Peruser System Configuration
On this page, we discuss configuration of the Peruser application container
(which usually runs inside cocoon, which runs inside a servlet container like
tomcat).
The subject of application configuration is more
fun and more interesting. Everything described on this page exists merely to
set up the application context so we can party there without having to write
java code.
The best way to start learning about Peruser system configuration is by
looking at the conf directory tree in your peruser
installation.
Note that a single peruser installation tree may contain configuration options
for multiple deployments.
There are five subdirectories of peruser/conf.
|
ant_prop
|
Contains ant properties controlling build and installation. See
Ant Properties Reference.
|
|
sh_env
|
Sets up operating system environment variables: JAVA_HOME, ANT_HOME and PATH.
|
|
cocoon
|
Contains configuration information used to embed peruser inside a cocoon
deployment. Changes to any of this information (except the root sitemap)
require a servlet container restart to take effect.
|
pre_build
|
Contains configuration used to patch your cocoon installation before it is
built. Changes here require that you rebuild and redeploy cocoon, as well as
restart your servlet container, hence these changes are usually kept to a
minimum.
|
properties
|
Properties used primarily to enable/disable various cocoon features.
|
local.blocks.properties
|
This file enables/disables cocoon
blocks.
Our default config disables all blocks that are not used by the peruser
samples. You can enable more blocks to access more cocoon features.
|
|
local.build.properties
|
High level build switches for cocoon. Peruser disables the cocoon
documentation and test-suite.
|
|
|
webxml_patch
|
These files use the cocoon "xweb" format to patch the
web.xml file that configures your servlet container (e.g.
tomcat) to handle the cocoon servlet. Global webapp behavior (for all your
nested peruser/cocoon apps) is controlled here. Ch
|
peruser_log4j_enable.xweb
|
Configures the cocoon servlet to use log4j, as configured in
cocoon/postbuild/log4j/peruser_log4j.xconf - described below.
|
|
peruser_session_timeout.xweb
|
Sets the HTTP session timeout for the cocoon servlet. Our default is 90
minutes. One HTTP session per browser is shared by all your peruser/cocoon
applications.
|
|
peruser_upload_config.xweb
|
Sets the maximum file upload size for the cocoon servlet. Our default is 5
Megabytes.
|
|
|
xconf_patch
|
These files use the cocoon "xconf" format to patch the
cocoon.xconf file which is the hub of low-level cocoon
framework config. Cocoon will apply these .xconf files to produce a patched
version of the cocoon.xconf file. This file is then deployed into
yourDeployedCocoonWebappDirectory/WEB-INF/cocoon.xconf. This file is
read during the cocoon webapp startup. So, to successfully change any of this
configuration
peruser_kernel.xconf
|
Tells cocoon that we want it to load a single instance of
PeruserCocoonKernel,
to be shared by all peruser applications. Configures that peruser-cocoon-kernel
to have an "indoor_kernel" of type
JenaKernel
, thereby giving your cocoon applications access to all the RDF management
goodies in jena and ARQ (arguably, the main benefit of deploying Peruser today).
This JenaKernel is configured with an RDF model, which by default is the file
cocoon_jena_boot.ttl, which is found in the root directory of
the application configuration tree below.
|
|
peruser_jdbc.xconf
|
Configures the
JDBC
datasources available to your peruser applications. By default, peruser
attempts to connect to a MySQL server on the same machine ("localhost") where
your servlet container (+ cocoon + peruser) is running.
|
|
peruser_input_module.xconf
|
Sets up the
PeruserInputModule
as an
Input
Module (see "step 2 - use it", also see these
javadocs,
and the "input modules examples" in your cocoon installation or
here)
for use in your Cocoon sitemaps. This allows you to use the results of Peruser
operations (e.g. SPARQL query results) in sitemap expressions (e.g. URLs,
arguments to transformers, etc.). Not essential, but handy.
|
|
peruser_cron_jobs.xconf
|
Establishes
PeruserCocoonCronJob
as a
schedulable
job that can be run like a unix "cron" job, according to the schedule of
triggers in peruser_cron_triggers.xconf below.
|
|
peruser_cron_triggers.xconf
|
Defines
scheduled
triggers to be sent to the PeruserCocoonCronJob defined by
peruser_cron_jobs.xconf above.
|
|
|
|
post_build
|
Contains configuration used to patch your cocoon installation after it is
built. Making a change here does not require a cocoon rebuild or redeploy, you
can simply run the peruser ant-task "deploy-into-cocoon", and restart your
servlet container.
|
log4j
|
Configuration of the
log4j logging
facilities used by peruser and cocoon. Contains a single file,
peruser_log4j.xconf, which uses the
Log4J XML Config
Format.
|
|
|
sitemap
|
The biggest advantage of the cocoon framework is the power and flexibility of
sitemaps. Each application you create will have one or more of its own
sitemaps. In this directory, a very simple root sitemap.xmap is setup for the
peruser container. This sitemap simply
- Establishes the
PeruserTransformer
as an available sitemap component.
- Delegates all inbound requests to be handled by the appropriate application
sitemaps.
|
|
|
standalone
|
Contains configuration information used to run the peruser as a standalone
application (e.g. during unit testing).
Currently, this consists only of a single file:
log4j/log4j.properties, which configures
log4j logging for
the peruser unit tests.
|
