.openoffice.org), if 
accepted.

The purpose is to allow to access the KDE address book from OpenOffice.org. It 
is seen like any database or the other address books (Mozilla, Evolution) 
through the "data source" abstraction. It is intended for UNIX-like platforms 
(the ones where KDE currently runs).

The KDE address book is shared by the KDE applications. That includes Kontact 
and KMail. The address book can reside in a file in the user's home 
directory, or on a LDAP server, for example. A practical application of the 
KDE address book driver could be to do serial letters in Writer, for example.

An issue (#52449) has been created to describe the task.

A screenshot can be seen at
	http://www.openoffice.org/nonav/issues/showattachment.cgi/28387/it_works.png

The source code can be downloaded at
	http://www.openoffice.org/nonav/issues/showattachment.cgi/28408/skeleton.tgz

The source code is licensed under SISSL and LGPL. My JCA should have been 
received by Sun now.


Details
=======

The current code is an almost functional and stable skeleton. It consists in 
three components :
- the source code, to be placed into
	connectivity/sources/drivers/kab
- a set of patches (currently agains m120) to integrate the address book into
  the OOo build process and the runtime execution.
- a README.txt file for people interested in compiling, testing, and
  helping out

The skeleton currently reads _all_ the fields for _all_ the people in the 
address book and delivers them to OpenOffice.org. It currently does not 
interpret the SQL queries to restrict the address fields or the people to a 
smaller subset. It just acts as if all the queries were
	SELECT * from addresses

Doing real interpretation of the SQL queries is what I intend to do next. The 
SQL parser is already built in, but its results are not used. First steps are 
probably to handle
	SELECT * from addresses WHERE 0=1
clauses and statements like
	SELECT "Name" from addresses
which are both heavily used by OpenOffice.org.

Later steps might include interprating all WHERE clauses, ORDER BY clauses. 
Better localization and quality assurance would be needed too.

Once that basic functionality works, one could imagine to handle DELETE and 
UPDATE statements to manipulate the KDE addresses from within OpenOffice.org.


Any comments and help offers welcome.