Partial string literals

[mike.p...@gmail.com]

What's the easiest way to accept a partial string literal?

For example, I currently have rules like this one, that matches 'p', 'pr', 'pro' or 'prod'.

prod
= 'p'i ('r'i ('o'i ('d'i)?)?)? { return 'prod'; }

Is there a more general way of doing this? It's rather painful to maintain for anything more than a 3 or 4 character literal.

[Zak Greant]

Hi Mike,

It might be easiest to do the matching in Javascript. For example:

{
// return an array of words that match prefix str
function lookup(str){
var words = ['alpine', 'apple', 'apricot'];

return words.filter( function(word){
return word.indexOf( str ) === 0;
});
}
}

start
= str:[a-z]i+ { return lookup( str.join('') ); }

Cheers!
--zak

[Scott Noel-Hemming]

Depending on the surrounding rules I would think this would also work:

prod
= ( 'prod' / 'pro' / 'pr' / 'p' ) { return 'prod'; }

--
() ascii ribbon campaign - against html e-mail
/\ www.asciiribbon.org - against proprietary attachments

[David Majda]

2012/12/2 <mike.p...@gmail.com>:

While there are some alternatives (as pointed out by others), there
isn't a non-painful way to do this in PEG.js, short of using some
templating engine to create rules like this mechanically or generating
the grammar from some source data.

What exactly are you trying to accomplish?

--
David Majda
Entropy fighter
http://majda.cz/