Re: yet another pattern-matching library for Lua

["Duncan Cross" <duncan.cross@...>]

I'm impressed, LPeg looks like a very powerful addition.

However, this seems like a bug:

>print( lpeg.match("wha&t", (GoodName) * lpeg.P(-1)) )
false

>print( lpeg.match("wha&t", (BadName) * lpeg.P(-1)) )
6

>print( lpeg.match("wha&t", (GoodName + BadName) * lpeg.P(-1)) )
false

...I'm fairly new to the concept of parsing grammars, so I may have
just misunderstood something, but shouldn't the third command get the
same response as the second one, regardless of what the patterns are?

On 12/27/06, Roberto Ierusalimschy <roberto@inf.puc-rio.br> wrote:
> I am releasing a prototype of (yet) another pattern-matching
> library for Lua, called LPeg:
>
>   http://www.inf.puc-rio.br/~roberto/lpeg.html
>
> Because the library is rather unconventional, all things in it are
> experimental (even its name).
>
> Unlike most pattern-matching libraries, LPeg is not based on regular
> expressions, but on Parsing Expression Grammars (which in turn are based
> on TDPLs, a technique using top-down parsers with limited backtracking
> developed in the 70s):
>
> Why LPeg?
>
> - It seems to be quite efficient.
>
> - It has a simple implementation; patterns are translated to programs
> for a small parsing machine.
>
> - It is quite complete. Besides the usual pattern-matching operators
> based on regular expressions (alternation, concatenation, and
> repetition, but with a twisted semantics), it also supports BNF-like
> grammars (again with a twisted semantics; see the docs).
>
> - It has a formal ground. (Unlike most pattern-matching libraries today,
> which are ad-hoc collections of facilities loosely inspired by regular
> expressions.)
>
> Have fun!
>
> -- Roberto