Christoph's solution is probably the easiest one, but here's an example
of a simple action I wrote to simply call a URL and put the response in
the returned map. Just compile it (need the cocoon jars on the
classpath), jar it up, and stick it in the web-inf/lib folder. If
anyone notices anything bad, by all means let me know - I have hardly
dipped my toes into programming with cocoon, and this was actually the
first class I created for it.
package com.rjlg.mkb.repo.controller.cocoon;
import java.io.ByteArrayOutputStream;
import java.io.InputStream;
import java.net.HttpURLConnection;
import java.net.URL;
import java.util.HashMap;
import java.util.Map;
import org.apache.avalon.framework.activity.Disposable;
import org.apache.avalon.framework.component.ComponentException;
import org.apache.avalon.framework.component.ComponentManager;
import org.apache.avalon.framework.component.Composable;
import org.apache.avalon.framework.parameters.Parameters;
import org.apache.avalon.framework.thread.ThreadSafe;
import org.apache.cocoon.acting.AbstractAction;
import org.apache.cocoon.environment.Redirector;
import org.apache.cocoon.environment.SourceResolver;
public class RemoteXQLAction extends AbstractAction implements
ThreadSafe, Composable, Disposable
{
long created = System.currentTimeMillis();
public void compose(ComponentManager manager) throws
ComponentException {}
public Map act( Redirector redirector, SourceResolver resolver,
Map objectModel, String source, Parameters param ) throws Exception
{
String remoteURL = param.getParameter("url");
if (remoteURL == null) {
getLogger().error("Missing a url param");
return null;
}
Map map = new HashMap();
try{
URL url = new URL(remoteURL);
HttpURLConnection urlConnection =
(HttpURLConnection)url.openConnection();
urlConnection.setDoInput(true);
urlConnection.setDoOutput(false);
urlConnection.connect();
InputStream is = urlConnection.getInputStream();
ByteArrayOutputStream os = new
ByteArrayOutputStream();
byte[] b = new byte[2048];
int ret = 0;
while ((ret = is.read(b)) != -1) {
os.write(b, 0, ret);
}
String response = os.toString();
is.close();
os.close();
getLogger().debug("got back: "+response);
map.put("response", response);
return(map);
} catch (Exception ex){
getLogger().error("Error connecting to/reading
from remote server: "+remoteURL);
return null;
}
}
public void dispose() {
}
}
Chris Marasti-Georg
> -----Original Message-----
> From: Jonas Lundberg [mailto:[EMAIL PROTECTED]
> Sent: Friday, September 30, 2005 3:09 AM
> To: [email protected]
> Subject: Re: Using the output of a pipeline to call another pipeline?
>
> Thank's for the replies,
> Both using an action, and using a flowscript seem promising,
> but also problematic.
>
> For the flowscript solution to work, I need to read the
> xQuery result from a cocoon pipeline. Unfortunately,
> cocoon.processPipelineTo() does not work as in this example:
> readImage() {
> // get request uri in here somehow
> var realfile =
> cocoon.processPipelineTo("flow/locateimage"+cocoon.getRequestUri());
>
> //parse response
> var realFilename = getFilenameFromResponse(realfile);
>
> Cocoon.sendPage("flow/readimage/"+realFileName");
> }
> processPipelineTo wants an outputstream to send the response
> to, rather than storing it in a variable, as I need to do. Is
> there a way around this?
>
> The Action solution also seems promising, but I do not know
> how to create actions, and I cannot find any tutorial about it.
> I would need to know:
> - how to read from a pipeline inside the action, and store
> the result in a variable.
> - how to store that resulting variable in a sitemap variable
> (I have found tutorials describing that specific part)
> - how to compile the action, and where to put it for Cocoon
> to find it.
>
> I would really appreciate any responses!
>
> Hans
>
>
>
> On 9/27/05, Chris Marasti-Georg <[EMAIL PROTECTED]> wrote:
> > I think you have 2 options - a reader, or an action.
> >
> > Your action could call the xquery, get the result, and put
> it in the
> > returned map, making it visible to a subsequent read call
> (my syntax
> > may not be perfect, it's been a while since I coded an action):
> >
> > <map:match pattern="*/*/*.jpg">
> > <map:act type="image-filename-resolver">
> > <map:parameter name="virtual-uri"
> > value="{1}/{2}/{3}.jpg"/>
> > <map:read
> > src="http://myexistserver/exist/db/jpeg/{filename}"/>
> > </map:act>
> > </map:match>
> >
> > Or, implement a special reader that performs both steps...
> Calling the
> > xquery, and then using the result to retrieve the actual
> image. I'm
> > not sure of the inner workings of readers, with respect to reading
> > from a source other than the one that is passed in from the
> sitemap...
> > My guess would be that the action would be a cleaner implementation.
> >
> > One other thought I just had, but have no idea how
> reasonable it would
> > be, is to use a flowscript function like
> >
> > readImage() {
> > // get request uri in here somehow
> > var realfile =
> > cocoon.processPipelineTo("flow/locateimage"+cocoon.getRequestUri());
> >
> > //parse response
> > var realFilename = getFilenameFromResponse(realfile);
> >
> > Cocoon.sendPage("flow/readimage/"+realFileName");
> > }
> >
> > External pipeline
> > <map:match pattern="**.jpg">
> > <map:call name="readImage()"/> </map:match>
> >
> > Internal pipeline:
> > <map:match pattern="flow/locateimage/**">
> > <map:read src="http://myexistserver/exist/db/jpeg/{1}"/>
> > </map:match>
> > <map:match pattern="flow/readimage/**">
> > <map:read src="http://myexistserver/exist/db/jpeg/{1}"/>
> > </map:match>
> >
> > Just some ideas, and I take no blame if nothing works... I
> waited to
> > see if anyone else would pitch in 2 cents, and they didn't,
> so you get
> > my penny instead ;) Good to see another person using eXist...
> > Chris Marasti-Georg
> >
> >
> > > -----Original Message-----
> > > From: Jonas Lundberg [mailto:[EMAIL PROTECTED]
> > > Sent: Tuesday, September 27, 2005 11:54 AM
> > > To: [email protected]
> > > Subject: Re: Using the output of a pipeline to call
> another pipeline?
> > >
> > > Thanks for the replies, I'll explain in more detail then.
> > >
> > > The images are referenced in an HTML page, which means that the
> > > browser will request them separately after the page is
> sent. Do you
> > > propose that I somehow embed the image data in the page?
> > >
> > > All paths in the system are virtual. I use XML files to create
> > > virtual hierarchies, that gives depth and structure to a
> site. That
> > > means that the same contents can be rendered using
> several different
> > > site layouts, at the same time.
> > >
> > > Therefore, all images are stored in one big collection in
> the eXist
> > > database. To avoid that any two images get the same file
> name, each
> > > file is renamed when uploaded, and given a unique name (a
> number).
> > > The actual name, and the virtual path where it is to be shown, is
> > > stored in the meta data file.
> > >
> > > In the html page generation pipeline, I could put the new
> file name
> > > directly in the html page, but that would violate the separation
> > > between storage and presentation. The author of the
> document should
> > > decide what file name to present to the user, not the system. A
> > > strict separation between storage and presentation is a
> main point
> > > with my system.
> > >
> > > Also, if I execute an xquery to provide the file name,
> then I can at
> > > the same time check whether the metadata file states that
> the file
> > > is "live" or "waiting". It should not be possible to retreive an
> > > image before it is published. And when I am doing that, I
> can also
> > > check whether the particular user requesting the file should be
> > > allowed to see it. If the user is not allowed to see the
> image, or
> > > if it is not there, I can simply send a different image.
> > >
> > > So, when the xquery has created an xml file with the image:
> > > <file>
> > > 44500.jpg
> > > </file>
> > > Then I cannot send a HTML file with an <img src...> because the
> > > browser has requested the actual image file. Also, if I do that,
> > > then I have to check the user premissions, and the live-status of
> > > the image again, when the browser requests the actual image.
> > >
> > > If I instead simply cinclude the image then I will get
> <file> image
> > > data </file> That would be an image embedded in XML, which is not
> > > what I want.
> > >
> > > So, what I want to do is very basic:
> > >
> > > First, I want to generate the internal file name using
> xQuery, given
> > > the virtual path and the virtual file name. At the same
> time I check
> > > whether the image is live ,and whether the particular
> user should be
> > > allowed to view it.
> > > Then, I want to call a different (internal) pipeline using the
> > > resulting file name (embedded in the <file></file> tag) (for
> > > instance
> > > 44400.jpg) as the request, and return the result of that pipeline.
> > >
> > > Any ideas now of how I can achieve that?
> > >
> > > Hans
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > On 9/27/05, Derek Hohls <[EMAIL PROTECTED]> wrote:
> > > >
> > > > Its not clear what you mean by the "second pipeline" - do
> > > you mean the
> > > > **.jpg ? If so, can you not simply call that from the page
> > > where it
> > > > needs to appear? (And the page itself is generated using
> > > the output
> > > > from the first pipeline; cinclude or simply a map:aggregate
> > > will feed
> > > > the XML into your page prior to a final transform.)
> > > >
> > > > If you need more help, you may have to lay out your problem
> > > a little
> > > > more clearly.
> > > >
> > > > >>> [EMAIL PROTECTED] 2005/09/27 11:40:16 AM >>>
> > > >
> > > > I've got a very basic question that I got stuck with....
> > > maybe someone
> > > > can help me out?
> > > >
> > > > I have a web publishing system that stores all images in
> > > eXist, using
> > > > a number as the file name, e.g. 44500.jpg
> > > >
> > > > Each image has a metadata file, describing at what URL
> the image
> > > > should be shown (and also some other things, not shown
> here), e.g.
> > > > <metadata>
> > > > <live><at>webdesign/cocoon</at></live>
> > > > <filename>banner.jpg</filename>
> > > > </metadata>
> > > >
> > > > In the html files, the url to the file points at its
> > > virutal location,
> > > > and virtual name, e.g. webdesign/cocoon/banner.jpg Thus, I
> > > could write
> > > > a cocoon sitemap matcher like this:
> > > >
> > > > <map:match pattern="*/*/*.jpg>
> > > > <map:generate src="xq/{1}.xq" type="xquery"> <map:parameter
> > > > name="file" value="{1}/{2}/{3}"/> </map:generate> ? How to call
> > > > the second pipeline with the result of the query?
> > > > </map:match>
> > > >
> > > > The xquery generates the file name to use in the second
> > > pipeline, like
> > > > this <file> 44500.jpg </file>
> > > >
> > > > The image is then to be fetched from eXist using this pipeline:
> > > > <map:match pattern="*.jpg">
> > > > <map:read type="image"
> > > >
> > > >
> > > src="http://guest:[EMAIL PROTECTED]:8080/cocoon/webdav/db/conten
> > > ts/jpeg/{1}.jpg"
> > > > "
> > > > mime-type="text/jpg "/>
> > > > </map:match>
> > > >
> > > > But how do I call the second pipeline, using the file name
> > > created by
> > > > the generator in the first pipeline? This is a very basic
> > > > question, but I got stuck with it anyway... Any ideas?
> > > >
> > > > Hans
> > > >
> > > >
> > >
> --------------------------------------------------------------------
> > > -
> > > > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > > > For additional commands, e-mail: [EMAIL PROTECTED]
> > > >
> > > >
> > > > --
> > > > This message is subject to the CSIR's copyright, terms and
> > > conditions
> > > > and e-mail legal notice.
> > > > Views expressed herein do not necessarily represent the
> > > views of the CSIR.
> > > >
> > > > CSIR E-mail Legal Notice
> > > >
> > > > CSIR Copyright, Terms and Conditions
> > > >
> > > > For electronic copies of the CSIR Copyright, Terms and
> > > Conditions and
> > > > the CSIR Legal Notice send a blank message with "REQUEST
> > > LEGAL" in the
> > > > subject line to CSIR HelpDesk
> > > >
> > > >
> > > > This message has been scanned for viruses and dangerous
> content by
> > > > MailScanner, and is believed to be clean. MailScanner
> > > thanks Transtec
> > > > Computers for their support.
> > >
> > >
> --------------------------------------------------------------------
> > > - To unsubscribe, e-mail: [EMAIL PROTECTED]
> > > For additional commands, e-mail: [EMAIL PROTECTED]
> > >
> > >
> >
> >
> ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
