[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: Re: Issue: restrict <interleave>
I certainly welcome some restrictions for <interleave>, but I have a
question about the reasons you mentioned.
For example, the following pattern still satisfies the imposed restriction
but it still causes a combinatory explosion.
<element>
<interleave>
<ref name="a"/>
<ref name="b"/>
...
<ref name="z"/>
</interleave>
</element>
So my understanding is that your proposed restrictions are not for
making the automaton-based validation possible, but for making a simple
workaround for them.
If so, what is the point of allowing <choice>, as suggested in your note?
It makes the workaround difficult, doesn't it.
Also, I think it would be nice if we can allow the following pattern (in
the normalized form)
<element>
<group>
<!-- interleaved element contents -->
<interleave>
<element A/>
<element B/>
..
</interleave>
<!-- a big tree of attributes -->
<choice>
<attribute A/>
<attribute B/>
</choice>
</group>
</element>
Because this <group> is essentially the same as an <interleave>. I think
this situation happens very frequently because we'are assuming implicit
<group> for <element> patterns. We'll get the above internal form if the
user writes the following external form:
<element>
<interleave>
<element A/>
<element B/>
</interleave>
<ref name="attributes"/>
</element>
<define name="attributes">
<choice>
<attribute A/>
<attribute B/>
</choioce>
</define>
regards,
--
Kohsuke KAWAGUCHI +1 650 786 0721
Sun Microsystems kohsuke.kawaguchi@sun.com
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC