I'm not sure where I suggested it's not appropriate to consider an
extension.  I believe I suggested as you say it might be better to rip
out the function of allowing XML and XSL as input, to provide a separate
convenience jar and/or the code to embed that process, since creating
the FO doesn't involve any fop classes.  I already split out that
process in my embedded code if you want to see the 2 step.  I don't know
what the wiki text looks like or how common it might be for users to
want to use it as input.  If it has a standard format which can be
translated to FO tags and someone wants to write the code to translate
it, it would be nice if code like that could be included as a separate
download on the fop site.  I'm not sure what you consider an extension
(plugin?) if they're normally compiled into the fop.jar, if they're
separate jars to download to call as separate tasks, or if they're
separate jars to put on the classpath to accept different types of
values from the mainline fop (ie the -xml -xsl tags)?
 
I think anything unique to one type of input or output can be a separate
jar with the code kept separately whether you call them extensions,
plugins, or subprojects.  They can either be documented second step
processes (output of this = input of that) or extension options (use
-xml and -xsl tags to send input as xml instead of the -fo tag).

________________________________

From: Glenn Adams [mailto:[email protected]] 
Sent: Tuesday, June 14, 2011 10:47 AM
To: [email protected]
Subject: Re: FOP Extension to handle Wiki Syntax


You know, given the time spent answering questions about XSL and the
XML+XSL -> XSL-FO front-end ("convenience mechanism") in FOP, I
sometimes wonder if it would be better to rip out that function. Perhaps
then folks would understand better that FOP is fundamentally an XSL-FO
-> output format processor. 

As to the original comment, I agree with Eric that is is not appropriate
to consider an FOP extension to accommodate semantics that apply to the
XML+XSL -> XSL-FO 'convenience mapping' mechanism.

G.

On Tue, Jun 14, 2011 at 8:29 AM, Eric Douglas <[email protected]>
wrote:


        It could handle any format you want to write your input in as
long as it
        can be translated to FO.  If another format is commonly used for
        generating FOP input someone would just have to write an input
        translator extension.
        
        FOP doesn't even do anything with XML/XSL.  It accepts input as
XML/XSL
        as a courtesy extension.  I wrote a transform with embedded code
        starting with data in XML using an XSL to translate it.  Then I
figured
        out how to generate FO and split that out as a separate step.
That uses
        the javax transformer.  That step doesn't use any FOP objects.
        
        If you think about it, output could be considered extensions
also.  The
        main task of the FOP is to input FO and generate the IF.  Once
that's
        laid out it can take various renderers and generate output so
you have a
        PDF extension, a PNG extension, a TIFF extension, etc.  They
don't need
        to be in the FOP package.  Someone who only wants to create PDFs
doesn't
        need any classes which create PNGs.  If you could break all the
        extensions out into subprojects it would make it a few extra
steps to
        download but it would be simplified into smaller jars which of
course
        load faster if you don't need them all.
        
        
        -----Original Message-----
        From: Christopher R. Maden [mailto:[email protected]]
        Sent: Tuesday, June 14, 2011 9:44 AM
        To: [email protected]
        Subject: Re: FOP Extension to handle Wiki Syntax
        
        On 06/14/2011 07:25 AM, kalgon wrote:
        > Yes I could transform the XML prior to rendering it to PDF but
that
        > wouldn't be as nice and clean as an extension which would take
care of
        
        > everything. Moreover, an extension can be reused whereas
        > pre-transforming the XML would require a specific XSLT for
each XML
        > schema... definitely not the way I want to go.
        
        Except that's how XSL works, and what FOP implements.  FOP takes
        exactly[*] 1 kind of input: the FO markup in XML defined by the
XSL
        Recommendations.
        
        Nearly all FOP users, as well as users of other XSL formatters,
        transform their source either with XSLT or some other tool into
FO for
        presentation to the formatter.
        
        If FOP supports MediaWiki syntax natively, why not MoinMoin or
some
        other wiki?  Why not HTML, DocBook, DITA, CALS, ...?
        
        ~Chris
        
        [*] approximately
        --
        Chris Maden, text nerd  <URL: http://crism.maden.org/ > "Before
I built
        a wall I'd ask to know / What I was walling in or  out, / And to
whom I
        was like to give offence." - RF, Mending Wall GnuPG Fingerprint:
C6E4
        E2A9 C9F8 71AC 9724 CAA3 19F8 6677 0077 C319
        
        

Reply via email to