[antlr-interest] ANTLR -vs- JB

Wed Jul 30 11:11:47 PDT 2003

How complex is the grammar?  I'd say the more complex it is the worse off
you will be using the .y file because the debugging and tweaking of the .y
will get out of hand and eventually overshadow the time it would have taken
you to learn antlr and do the grammar from scratch.  More important are your
timeline, your scope, and the future of this piece of software.  

You can always take a stab at it one way and change your mind if it is not
looking good.

Monty

-----Original Message-----
From: Jason [mailto:jasonriz at yahoo.com] 
Sent: Wednesday, July 30, 2003 9:43 AM
To: antlr-interest at yahoogroups.com
Subject: RE: [antlr-interest] ANTLR -vs- JB

Monty,

Thanks for taking the time to respond.  It's not
actually SQL - it's MDX, very roughly speaking, it's
SQL for OLAP.  I'm also not thrilled with the idea of maintaining the
'hybrid' .y file, but I guess it's better than trying to write my own MDX
grammar. 
Thanks again for the reply.

-jason

--- mzukowski at yci.com wrote:
> There is already an SQL grammar to download from the antlr.org site.
> Personally I wouldn't want to maintain a .y file
> that has been tweaked to
> generate java.  Not to mention that I wouldn't want
> to debug it.  ANTLR's
> strength is in its sane, human readable generated
> code.
> 
> I haven't heard of JB so I have no opinion of it.
> Do you still need C
> output as well?  From antlr you could generate Java
> or C++.
> 
> Monty
> 
> -----Original Message-----
> From: Jason [mailto:jasonriz at yahoo.com]
> Sent: Wednesday, July 30, 2003 6:36 AM
> To: antlr-interest at yahoogroups.com
> Subject: [antlr-interest] ANTLR -vs- JB
> 
> 
> Hello,
> 
> I've been tasked with writing a parser for a SQL
> like
> query language.  Another team of developers in my
> company has already written a c-language parser
> based
> on the grammar using Flex/Bison.  I've obtained
> their
> input files (.y and .l) and obviously it makes sense
> for me to take advantage of them rather than trying
> to
> independently construct the grammar.   My parser,
> however, is to be implemented in Java.
> 
> One of my co-workers found something called 'jb'
> (http://www.cs.colorado.edu/serl/misc/jb.html) which
> purports to convert the output of bison into a java
> parser.  This seems like the ideal solution: all I
> need to do is replace the C in the .y file I
> obtained
> with Java and I'm all set.  A few questions:
> 
> 1) Does anyone on the list have any experience with
> jb.  The documentation seems pretty sketchy and I'm
> not even sure that it's any longer maintained.
> Neither of these fills me with conifidence?
> 
> 2) If I were to use ANTLR, of what use would the .y
> and .l files be?  Would I need to manually translate
> the files to syntax ANTLR understands?  I'd really
> like to use ANTLR because of how well it's
> documented,
> how actively it's maintained, and how helpful the
> user community seems.
> 
> Thanks in advance for any thoughts anyone might have
> 
> -jason
> 
> __________________________________
> Do you Yahoo!?
> Yahoo! SiteBuilder - Free, easy-to-use web site
> design software
> http://sitebuilder.yahoo.com
> 
>  
> 
> Your use of Yahoo! Groups is subject to 
> http://docs.yahoo.com/info/terms/
> 
> 
>  
> 
> Your use of Yahoo! Groups is subject to 
> http://docs.yahoo.com/info/terms/
> 
> 

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com

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