forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <rgard...@apache.org>
Subject Re: RT: Commandline conversion
Date Fri, 02 Mar 2007 14:26:50 GMT
C.Grobmeier wrote:
> Hi all,
> 
> we now use Forrest here at my latest project for software documentation. Before i brought
up Forrest, the guys here used word documents. There is no plugin at the moment for word,
but with POI it may be possible to write one. 

Actually there is a word plugin. The reason it is not in our plugins 
list is because a) there are licence issues that mean we can't host it 
at Apache and b) I never found the time to package it and put it on the 
sourceforge project I created to house it and other similar plugins.

I've not used it for about six months, but see no reason why it should 
not work.

Anyway, this is off topic for this thread. If you want to start another 
thread to find out more and/or to discuss alternative implementations go 
ahead.

> However, i thought it would be quite cool not to interpret the word document everytime
i have a request. Instead of this would like to have a forrest document, but i want to use
Forrest translation possibilitys. If dream of something like:
> 
> $> forrest-convert concept.doc concept.xml

I'm not sure of the value of losing your original source format, unless 
the internal format is better suited to your needs. I guess you know 
your use case best.

Why not use an existing tool such as wget?

In this case you would do "forrest run" followed by "wget 
http://localhost:8888/concept.xml concept.xml"

The downside of this is that you need the source document to be located 
in the Forrest content object. I'm guessing you would like to be able to 
do it on an arbitrary file.

> I don't know if this is possible but i would think that this feature would be an good
enhancement for Forrest, maybe for Forrest 2. Wdyt?

For Forrest 2 it would be really easy, it has been designed to make this 
kind of embedding in different applications possible (and easy). 
However, Forrest2 is only experimental and there is no buy in from the 
community at this point. This means that the functionality is minimal 
and I have not yet addressed the reuse of plugin code from Forrest 1.

Doing it in Forrest1 would be a hack because the crawler we use is not 
very flexible. That is why Cocoon have dropped the crawler in 2.2.

We need to create a new crawler for Forrest and it would certainly be 
within the realms of that crawler to do something like this. Thorsten is 
experimenting with a lab called Droids at the moment that looks very 
promising. However, it is not integrated into Forrest yet and it will be 
a while before that happens.

So, back to the current working code base...

This faq entry may give you some clues about how to hack something 
together now:

http://forrest.apache.org/docs_0_80/faq.html#cli-xconf

My immediate thoughts are to write a script (ANT for cross platform?) 
that will:

- mkdir tempWork
- cd tempWork
- forrest seed-basic
- copy source file in src/content/documentation/xdocs/
- use XSL to create the necessary cli.xconf file (see above faq)
- use XSL to set the start page to foo.bar 
(http://forrest.apache.org/docs_0_80/faq.html#defaultStartPage)
- forrest site
- copy build/site/foo.xml to the required output directory
- cd ..
- rm -Rf tempWork

Like I said, it is really hacky and it'll be fairly slow (due to the 
overhead of starting up Forrest), but I'm fairly sure it will work.

Now, if you fancy working on a better solution using a Droid crawler for 
use in Forrest that would be really valuable to the community as a 
whole, such a contribution would be useful to both Forrest and the 
Forrest2 experiments.

Ross

Mime
View raw message