This extension allows an export XML of a class in particular. (Tested on
EzPublish 3.6.1) It provides a module of administration which will
give the possibility of defining syndications XML.
You can specify several things:
* The name of flow.
* The description of flow.
* The site which refers to flow.
* An image which refers to flow.
* The possibility of activate/désactivate the flow.
* The possibility of indicating the ID of the node relative of each
object.
* The possibility of choosing the end of the URL of access of flow.

* The ID of the node source which will be used as root for export.

* The class of type of exported object.
* The fields which one wants to export by redefining the tags which
corresponds to them.
Then flow is accesssible with the URL of access.
Installation
Decompress the extension in the directory extension. Once that made it is
necessary to activate the extension in the file site.ini siteaccess
FACE and BACK office. It is necessary to add a table in the data
base:
/************************************************************/
CREATE TABLE `FLUX_xml` (
`flux_identifiant` varchar(255) NOT NULL default '',
`flux_name` varchar(255) default NULL,
`flux_description` text,
`flux_url_site` varchar(255) default NULL,
`flux_is_actif` tinyint(3) default NULL,
`flux_for_arbo` tinyint(3) default NULL,
`flux_source_node` int(11) default NULL,
`flux_arbo_fille` tinyint(3) default NULL,
`flux_class_identifier` varchar(255) default NULL,
`flux_champs` text,
`flux_image` varchar(10),
`flux_date_lastmodif` datetime default NULL,
PRIMARY KEY (`flux_identifiant`)
);
/************************************************************/
It is also necessary to add a rewriting rule in the virtualhost
corresponding to the siteaccess:
RewriteRule fluxml/syndicate~(.*)$ fluxml/syndicate?id=$1 [PT]
There remains a thing to be made, to list the classes which you allow to
export, this this fact in the file flux.ini of the repertory setting
of the extension.
[ClassesExportables]
identifier[]=folder
identifier[]=article
Possible improvements
I think that experts of EzPublish will improve the input which makes
it possible to inform the id node source.
Ajax would be cool with the selection of the classes
But especially a system of cache could really optimize the syndications.