For XML Schema the context is in general the path up to the root, as the
root must be either a typed element or a global element or the path up
to the first encountered typed element, where by typed element I mean an
element with an xsi:type attribute.
For Relax NG the context is the whole document as both what is before
and what is after determines what can be at a certain point.

But I do not think that the context creates the hardest difficulty but
the name classes and the fact that there is no such thing as global
elements. One can express for instance a name class as nsName test
except element x from test namespace and that covers y from test
namespace. So in the general case it is hard to try to present single
elements and attributes and define them. Also when one specifies a
wildcard in XML Schema that defines only what global elements are
accepted, while in Relax NG we have a single definition (pattern) as a
locally defined common content for all the elements with names defined
by the name class.

As the working unit in Relax NG is the pattern I think that should be
the base for the documentation. Each pattern should be expanded until we
reach elements and attributes and show their name classes then link to
the documentation for the patterns corresponding to those name classes.

Regards,
George
------------------------
George Cristian Bina
http://www.oxygenxml.com


John.Cowan wrote:
> Bob Foster scripsit:
>
>
>>1. The same element name may be reused with many different content
>>models. This is true in XML Schema, as well. The only thing that
>>differentiates the element definitions is the path that reaches the
>>element, and sometimes not even that. The simplest representation is
>>probably to list all of the alternate forms on a single page as
>>alternatives, displaying the path(s) to the element in sufficient detail
>>to distinguish between them, if possible. Actually, just the parent
>>element(s) might be sufficient for most practical purposes.
>
>
> In WXS, the combination of a child element's name and its parent's type
> determines the child's type uniquely.  In RNG there are no guarantees.
>

Bob Foster scripsit:

> 1. The same element name may be reused with many different content
> models. This is true in XML Schema, as well. The only thing that
> differentiates the element definitions is the path that reaches the
> element, and sometimes not even that. The simplest representation is
> probably to list all of the alternate forms on a single page as
> alternatives, displaying the path(s) to the element in sufficient detail
> to distinguish between them, if possible. Actually, just the parent
> element(s) might be sufficient for most practical purposes.

In WXS, the combination of a child element's name and its parent's type
determines the child's type uniquely.  In RNG there are no guarantees.

--
John Cowan  jcowan@...  www.reutershealth.com  www.ccil.org/~cowan
If a soldier is asked why he kills people who have done him no harm, or a
terrorist why he kills innocent people with his bombs, they can always
reply that war has been declared, and there are no innocent people in an
enemy country in wartime.  The answer is psychotic, but it is the answer
that humanity has given to every act of aggression in history.  --Northrop Frye

Thanks, Dave. Now we're getting down to it.

So far we've identified three kinds of documentation that can be
automatically generated from a schema and have exemplars for the two
report forms:

1. End user documentation
     a. Structured report
2. Schema author documentation
     a. Structured report
     b. Hyperlinked document

Notes:

The big difference between end user and schema author documentation is
that the latter refer to schema artifacts, like parameter entities in
DTDs, complex types in XML Schema or defines in RELAX NG. Otherwise, the
reports have much in common: an index list of elements, detailed
descriptions of each element, some form of parent and child lists,
examples of use and any annotations provided in the schema.

Many editors already provide hyperlinking within the editor, but the
hyperlinked document form can be published on the web.

RELAX NG-specific issues:

1. The same element name may be reused with many different content
models. This is true in XML Schema, as well. The only thing that
differentiates the element definitions is the path that reaches the
element, and sometimes not even that. The simplest representation is
probably to list all of the alternate forms on a single page as
alternatives, displaying the path(s) to the element in sufficient detail
to distinguish between them, if possible. Actually, just the parent
element(s) might be sufficient for most practical purposes.

When linking to an element with multiple definitions from a given
context, one should obviously link directly to the forms of the element
that can appear in the context.

2. The same attribute name may have different definitions, including
whether it is optional, in the (identically) same element. This is a
tough one to document. Attributes are commonly used in sets; this set
for this use case, that set for that use case. It gets down to how much
detail of the grammar you want to present, esp. for attributes where
people don't expect any grammar at all. One possible solution would be
just to use three different representations: optional, required and
indeterminate. When the same attribute name has multiple types, show all
of them.

3. Co-constraints. Interesting problem. ;-} In fact, both of the above
are special cases of this problem, which is in general to find a concise
description of the applicable constraints under which an element or
attribute may appear.

4. As always, annotations are the fallback. A simple description of a
grammar may be quite adequate when accompanied by the appropriate text.
This suggests that tools should invest some effort in helping people
annotate schemas effectively. Some people might not want their schemas
cluttered up with the amount of annotation required to produce decent
documentation; also there are some constraints about where annotations
can be placed in compact syntax. This suggests use of an auxiliary
document that would be combined with the schema, e.g., by XSLT, prior to
document generation. Are there any such standard formats? Another thing
worthy of study is what annotation guidelines are in use in the various
schemas, e.g., docbook or tei, to promote effective documentation?

Bob

Dave Pawson wrote:
> On Fri, 2005-07-01 at 08:29 -0700, Bob Foster wrote:
>
>
>>As I recall, this style of documentation for XSLT, showing the valid
>>parents and children, is quite popular, but another example doesn't come
>>readily to Google.
>
>
>
> http://www.docbook.org/tdg/en/html/para.html
>
> They don't come much better.
> (or more useful)
>
>
>

On Fri, 2005-07-01 at 08:29 -0700, Bob Foster wrote:

> As I recall, this style of documentation for XSLT, showing the valid
> parents and children, is quite popular, but another example doesn't come
> readily to Google.


http://www.docbook.org/tdg/en/html/para.html

They don't come much better.
(or more useful)



--
Regards,

Dave Pawson
XSLT + Docbook FAQ
http://www.dpawson.co.uk

All the examples so far, including this one, are of documentation aimed
at XML authors. Here's an entirely different take, aimed at schema
authors (?):

http://www.stylusstudio.com/examples/schemadoc.html

Bob

Sebastian Rahtz wrote:
> Bob Foster wrote:
>
>
>>>Finding the balance between the constraints and the usefulness is the
>>
>>hard bit.
>>
>>Right. That's why I was hoping some user would point to an existing
>>example of documentation produced for DTDs or XML Schema and say, do it
>>like that.
>>
>>
>>
>
> http://www.tei-c.org/P5/Guidelines/REFTAG.html
>
> do it like that
>

Thanks. And (aside from the ugly but maybe useful first page) wow. This
has the look of something produced automatically. Do you happen to have
a link to the original schema(s)?

Bob

Sebastian Rahtz wrote:
> Bob Foster wrote:
>
>
>>>Finding the balance between the constraints and the usefulness is the
>>
>>hard bit.
>>
>>Right. That's why I was hoping some user would point to an existing
>>example of documentation produced for DTDs or XML Schema and say, do it
>>like that.
>>
>>
>>
>
> http://www.tei-c.org/P5/Guidelines/REFTAG.html
>
> do it like that
>

Hi Dave,

A schema documentation tool should present schema information and use
the annotations to show additional developer to user information. The
main idea is to have indeed a linked representation of the schema and a
little more than that if possible.
The main advantage over the schema itself will be that one can freely
share this documentation and will need basically only a browser or in
general a viewer in order to browse the schema (no need for getting the
schema in an authoring environment).

We extended the XS3P XML Schema documentation adding support for
chunking, schema diagrams and Japanese names, see links below. My
intention is to have something similar for Relax NG. My main question is
not how that documentation should be done but what it should look like.
In Relax NG you can have the same name class in different places with
different content each time, there is no such thing as global element.
An attribute can be required if you follow one pattern but can be
optional if you follow some other pattern. If you handle a schema that
has limited Relax NG usage and can be very well supported by a DTD then
the above problems are not seen but if you think at a general schema
then they appear. I currently think to provide a pattern based
documentation only for start, see how that ends up and decide later what
should be added.

I have uploaded the documentation for a few XML schemas. Most of them
are big schemas so I generated the chunked version.

XML Schema
http://www.oxygenxml.com/schemaDoc/xsltAndXsd/XMLSchema.xsd.html
XSLT 2.0
http://www.oxygenxml.com/schemaDoc/xsltAndXsd/XSL2.0Schema.xsd.html
DocBook
http://www.oxygenxml.com/schemaDoc/docbook/docbook.xsd.html
TEI
http://www.oxygenxml.com/schemaDoc/tei/tei.xsd.html
Japanese sample full and chunked versions
http://www.oxygenxml.com/schemaDoc/Japanese/sample.xsd.full.html
http://www.oxygenxml.com/schemaDoc/Japanese/sample.xsd.html

Best Regards,
George
---------------------------------------------------------------------
George Cristian Bina
<oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger
http://www.oxygenxml.com


Dave Pawson wrote:
> On Fri, 2005-07-01 at 14:55 +0300, George Cristian Bina wrote:
>
>
>> > Look to rng itself, and what's available.
>> > We could all do it, if the schema is 'special' to our way of working.
>> > What can you do that is within the rng rec itself?
>
>
>>Sorry but I cannot follow you. Can you please expand a little?
>>My question is if we should focus on patterns only or not. A schema is
>>formed from patterns and inside patterns we can find name classes for
>>elements and attributes. We can present for instance each pattern and
>>specify its content until we reach pattern references. Is that enough?
>>Another possible approach will be to try to determine for instance that
>>there is a person element that has certain attributes and that can
>>contain some other elements and document that (like expanding from a
>>name class following eventual pattern references until we reach other
>>name classes).
>
>
> At some point your output, if XML will begin to look like the rng
> schema,
> perhaps with a variant syntax.
>
> If you constrained the schema author in some way, perhaps using
> the format Eric does in his book,
>
>  <define name="author-element">
>      <element name="author">
>        <attribute name="id"/>
>        <ref name="name-element"/>
>        <ref name="born-element"/>
>        <xhtml:p>After this point, everything is optional.</xhtml:p>
>        <optional>
>          <ref name="died-element"/>
>        </optional>
>      </element>
>    </define>
>
> Then the embedded documentation would make a worthwhile addition to the
schema.
>
> Without some form of additions in a regular format, how can you do other
> than repeat the schema? There may be some value in converting it to a linked
> html, but I doubt it would please many people?
>
> Finding the balance between the constraints and the usefulness is the hard
bit.
>
> regards DaveP
>
>
>
>
>>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>

Bob Foster wrote:

> > Finding the balance between the constraints and the usefulness is the
>hard bit.
>
>Right. That's why I was hoping some user would point to an existing
>example of documentation produced for DTDs or XML Schema and say, do it
>like that.
>
>
>
http://www.tei-c.org/P5/Guidelines/REFTAG.html

do it like that

--
Sebastian Rahtz
Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431

OSS Watch: JISC Open Source Advisory Service
http://www.oss-watch.ac.uk

> Finding the balance between the constraints and the usefulness is the
hard bit.

Right. That's why I was hoping some user would point to an existing
example of documentation produced for DTDs or XML Schema and say, do it
like that.

For example,
http://msdn.microsoft.com/workshop/author/html/reference/elements.asp
provides an alphabetical list of element names and annotations with
links from each element to an alphabetical list of attributes, possible
values and annotations. Obviously, this requires the schema be annotated
to product the text, but even that simple organization would be more
useful to some people than the raw XML.

Here
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/xmlsdk/html/637\
7ce5f-3c45-42a6-b7a9-ec8da588b60c.asp
is a variant of this format for XSLT. Again an alphabetical list of
elements and annotations, with links to a page per element. If you look
at the page in terms of what might be automatically generated, the
element annotation is repeated. The list of attributes and annotations,
and the Element Information section, containing Number of Occurrences,
Parent Elements and Child Elements. The (very useful) example near the
top would probably have to come from an annotation, though one could be
automatically generated showing only required attributes in a RELAX NG
schema the example would have a real risk of being invalid.

As I recall, this style of documentation for XSLT, showing the valid
parents and children, is quite popular, but another example doesn't come
readily to Google.

Please everyone feel free to chime in with examples of documentation you
have seen and found useful.

Bob


Dave Pawson wrote:
> On Fri, 2005-07-01 at 14:55 +0300, George Cristian Bina wrote:
>
>
>> > Look to rng itself, and what's available.
>> > We could all do it, if the schema is 'special' to our way of working.
>> > What can you do that is within the rng rec itself?
>
>
>>Sorry but I cannot follow you. Can you please expand a little?
>>My question is if we should focus on patterns only or not. A schema is
>>formed from patterns and inside patterns we can find name classes for
>>elements and attributes. We can present for instance each pattern and
>>specify its content until we reach pattern references. Is that enough?
>>Another possible approach will be to try to determine for instance that
>>there is a person element that has certain attributes and that can
>>contain some other elements and document that (like expanding from a
>>name class following eventual pattern references until we reach other
>>name classes).
>
>
> At some point your output, if XML will begin to look like the rng
> schema,
> perhaps with a variant syntax.
>
> If you constrained the schema author in some way, perhaps using
> the format Eric does in his book,
>
>  <define name="author-element">
>      <element name="author">
>        <attribute name="id"/>
>        <ref name="name-element"/>
>        <ref name="born-element"/>
>        <xhtml:p>After this point, everything is optional.</xhtml:p>
>        <optional>
>          <ref name="died-element"/>
>        </optional>
>      </element>
>    </define>
>
> Then the embedded documentation would make a worthwhile addition to the
schema.
>
> Without some form of additions in a regular format, how can you do other
> than repeat the schema? There may be some value in converting it to a linked
> html, but I doubt it would please many people?
>
> Finding the balance between the constraints and the usefulness is the hard
bit.
>
> regards DaveP
>
>
>
>
>>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>
>

On Fri, 2005-07-01 at 14:55 +0300, George Cristian Bina wrote:

>  > Look to rng itself, and what's available.
>  > We could all do it, if the schema is 'special' to our way of working.
>  > What can you do that is within the rng rec itself?

> Sorry but I cannot follow you. Can you please expand a little?
> My question is if we should focus on patterns only or not. A schema is
> formed from patterns and inside patterns we can find name classes for
> elements and attributes. We can present for instance each pattern and
> specify its content until we reach pattern references. Is that enough?
> Another possible approach will be to try to determine for instance that
> there is a person element that has certain attributes and that can
> contain some other elements and document that (like expanding from a
> name class following eventual pattern references until we reach other
> name classes).

At some point your output, if XML will begin to look like the rng
schema,
perhaps with a variant syntax.

If you constrained the schema author in some way, perhaps using
the format Eric does in his book,

  <define name="author-element">
      <element name="author">
        <attribute name="id"/>
        <ref name="name-element"/>
        <ref name="born-element"/>
        <xhtml:p>After this point, everything is optional.</xhtml:p>
        <optional>
          <ref name="died-element"/>
        </optional>
      </element>
    </define>

Then the embedded documentation would make a worthwhile addition to the schema.

Without some form of additions in a regular format, how can you do other
than repeat the schema? There may be some value in converting it to a linked
html, but I doubt it would please many people?

Finding the balance between the constraints and the usefulness is the hard bit.

regards DaveP



>
>

Hi Dave,

  >>* Will a documentation of the patterns be enough? Or one should try to
  >>determine the name classes and start from there?
  >
  >
  > Look to rng itself, and what's available.
  > We could all do it, if the schema is 'special' to our way of working.
  > What can you do that is within the rng rec itself?
  >
  > regards DaveP

Sorry but I cannot follow you. Can you please expand a little?
My question is if we should focus on patterns only or not. A schema is
formed from patterns and inside patterns we can find name classes for
elements and attributes. We can present for instance each pattern and
specify its content until we reach pattern references. Is that enough?
Another possible approach will be to try to determine for instance that
there is a person element that has certain attributes and that can
contain some other elements and document that (like expanding from a
name class following eventual pattern references until we reach other
name classes).

Best Regards,
George
---------------------------------------------------------------------
George Cristian Bina
<oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger
http://www.oxygenxml.com

Dave Pawson wrote:
> On Wed, 2005-06-29 at 15:39 -0700, Bob Foster wrote:
>>What format would you (and anyone else who's interested) like to see the
>>documentation in?
>
> XML Bob?
>   Then transform it to what is required.

Sigh. RELAX NG is already XML. What is required?

Bob

/ Dave Pawson <DaveP@...> was heard to say:
| On Wed, 2005-06-29 at 22:25 +0200, Jirka Kosek wrote:
|
|> My student started working on this. Code should be finished during
|> autumn if everything goes well. I did quite extensive research few
|> months ago and I didn't find any really useful package that generates
|> documentation from RNG.
|
| I believe docbook, tdg, is generated like this Jirka, but
| the schema has certain features that are written deliberately
| to enable the trace through the schema.

I think we're drifting away from the original question here. DocBook
NG contains additional markup designed to make it possible to generate
documentation for the scheam from the schema sources. That's distinct
From generating example documents that conform to the schema.

                                         Be seeing you,
                                           norm

--
Norman Walsh <ndw@...> | Greatness of soul is never apparent,
http://nwalsh.com/            | for it conceals itself; a little
                               | originality is usually all that shows.
                               | Greatness of soul is more frequent than
                               | one would suppose.-- Stendhal

On Thu, 2005-06-30 at 16:26 +0300, George Cristian Bina wrote:
> Thanks Sebastian and everybody else that responded to this.

> * Will a documentation of the patterns be enough? Or one should try to
> determine the name classes and start from there?

Look to rng itself, and what's available.
We could all do it, if the schema is 'special' to our way of working.
What can you do that is within the rng rec itself?

regards DaveP

Thanks Sebastian and everybody else that responded to this.
What I mean is not a meta language, but the possibility to generate the
documentation directly from the schema file.
Jirka, we also plan to develop something in this direction.

IMO the problem is not so much how that should be done but what should
be done and this list should be the best place to find out :).

* Will a documentation of the patterns be enough? Or one should try to
determine the name classes and start from there?
* Shall the documentation be one documentation file for one schema file,
a single documentation file for all the schema files or we shall have
documentation for each "grammar"? (Any of these can be eventually chunked).
* should the documentation be based on the physical representation, on
the compiled pattern or on both?
* how references should be solved? Imagine a parentRef in a module
included from different grammars, then finding the reference depends on
the location this module was included from.

And I think there will be more issues once somebody starts working on this.

Best Regards,
George
---------------------------------------------------------------------
George Cristian Bina
<oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger
http://www.oxygenxml.com


Sebastian Rahtz wrote:
> The Text Encoding Initiative schemas are maintained using a module
> of the TEI called "tagdocs", which allows for describing elements,
> attributes,
> classes of elements, 'entities', and so on. The whole of the large TEI
> Guidelines
> are written in the same single document; hence the name of "ODD" for
> this system - One Document Does it all.
>
> The ODD format reverts to Relax NG markup for specifing the content
> model of elements. From this we derive, as needed,
>
>  - documentaton (HTML,  PDF, LaTeX, TEI XML etc)
>  - Relax NG schemas (RNC using trang)
>  - XSD schemas (using trang)
>  - DTDs (direct translation)
>
> There is a complex and powerful system for writing customizations
> (additions, deletions, changes, internationalization), expressed
> in the same language.
>
> Some pointers on http://www.tei-c.org/P5/; or I can describe
> in detail as desired.
>
>

The Text Encoding Initiative schemas are maintained using a module
of the TEI called "tagdocs", which allows for describing elements,
attributes,
classes of elements, 'entities', and so on. The whole of the large TEI
Guidelines
are written in the same single document; hence the name of "ODD" for
this system - One Document Does it all.

The ODD format reverts to Relax NG markup for specifing the content
model of elements. From this we derive, as needed,

  - documentaton (HTML,  PDF, LaTeX, TEI XML etc)
  - Relax NG schemas (RNC using trang)
  - XSD schemas (using trang)
  - DTDs (direct translation)

There is a complex and powerful system for writing customizations
(additions, deletions, changes, internationalization), expressed
in the same language.

Some pointers on http://www.tei-c.org/P5/; or I can describe
in detail as desired.


--
Sebastian Rahtz
Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431

OSS Watch: JISC Open Source Advisory Service
http://www.oss-watch.ac.uk

On Wed, 2005-06-29 at 15:39 -0700, Bob Foster wrote:
> Oh. Well then I know of one editor that does it. ;-}
>
> What format would you (and anyone else who's interested) like to see the
> documentation in?

XML Bob?
   Then transform it to what is required.

regards DaveP

>
>
>

On Wed, 2005-06-29 at 22:25 +0200, Jirka Kosek wrote:

> My student started working on this. Code should be finished during
> autumn if everything goes well. I did quite extensive research few
> months ago and I didn't find any really useful package that generates
> documentation from RNG.

I believe docbook, tdg, is generated like this Jirka, but
the schema has certain features that are written deliberately
to enable the trace through the schema.

I'll look forward to your students work.



--
Regards,

Dave Pawson
XSLT + Docbook FAQ
http://www.dpawson.co.uk

Oh. Well then I know of one editor that does it. ;-}

What format would you (and anyone else who's interested) like to see the
documentation in?

Bob Foster
http://xmlbuddy.com/

George Cristian Bina wrote:
> Hi Bob,
>
> Thanks for mentioning oXygen, but there are a few things in oXygen wrt
> Relax NG support that are missing compared with the support we have for
> DTD and/or XML Schema but they are on our todo list. One is exactly
> this, we generate content only based on a DTD or based on an XML Schema.
> Another one is presenting the schema annotations (comments for DTDs)
> next to the content completion popup when the user selects a proposal.
> And the last one that we have only for XML Schema is support for
> generating schema documentation for Relax NG, eventually with diagrams.
> BTW, does anyone know of an existing project/application that generates
> documentation for a Relax NG schema?
>
> Best Regards,
> George
> ------------------------
> George Cristian Bina
> http://www.oxygenxml.com
>
> Bob Foster wrote:
>
>>Several commercial XML editors do this. I don't know them all but our
>>own XMLBuddy Pro Eclipse plug-in will generate instances from RELAX NG
>>schemas (either syntax), DTDs and XML Schema. I believe Oxygen can as well.
>>
>>Bob Foster
>>http://xmlbuddy.com/
>>
>>Sylvain Hellegouarch wrote:
>>
>>
>>>Hi people,
>>>
>>>I've just registered to the list so I'd like to say to everyone and also
>>>congratulate those who brought RelaxNG to the world :)
>>>
>>>Anyway, I've been playing with Relax for some time now but from the very
>>>begining I have had a question.
>>>
>>>Is it possible (and does it make sense) to generate an XML document from a
Relax
>>>schema?
>>>
>>>Let's say that I have a set of data contained in a non XML format (say a
class
>>>instance) and I'd like to serialise them.
>>>
>>>* I could do it by hand and building myself the XML document.
>>>* I could have an XSLT stylesheet and use parameters.
>>>* I could also have an empty doc, load it as a DOM tree and simply put the
value
>>>at the right place within the doc.
>>>
>>>But my wish would be to define a RELAX document, then with some magic ;)
being
>>>able to take it one hand, my data on the other hand to finally generate the
XML
>>>document while validating the data in one shot. Is it possible?
>>>
>>>Forgive me for being naive but it'd be quite handy. :)
>>>
>>>Thanks for any thoughts.
>>>
>>>- Sylvain

George Cristian Bina wrote:

> BTW, does anyone know of an existing project/application that generates
> documentation for a Relax NG schema?

My student started working on this. Code should be finished during
autumn if everything goes well. I did quite extensive research few
months ago and I didn't find any really useful package that generates
documentation from RNG.

				 Jirka

--
------------------------------------------------------------------
    Jirka Kosek     e-mail: jirka@...     http://www.kosek.cz
------------------------------------------------------------------
    Profesionální školení a poradenství v oblasti technologií XML.
       Podívejte se na náš nově spuštěný web http://DocBook.cz
         Podrobný přehled školení http://xmlguru.cz/skoleni/
------------------------------------------------------------------

Hi Bob,

Thanks for mentioning oXygen, but there are a few things in oXygen wrt
Relax NG support that are missing compared with the support we have for
DTD and/or XML Schema but they are on our todo list. One is exactly
this, we generate content only based on a DTD or based on an XML Schema.
Another one is presenting the schema annotations (comments for DTDs)
next to the content completion popup when the user selects a proposal.
And the last one that we have only for XML Schema is support for
generating schema documentation for Relax NG, eventually with diagrams.
BTW, does anyone know of an existing project/application that generates
documentation for a Relax NG schema?

Best Regards,
George
------------------------
George Cristian Bina
http://www.oxygenxml.com

Bob Foster wrote:
> Several commercial XML editors do this. I don't know them all but our
> own XMLBuddy Pro Eclipse plug-in will generate instances from RELAX NG
> schemas (either syntax), DTDs and XML Schema. I believe Oxygen can as well.
>
> Bob Foster
> http://xmlbuddy.com/
>
> Sylvain Hellegouarch wrote:
>
>>Hi people,
>>
>>I've just registered to the list so I'd like to say to everyone and also
>>congratulate those who brought RelaxNG to the world :)
>>
>>Anyway, I've been playing with Relax for some time now but from the very
>>begining I have had a question.
>>
>>Is it possible (and does it make sense) to generate an XML document from a
Relax
>>schema?
>>
>>Let's say that I have a set of data contained in a non XML format (say a class
>>instance) and I'd like to serialise them.
>>
>>* I could do it by hand and building myself the XML document.
>>* I could have an XSLT stylesheet and use parameters.
>>* I could also have an empty doc, load it as a DOM tree and simply put the
value
>>at the right place within the doc.
>>
>>But my wish would be to define a RELAX document, then with some magic ;) being
>>able to take it one hand, my data on the other hand to finally generate the
XML
>>document while validating the data in one shot. Is it possible?
>>
>>Forgive me for being naive but it'd be quite handy. :)
>>
>>Thanks for any thoughts.
>>
>>- Sylvain
>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>

Several commercial XML editors do this. I don't know them all but our
own XMLBuddy Pro Eclipse plug-in will generate instances from RELAX NG
schemas (either syntax), DTDs and XML Schema. I believe Oxygen can as well.

Bob Foster
http://xmlbuddy.com/

Sylvain Hellegouarch wrote:
> Hi people,
>
> I've just registered to the list so I'd like to say to everyone and also
> congratulate those who brought RelaxNG to the world :)
>
> Anyway, I've been playing with Relax for some time now but from the very
> begining I have had a question.
>
> Is it possible (and does it make sense) to generate an XML document from a
Relax
> schema?
>
> Let's say that I have a set of data contained in a non XML format (say a class
> instance) and I'd like to serialise them.
>
> * I could do it by hand and building myself the XML document.
> * I could have an XSLT stylesheet and use parameters.
> * I could also have an empty doc, load it as a DOM tree and simply put the
value
> at the right place within the doc.
>
> But my wish would be to define a RELAX document, then with some magic ;) being
> able to take it one hand, my data on the other hand to finally generate the
XML
> document while validating the data in one shot. Is it possible?
>
> Forgive me for being naive but it'd be quite handy. :)
>
> Thanks for any thoughts.
>
> - Sylvain

On Wed, 2005-06-29 at 13:24 +0000, mktmurata wrote:
> Thank you for your encouraging mail.
>
> > Is it possible (and does it make sense) to generate an XML document
> from a Relax
> > schema?
>
> Kohsuke created Sun XML Instance Generator, which generates
> XML instances from schemas.  But I am not sure if it supports
> RELAX NG.

My experience is that it supports both DTD's and rng.
Just that once it gets complex, it breaks too much to be of use.
   Pity, could be a really useful tool


--
Regards,

Dave Pawson
XSLT + Docbook FAQ
http://www.dpawson.co.uk

Cheers.

I do realise that this is not the purpose of Relax and I wouldn't be surprised
if nobody else could suggest a solution.

I guess that I'd better workf with XSLT for generating my document and then
validating it with Relax.

Thanks again anyway

- Sylvain

Selon mktmurata <EB2M-MRT@...>:

> Thank you for your encouraging mail.
>
> > Is it possible (and does it make sense) to generate an XML document
> from a Relax
> > schema?
>
> Kohsuke created Sun XML Instance Generator, which generates
> XML instances from schemas.  But I am not sure if it supports
> RELAX NG.  Its README and the following web page do not mention
> RELAX NG, but this program is mentioned by the RELAX NG home page
> (relaxng.org).
>
> http://www.sun.com/software/xml/developers/instancegenerator/
>
> Cheers,
>
> Makoto
>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>




----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

Thank you for your encouraging mail.

> Is it possible (and does it make sense) to generate an XML document
from a Relax
> schema?

Kohsuke created Sun XML Instance Generator, which generates
XML instances from schemas.  But I am not sure if it supports
RELAX NG.  Its README and the following web page do not mention
RELAX NG, but this program is mentioned by the RELAX NG home page
(relaxng.org).

http://www.sun.com/software/xml/developers/instancegenerator/

Cheers,

Makoto

Hi people,

I've just registered to the list so I'd like to say to everyone and also
congratulate those who brought RelaxNG to the world :)

Anyway, I've been playing with Relax for some time now but from the very
begining I have had a question.

Is it possible (and does it make sense) to generate an XML document from a Relax
schema?

Let's say that I have a set of data contained in a non XML format (say a class
instance) and I'd like to serialise them.

* I could do it by hand and building myself the XML document.
* I could have an XSLT stylesheet and use parameters.
* I could also have an empty doc, load it as a DOM tree and simply put the value
at the right place within the doc.

But my wish would be to define a RELAX document, then with some magic ;) being
able to take it one hand, my data on the other hand to finally generate the XML
document while validating the data in one shot. Is it possible?

Forgive me for being naive but it'd be quite handy. :)

Thanks for any thoughts.

- Sylvain



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

On Tue, 28 Jun 2005, Michael Smith wrote:

> I just sent out announcements to the xml-doc and docbook-apps
> mailing lists. That should bring in a few more subscribers. I know
> there are a number of people on those lists who were also
> subscribed to the relaxng-user@... list, and some of them
> may not realize yet that the relaxng-user@... list is not
> working any longer.

And I sent it to TEI-L (email list of the Text Encoding Initiative,
whose markup standard is being revised from a DTD-based to a Relax
NG-based language.)

DS

--
David Sewell, Editorial and Technical Manager
Electronic Imprint, The University of Virginia Press
PO Box 400318, Charlottesville, VA 22904-4318 USA
Courier: 310 Old Ivy Way, Suite 302, Charlottesville VA 22903
Email: dsewell@...   Tel: +1 434 924 9973
Web: http://www.ei.virginia.edu/

I just sent out announcements to the xml-doc and docbook-apps
mailing lists. That should bring in a few more subscribers. I know
there are a number of people on those lists who were also
subscribed to the relaxng-user@... list, and some of them
may not realize yet that the relaxng-user@... list is not
working any longer.

"MURATA Makoto (FAMILY Given)" <EB2M-MRT@...> writes:

> Folks,
>
> This list has 25 members now.  But I strongly believe that there
> are many more users of RELAX NG.  Please mention this mailing list
> to your colleagues who use RELAX NG.
>
>
> Cheers,
>

--
Michael Smith
http://logopoeia.com/  http://www.oreillynet.com/pub/au/890

Folks,

This list has 25 members now.  But I strongly believe that there
are many more users of RELAX NG.  Please mention this mailing list
to your colleagues who use RELAX NG.


Cheers,

--
MURATA Makoto (FAMILY Given) <EB2M-MRT@...>

--
Regards,

Dave Pawson
XSLT + Docbook FAQ
http://www.dpawson.co.uk