[ragel-users] ragel scanner construct question

Pramukta Kumar prak at mac.com
Tue Apr 21 01:48:45 UTC 2009

I understand that an algebraic expression can be described by a  
context free grammar and not by a regular grammar (I thought due to  
the recursive requirement).  However, i'm only referring to the  
validation of the next symbol.  All I want to express within ragel is  
that within the buffer, "2 3" shouldn't match since it doesn't make  
sense for a number to be directly followed by another number whereas  
"2 +" or "2 )" would make sense.  I can see doing this via the  
lookahead strategy you talk about in the manual, but that seems to be  
a little excessive.  It just seems like all I am making use of to  
implement this portion is a next-state table so it can't be so off the  
wall.  Am I still totally off the mark?


also, is the dragon book still the best place to look for this stuff?

On Apr 20, 2009, at 7:16 PM, Adrian Thurston wrote:

> You can't do that in Ragel. To understand why you'll have to do some
> reading on context-free VS regular languages. Pick up a compiler book.
> It's a big topic.
> -Adrian
> Pramukta Kumar wrote:
>> I am experimenting with ragel to validate and tokenize an algebraic
>> expression in one step.  Basically convert something like "3 + A *  
>> 2 /
>> ( 1 - 5 )" to "[3.0, 'A' , 2.0, :*, 1.0, 5.0, :-, :/, :+]" if it's
>> valid, and throw an error otherwise. I'm using the ruby target, and
>> the standard "shunting yard" algorithm or whatever.
>> I have the thing working (I think, still testing) by using the  
>> scanner
>> construct and making an "allowed transition" table by hand.  Each of
>> the tokenizing actions validates against this table to see whether it
>> should continue.  It seems to me that there is probably a nice way to
>> do this within ragel itself but I can't seem to figure it out.  Can
>> somebody point me in the right direction?
>> the files are up on github: http://github.com/prakatmac/expression-parser/tree/master
>> Sorry if this is a dumb question.  I'm pretty new at this stuff.
>> Thanks.
>> ~pramukta
>> _______________________________________________
>> ragel-users mailing list
>> ragel-users at complang.org
>> http://www.complang.org/mailman/listinfo/ragel-users
> _______________________________________________
> ragel-users mailing list
> ragel-users at complang.org
> http://www.complang.org/mailman/listinfo/ragel-users

More information about the ragel-users mailing list