You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@cocoon.apache.org by Vadim Gritsenko <va...@reverycodes.com> on 2007/12/10 16:00:48 UTC
Re: Pros and cons of extension less requests (was Re: Url without extensions - how to match/rewrite)
On Dec 10, 2007, at 3:40 AM, Thorsten Scherler wrote:
> I ended up with:
>
> <!-- This will match all extension less requests -->
> <map:match type="regexp" pattern="^[^.]+$">
> <!-- This will match all requests having "/" on the end -->
> <map:match type="regexp" pattern="^(.*)/$">
> <map:match pattern="*/">
> <map:redirect-to uri="../{1}.html" />
This is a browser redirect, so user will ultimately see .../foo.html
URL in the browser. In this case, why would you start with a URL
without extension? I does not make any sense. Just use URL with
extension from the beginning.
> </map:match>
> <map:match pattern="**/*/">
> <map:redirect-to uri="../{2}.html" />
> </map:match>
> </map:match>
> <map:match pattern="**/*">
> <map:redirect-to uri="{2}.html" />
> </map:match>
> <map:match pattern="*">
> <map:redirect-to uri="{1}.html" />
> </map:match>
> </map:match>
Regardless of comment above, what you got here is too verbose, you can
DRY it down using smarter match pattern. For example:
<map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)\/$">
<map:redirect-to uri="../{1}.html"/>
</map:match>
<map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)$">
<map:redirect-to uri="{1}.html"/>
</map:match>
> One side effect is that the app is now way slower if you use
> "extensionless requests". In our application we need the extensions to
> trigger the correct pipelines, which I consider the normal usage of
> cocoon. By redirecting the request one is causing more processing on
> the
> app which explains the higher response time.
You probably mean "more round trips". I don't see any extra processing
on the app.
Vadim
> That makes me wonder whether extension less requests makes sense at
> all
> in a multi-output-format environment.
>
> WDYT?
>
> salu2
> --
> Thorsten Scherler
> thorsten.at.apache.org
> Open Source Java consulting, training and
> solutions
Re: Pros and cons of extension less requests (was Re: Url without extensions - how to match/rewrite)
Posted by Vadim Gritsenko <va...@reverycodes.com>.
On Dec 10, 2007, at 11:34 AM, Thorsten Scherler wrote:
> On Mon, 2007-12-10 at 10:00 -0500, Vadim Gritsenko wrote:
>> On Dec 10, 2007, at 3:40 AM, Thorsten Scherler wrote:
>>
>>> I ended up with:
>>>
>>> <!-- This will match all extension less requests -->
>>> <map:match type="regexp" pattern="^[^.]+$">
>>> <!-- This will match all requests having "/" on the end -->
>>> <map:match type="regexp" pattern="^(.*)/$">
>>> <map:match pattern="*/">
>>> <map:redirect-to uri="../{1}.html" />
>>
>> This is a browser redirect, so user will ultimately see .../foo.html
>> URL in the browser. In this case, why would you start with a URL
>> without extension? I does not make any sense. Just use URL with
>> extension from the beginning.
>
> lol, tell it my client.
>
> The problem is when you do not do the rewrite
Solution's simple - since you really don't need them, just do not use
such URLs at all :)
> you will end up to use
> absolute urls for all images, css, etc. since the client browser will
> interpret the "/" on the end as directory screwing up the relative
> imports. Meaning you have either lots of test whether or not we have
> "/"
> on the end or do a rewrite.
>
>>
>>> </map:match>
>>> <map:match pattern="**/*/">
>>> <map:redirect-to uri="../{2}.html" />
>>> </map:match>
>>> </map:match>
>>> <map:match pattern="**/*">
>>> <map:redirect-to uri="{2}.html" />
>>> </map:match>
>>> <map:match pattern="*">
>>> <map:redirect-to uri="{1}.html" />
>>> </map:match>
>>> </map:match>
>>
>> Regardless of comment above, what you got here is too verbose, you
>> can
>> DRY it down using smarter match pattern. For example:
>>
>> <map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)\/$">
>> <map:redirect-to uri="../{1}.html"/>
>> </map:match>
>> <map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)$">
>> <map:redirect-to uri="{1}.html"/>
>> </map:match>
>>
>
> Hmm, sorry but the above is not working for me.
I tested with the applet, was working fine. Probably you need to
double up each '\' to work around sitemap escaping syntax.
Vadim
Re: Pros and cons of extension less requests (was Re: Url without
extensions - how to match/rewrite)
Posted by Thorsten Scherler <th...@juntadeandalucia.es>.
On Mon, 2007-12-10 at 10:00 -0500, Vadim Gritsenko wrote:
> On Dec 10, 2007, at 3:40 AM, Thorsten Scherler wrote:
>
> > I ended up with:
> >
> > <!-- This will match all extension less requests -->
> > <map:match type="regexp" pattern="^[^.]+$">
> > <!-- This will match all requests having "/" on the end -->
> > <map:match type="regexp" pattern="^(.*)/$">
> > <map:match pattern="*/">
> > <map:redirect-to uri="../{1}.html" />
>
> This is a browser redirect, so user will ultimately see .../foo.html
> URL in the browser. In this case, why would you start with a URL
> without extension? I does not make any sense. Just use URL with
> extension from the beginning.
lol, tell it my client.
The problem is when you do not do the rewrite you will end up to use
absolute urls for all images, css, etc. since the client browser will
interpret the "/" on the end as directory screwing up the relative
imports. Meaning you have either lots of test whether or not we have "/"
on the end or do a rewrite.
>
>
> > </map:match>
> > <map:match pattern="**/*/">
> > <map:redirect-to uri="../{2}.html" />
> > </map:match>
> > </map:match>
> > <map:match pattern="**/*">
> > <map:redirect-to uri="{2}.html" />
> > </map:match>
> > <map:match pattern="*">
> > <map:redirect-to uri="{1}.html" />
> > </map:match>
> > </map:match>
>
> Regardless of comment above, what you got here is too verbose, you can
> DRY it down using smarter match pattern. For example:
>
> <map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)\/$">
> <map:redirect-to uri="../{1}.html"/>
> </map:match>
> <map:match pattern="^(?:[^.\/]*\/)*([^.\/]+)$">
> <map:redirect-to uri="{1}.html"/>
> </map:match>
>
Hmm, sorry but the above is not working for me.
>
> > One side effect is that the app is now way slower if you use
> > "extensionless requests". In our application we need the extensions to
> > trigger the correct pipelines, which I consider the normal usage of
> > cocoon. By redirecting the request one is causing more processing on
> > the
> > app which explains the higher response time.
>
> You probably mean "more round trips". I don't see any extra processing
> on the app.
Yeah, I meant more round trips.
Thanks for your feedback.
salu2
>
> Vadim
>
>
> > That makes me wonder whether extension less requests makes sense at
> > all
> > in a multi-output-format environment.
> >
> > WDYT?
> >
> > salu2
> > --
> > Thorsten Scherler
> > thorsten.at.apache.org
> > Open Source Java consulting, training and
> > solutions
>
--
Thorsten Scherler thorsten.at.apache.org
Open Source Java consulting, training and solutions