...
> > I've implemented some configuration possibilities for the LangSelect,
> > but
> > I'm not sure that they can be very useful in the case, when lang for the
> > session differs from the lang in a cookie. Now the action can be used as
> > this:
> >
> > <map:action name="lang-select"
> > src="org.apache.cocoon.acting.LangSelect">
> > <!-- Should the action create a session to store the language
> > code. -->
> > <create-session>true</create-session>
> > <!-- Should the action store the lang code in the session -->
> > <store-in-session>true</store-in-session>
> > <!-- Should the action store the lang code in a cookie -->
> > <store-in-cookie>true</store-in-cookie>
>
> I can see the need for a <store-in-request> as an request attribuite as
well
> (for those not using sessions nor cookies at all).
Ok. I will add that.
> > <!--
> > I am going to add other configuration params:
> > cookie-age - the cookieAge attribute of the cookie with the
> > lang
> > code
> > param-name - the name of the request param, the session
> > attribute and the name of the cookie.
> > Now it's hard coded and equals to 'lang'.
> > -->
> > </map:action>
> >
> > Maybe it's better to make these settings params, so for some pipelines
> > thay
> > can be defined dynamicaly from the sitemap?
>
> I think it's better to have config and parameters as well. Former is meant
as
> default behaviour later as runtime overriding value. I've made that for
the
> TraxTransformer as well. This eliminate the need to define differently
> configured components of the same implementation where you have default
for the
> usual case and are able to change them in the pipeline for special use.
>
> > Btw, why the syntax for configuration and sitemap params are different?
> > Why
> > not to use:
> > <parameter name="create-session" value="true" />?
>
> The configuration is a hierarchical structure (XML fragment) where as the
> Parameters are name/value pairs.
So, I have to implement Configurable interface (as it is now) and at the
same time process parameters passed to the act method, right?
Btw, in the main sitemap there is a match with pattern "i18n", but there is
no corresponding files (navigation.xsp, etc.). Won't it have a conflict with
the separate pipeline for my i18n sample? I think that either it must be
removed or it'd better that the author (is that you Giacomo?) add his i18n
sample to the CVS.
>
> Giacomo
>
Kot.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]
