[antlr-interest] Re: XML parsing

Thu Jun 12 09:26:23 PDT 2003

This sounds really useful, guys.  Keep us posted.  I know some day I'll have
to deal with XML.  Will be cool to be able to use real tree parsers for it.

Monty

-----Original Message-----
From: Oliver Zeigermann [mailto:oliver at zeigermann.de] 
Sent: Thursday, June 12, 2003 1:14 AM
To: antlr-interest at yahoogroups.com
Subject: [antlr-interest] Re: XML parsing

Matt!

Really good ideas! Better than mine!

Actually, using wildcards and a generic token type for it already works with
XPA and the token type manager. If XPA sees an XML element that is not
definied in the grammar, it tries to map it to the generic token type
"wildcard". To make this known to your grammar you will have to put
something like

tokens {
    "<wildcard>";
}

into it. 

>>From then on you can use "." to match any tag. Using your alternative 1 and
2 together most likely will do the job. 

As a reference have a look at the example in 

examples/treeParser/componentTransformer/ComponentTreeParser.g 

or 

examples/tokenParser/transformingComponentParser/ComponentParser.g

in the XPA distribution.

Oliver

--- In antlr-interest at yahoogroups.com, Matt Benson <gudnabrsam at y...>
wrote:
> It seems there would be a couple of alternatives:
> 
> 1)  Use the wildcard feature of XPA with semantic
> predicates to test actual text value.
> 
> 2)  Modify XPA's XMLTokenTypeManager to map the
> alternatives to one generic token type.
> 
> -Matt
>  
> --- Oliver Zeigermann <oliver at z...> wrote:
> > As with ANTLR every token needs to have a type which
> > can be matched
> > by the parser, there is no way to do it dynamically.
> > You would have to list all the alternatives. If this
> > is your main 
> > problem you might consider using SAX and RE.
> > 
> > --- In antlr-interest at yahoogroups.com, Matt Benson <gudnabrsam at y...>
> > wrote:
> > > I am staying on the list in case any of this is of interest to 
> > > anyone else--if not, I apologize.
> > What I
> > > have, Oliver, is an XML file that has been created
> > so
> > > that tags of similar types, that I could process
> > in
> > > the same way, have different names which are an indication of 
> > > context.  The names might both match
> > a
> > > given RE, but be different.  Can you think of any
> > > slick way I could use the RE with XPA, or would I
> > have
> > > to allow for each possible value as alternatives
> > of
> > > the same rule (doable, but messy)?
> > > 
> > > Thanks,
> > > Matt
> > > 
> > > --- Oliver Zeigermann <oliver at z...> wrote:
> > > > ASTFactory had to be extended in order to allow
> > AST
> > > > creation from
> > > > strings (no type). This is fully compatible with
> > the
> > > > original ANTLR
> > > > code as it simply involves making "AST
> > > > create(String)" public (was 
> > > > protected). You will find the new source of
> > > > ASTFactory in the 
> > > > distribution. 
> > > > 
> > > > You will also find the file "Open-Issues.txt" in
> > the
> > > > distribution
> > > > which summarises what is missing. For short I
> > can
> > > > tell, XPA is very
> > > > stable and functional. I left the status to
> > alpha to
> > > > indicated
> > > > things may still change a bit on request of
> > users.
> > > > So, this is your
> > > > chance to make me adapt XPA to your needs ;)
> > > > 
> > > > Oliver
> > > > 
> > > > --- In antlr-interest at yahoogroups.com, Matt
> > Benson
> > > > <gudnabrsam at y...>
> > > > wrote:
> > > > > The first things I notice are that the latest
> > > > version
> > > > > is named like an alpha, and that the ANTLR jar
> > is
> > > > > labeled as having been patched.  Can you give
> > a
> > > > quick
> > > > > run-down on the current status of XPA and the
> > > > nature
> > > > > of the ANTLR patching taking place?
> > > > > 
> > > > > Thanks,
> > > > > Matt
> > > > > 
> > > > > 
> > > > > --- Oliver Zeigermann <oliver at z...> wrote:
> > > > > > I have implemented a framework of what you
> > > > describe:
> > > > > > 
> > > > > > http://www.zeigermann.de/xpa/index.html
> > > > > > 
> > > > > > It allows you to feed XML SAX events into
> > ANTLR
> > > > > > parsers as token
> > > > > > streams. Optionally, if you do not care for
> > > > space,
> > > > > > you can create an
> > > > > > AST from a SAX parser and transform it using
> > > > ANTLR
> > > > > > tree parsers.
> > > > > > 
> > > > > > Oliver
> > > > > > 
> > > > > > --- In antlr-interest at yahoogroups.com, Matt
> > > > Benson
> > > > > > <gudnabrsam at y...>
> > > > > > wrote:
> > > > > > > I am involved with a project of which one
> > of
> > > > the
> > > > > > most
> > > > > > > daunting components is to take a large XML
> > > > > > document
> > > > > > > and insert its contents in a relational
> > > > database.
> > > > > > The
> > > > > > > structure of the document allows for
> > > > repetitions
> > > > > > of
> > > > > > > many of its elements, and may include the
> > same
> > > > > > types
> > > > > > > of elements with different meanings based
> > on
> > > > > > context.
> > > > > > > I was trying to decide whether something
> > like
> > > > XSLT
> > > > > > > might help with this problem, when I was
> > > > struck by
> > > > > > the
> > > > > > > idea that a recursive-descent parser might
> > be
> > > > the
> > > > > > > thing to use.  This way we could specify
> > > > structure
> > > > > > by
> > > > > > > use of rules, and Antlr's use of
> > parameters
> > > > and
> > > > > > return
> > > > > > > values could allow us to specify a
> > contextual
> > > > > > > interpretation to an element, and receive
> > a
> > > > value
> > > > > > by
> > > > > > > which to associate it with a parent
> > relational
> > > > > > > database record.  Does this sound like a
> > good
> > > > > > > approach?  Can anyone offer a better
> > > > alternative?
> > > > > > > 
> > > > > > > TIA,
> > > > > > > Matt
> > > > > > > 
> > > > > > > __________________________________
> > > > > > > Do you Yahoo!?
> > > > > > > Yahoo! Calendar - Free online calendar
> > with
> > > > sync
> > > > > > to Outlook(TM).
> > > > > > > http://calendar.yahoo.com
> > > > > > 
> > > > > > 
> > > > > >  
> > > > > > 
> > > > > > Your use of Yahoo! Groups is subject to 
> > > > > > http://docs.yahoo.com/info/terms/
> > > > > > 
> > > > > > 
> > > > > 
> > > > > 
> > > > > __________________________________
> > > > > Do you Yahoo!?
> > > > > Yahoo! Calendar - Free online calendar with
> > sync
> > > > to Outlook(TM).
> > > > > http://calendar.yahoo.com
> > > > 
> > > > 
> > > >  
> > > > 
> > > > Your use of Yahoo! Groups is subject to 
> > > > http://docs.yahoo.com/info/terms/
> > > > 
> > > > 
> > > 
> > > 
> > > __________________________________
> > > Do you Yahoo!?
> > > Yahoo! Calendar - Free online calendar with sync
> > to Outlook(TM).
> > > http://calendar.yahoo.com
> > 
> > 
> >  
> > 
> > Your use of Yahoo! Groups is subject to 
> > http://docs.yahoo.com/info/terms/
> > 
> > 
> 
> 
> __________________________________
> Do you Yahoo!?
> Yahoo! Calendar - Free online calendar with sync to Outlook(TM). 
> http://calendar.yahoo.com

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 