You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@forrest.apache.org by Bert Van Kets <be...@vankets.com> on 2002/05/31 10:51:25 UTC

images problem in css during build

Building my new skin (gosh that sounds weird!) I noticed that due to the 
crawler there is a problem with the images that are used in CSS.
when doing a build webapp the images are perfectly copied, but doing a 
build docs they are not.  Since there is no <img> tag calling the images 
they are not put in the docs directory.  Putting the images in the css 
directory or in an images directory in the css directory does not help 
either (of course not).
As a workaround I could add a link to the images on a 1px image somewhere 
hidden and thus make sure they are copied.
Can we solve this?  I suppose we need to adjust Ant to get this 
done?  Suggestions?
Bert

This mail is written in 100% recycled electrons.


Re: images problem in css during build

Posted by Nicola Ken Barozzi <ni...@apache.org>.
From: "Bert Van Kets" <be...@vankets.com>

> In the current forrest skin the images are in the css/images skin.  I am
> not very familiar with the build process, but wouldn't it be possible to
> let Ant just copy certain directories (like the css/images one)?

Could be done, but not to do.
We should fix the Cocoon CLI instead.

> What are we going to do if someone wants a part in the site that is not
> linked to the main site?  Two consecutive builds?

If part of a site is not linked to the other, well it's basically another
site.
So two consecutive builds make sense.

A way of doing it with one build is to make an entry page (start.html) that
contains links to all the starting points for all the sites.

--
Nicola Ken Barozzi                   nicolaken@apache.org
            - verba volant, scripta manent -
   (discussions get forgotten, just code remains)
---------------------------------------------------------------------


RE: images problem in css during build

Posted by Bert Van Kets <be...@vankets.com>.
In the current forrest skin the images are in the css/images skin.  I am 
not very familiar with the build process, but wouldn't it be possible to 
let Ant just copy certain directories (like the css/images one)?
What are we going to do if someone wants a part in the site that is not 
linked to the main site?  Two consecutive builds?
Bert
At 11:04 31/05/2002 +0200, you wrote:
> > From: Bert Van Kets [mailto:bert@vankets.com]
>
> > Building my new skin (gosh that sounds weird!) I noticed that
> > due to the
> > crawler there is a problem with the images that are used in CSS.
> > when doing a build webapp the images are perfectly copied,
> > but doing a
> > build docs they are not.  Since there is no <img> tag calling
> > the images
> > they are not put in the docs directory.
>
>Yep, another missing feature in our beloved crawler ... while this one
>might be a hard to patch, since it would mean building a full-blown user
>agent capable of interpreting CSS.
>
>The 1 pixel idea was one of the things we came across doing our own
>outerthought-skin, but it seemed kind of patchy to me - anyone else an
>idea?
>
></Steven>


RE: images problem in css during build

Posted by Conal Tuohy <co...@paradise.net.nz>.
> >Put the image url in a commented link tag, and I think it 
> will be picked up.
> ><!-- <a href="mystylesheet">Dummy link</a> -->
> 
> Just tested it, they are not picked up.  Without the comment 
> tags it works.
> Bert

Try <a href="mystylesheet.css"></a>.

---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org


Re: images problem in css during build

Posted by Bert Van Kets <be...@vankets.com>.
At 11:02 31/05/2002 +0200, Nicola Ken Barozzi wrote:
>From: "Bert Van Kets" <be...@vankets.com>
>
> > Building my new skin (gosh that sounds weird!) I noticed that due to the
> > crawler there is a problem with the images that are used in CSS.
> > when doing a build webapp the images are perfectly copied, but doing a
> > build docs they are not.  Since there is no <img> tag calling the images
> > they are not put in the docs directory.  Putting the images in the css
> > directory or in an images directory in the css directory does not help
> > either (of course not).
> > As a workaround I could add a link to the images on a 1px image somewhere
> > hidden and thus make sure they are copied.
> > Can we solve this?  I suppose we need to adjust Ant to get this
> > done?  Suggestions?
>
>Put the image url in a commented link tag, and I think it will be picked up.
><!-- <a href="mystylesheet">Dummy link</a> -->

Just tested it, they are not picked up.  Without the comment tags it works.
Bert


>I know, it's a hack, but it will get it working 4 now.
>
>Someone has time to fix-propose something better for link finding in the
>crawler?
>
>--
>Nicola Ken Barozzi                   nicolaken@apache.org
>             - verba volant, scripta manent -
>    (discussions get forgotten, just code remains)
>---------------------------------------------------------------------
>
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
>For additional commands, email: cocoon-dev-help@xml.apache.org


Re: images problem in css during build

Posted by Bert Van Kets <be...@vankets.com>.
At 11:02 31/05/2002 +0200, Nicola Ken Barozzi wrote:
>From: "Bert Van Kets" <be...@vankets.com>
>
> > Building my new skin (gosh that sounds weird!) I noticed that due to the
> > crawler there is a problem with the images that are used in CSS.
> > when doing a build webapp the images are perfectly copied, but doing a
> > build docs they are not.  Since there is no <img> tag calling the images
> > they are not put in the docs directory.  Putting the images in the css
> > directory or in an images directory in the css directory does not help
> > either (of course not).
> > As a workaround I could add a link to the images on a 1px image somewhere
> > hidden and thus make sure they are copied.
> > Can we solve this?  I suppose we need to adjust Ant to get this
> > done?  Suggestions?
>
>Put the image url in a commented link tag, and I think it will be picked up.
><!-- <a href="mystylesheet">Dummy link</a> -->

Just tested it, they are not picked up.  Without the comment tags it works.
Bert


>I know, it's a hack, but it will get it working 4 now.
>
>Someone has time to fix-propose something better for link finding in the
>crawler?
>
>--
>Nicola Ken Barozzi                   nicolaken@apache.org
>             - verba volant, scripta manent -
>    (discussions get forgotten, just code remains)
>---------------------------------------------------------------------
>
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
>For additional commands, email: cocoon-dev-help@xml.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org


Re: images problem in css during build

Posted by Stefano Mazzocchi <st...@apache.org>.
Cross-posting because this requirement originated on forrest but hits
Cocoon design.

Nicola Ken Barozzi wrote:

> Someone has time to fix-propose something better for link finding in the
> crawler?

The problem is: CSS may link to images and the Cocoon crawler isn't able
to recognize any hyperlinking behavior which is not included in a stream
of SAX events.

This is due to a design limitation of Cocoon views: only multi-staged
pipelines can have views.

If we had the ability to propose 'variants' of resources read by Cocoon
(via "map:read") then no changes to the Cocoon crawler will be required.

I feel this is the most elegant way to deal with the issue.

The question is: how do we assign a view to a read resource?

 <map:readers>
  <map:reader name="resource" class="default">
   <map:view name="variantA" class="..."/>
   <map:view name="variantB" class="..."/>
  </map:reader>
 </map:readers>

then, the usual, 

 <map:match pattern="...">
  <map:read type="resource" src="..."/>
 </map:match>

where is the sitemap engine to choose which class to use to read the
resource depending on the asked view.

For CSS, we could do

 <map:readers>
  <map:reader name="CSS-reader" class="ResourceReader">
   <map:view name="links" class="CSSLinkReader"/>
  </map:reader>
 </map:readers>

then

 <map:match pattern="**.css">
  <map:read type="CSS-reader" src="{1}.css"/>
 </map:match>

NOTE: asking for a different view might change the MIME-type.

What do you think?

-- 
Stefano Mazzocchi      One must still have chaos in oneself to be
                          able to give birth to a dancing star.
<st...@apache.org>                             Friedrich Nietzsche
--------------------------------------------------------------------


---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org


Re: images problem in css during build

Posted by Stefano Mazzocchi <st...@apache.org>.
Cross-posting because this requirement originated on forrest but hits
Cocoon design.

Nicola Ken Barozzi wrote:

> Someone has time to fix-propose something better for link finding in the
> crawler?

The problem is: CSS may link to images and the Cocoon crawler isn't able
to recognize any hyperlinking behavior which is not included in a stream
of SAX events.

This is due to a design limitation of Cocoon views: only multi-staged
pipelines can have views.

If we had the ability to propose 'variants' of resources read by Cocoon
(via "map:read") then no changes to the Cocoon crawler will be required.

I feel this is the most elegant way to deal with the issue.

The question is: how do we assign a view to a read resource?

 <map:readers>
  <map:reader name="resource" class="default">
   <map:view name="variantA" class="..."/>
   <map:view name="variantB" class="..."/>
  </map:reader>
 </map:readers>

then, the usual, 

 <map:match pattern="...">
  <map:read type="resource" src="..."/>
 </map:match>

where is the sitemap engine to choose which class to use to read the
resource depending on the asked view.

For CSS, we could do

 <map:readers>
  <map:reader name="CSS-reader" class="ResourceReader">
   <map:view name="links" class="CSSLinkReader"/>
  </map:reader>
 </map:readers>

then

 <map:match pattern="**.css">
  <map:read type="CSS-reader" src="{1}.css"/>
 </map:match>

NOTE: asking for a different view might change the MIME-type.

What do you think?

-- 
Stefano Mazzocchi      One must still have chaos in oneself to be
                          able to give birth to a dancing star.
<st...@apache.org>                             Friedrich Nietzsche
--------------------------------------------------------------------


Re: images problem in css during build

Posted by Nicola Ken Barozzi <ni...@apache.org>.
From: "Bert Van Kets" <be...@vankets.com>

> Building my new skin (gosh that sounds weird!) I noticed that due to the
> crawler there is a problem with the images that are used in CSS.
> when doing a build webapp the images are perfectly copied, but doing a
> build docs they are not.  Since there is no <img> tag calling the images
> they are not put in the docs directory.  Putting the images in the css
> directory or in an images directory in the css directory does not help
> either (of course not).
> As a workaround I could add a link to the images on a 1px image somewhere
> hidden and thus make sure they are copied.
> Can we solve this?  I suppose we need to adjust Ant to get this
> done?  Suggestions?

Put the image url in a commented link tag, and I think it will be picked up.
<!-- <a href="mystylesheet">Dummy link</a> -->

I know, it's a hack, but it will get it working 4 now.

Someone has time to fix-propose something better for link finding in the
crawler?

--
Nicola Ken Barozzi                   nicolaken@apache.org
            - verba volant, scripta manent -
   (discussions get forgotten, just code remains)
---------------------------------------------------------------------



---------------------------------------------------------------------
To unsubscribe, e-mail: cocoon-dev-unsubscribe@xml.apache.org
For additional commands, email: cocoon-dev-help@xml.apache.org


Re: images problem in css during build

Posted by Nicola Ken Barozzi <ni...@apache.org>.
From: "Bert Van Kets" <be...@vankets.com>

> Building my new skin (gosh that sounds weird!) I noticed that due to the
> crawler there is a problem with the images that are used in CSS.
> when doing a build webapp the images are perfectly copied, but doing a
> build docs they are not.  Since there is no <img> tag calling the images
> they are not put in the docs directory.  Putting the images in the css
> directory or in an images directory in the css directory does not help
> either (of course not).
> As a workaround I could add a link to the images on a 1px image somewhere
> hidden and thus make sure they are copied.
> Can we solve this?  I suppose we need to adjust Ant to get this
> done?  Suggestions?

Put the image url in a commented link tag, and I think it will be picked up.
<!-- <a href="mystylesheet">Dummy link</a> -->

I know, it's a hack, but it will get it working 4 now.

Someone has time to fix-propose something better for link finding in the
crawler?

--
Nicola Ken Barozzi                   nicolaken@apache.org
            - verba volant, scripta manent -
   (discussions get forgotten, just code remains)
---------------------------------------------------------------------



RE: images problem in css during build

Posted by Steven Noels <st...@outerthought.org>.
> From: Bert Van Kets [mailto:bert@vankets.com]

> Building my new skin (gosh that sounds weird!) I noticed that
> due to the
> crawler there is a problem with the images that are used in CSS.
> when doing a build webapp the images are perfectly copied,
> but doing a
> build docs they are not.  Since there is no <img> tag calling
> the images
> they are not put in the docs directory.

Yep, another missing feature in our beloved crawler ... while this one
might be a hard to patch, since it would mean building a full-blown user
agent capable of interpreting CSS.

The 1 pixel idea was one of the things we came across doing our own
outerthought-skin, but it seemed kind of patchy to me - anyone else an
idea?

</Steven>