Comp.compilers newsgroup

Making semicolons optional moves LALR(1) language to LALR(2)?

Sat, 6 Feb 2010 14:59:58 +0000 (UTC)

For a hypothetical programming language that is LALR(1) and uses semicolons as statement terminators, would a change that makes semicolons only required on multi-statement lines and using the newline as an implicit statement terminator make the language LALR(2)?

Infinite look ahead required by C++?

Sat, 6 Feb 2010 14:58:38 +0000 (UTC)

What elements of C++ make it so hard to parse? Is it a weakness of compiler designs rather than a weakness of the language design? I've read somewhere that the language requires potentially infinite look ahead. Why? And how do compilers handle it? [It's ambiguous syntax. Others can doubtless fill in the details. -John]

Re: An example LL(K) language that is not LL(K-1) ?

Sat, 6 Feb 2010 14:57:58 +0000 (UTC)

On 2010-02-01, Chariton Karamitas wrote: > Hello, > > I don't think your assumption that any LL(k) can be transformed into > an LL(k-1) is correct. The 'k' in LL(k) is assumed to be the supremum > of lookahead symbols that you need in order to parse your input. So,

Re: An example LL(K) language that is not LL(K-1) ?

Sat, 6 Feb 2010 14:56:12 +0000 (UTC)

SLK Mail writes: > Example: > > S -> a A a > S -> b A b a > A -> b > A -> > > Can you convert this to LL(1)? Yes, there is no recursion in this grammar, therefore the language is trivially regular and thus trivally, LL(1).

Re: An example LL(K) language that is not LL(K-1) ?

Fri, 5 Feb 2010 22:33:07 +0000 (UTC)

The following is the reference for the original paper on LL(k). It should have a proof of the superset relationship of k to k-1 for the LL languages. Rosenkrantz, D.J. and R.E. Stearns (1970). "Properties of Deterministic Top-Down Grammars," Inf. and Control, 17 (3), pp

Re: An example LL(K) language that is not LL(K-1) ?

Fri, 5 Feb 2010 22:32:32 +0000 (UTC)

On Feb 2, 6:57 am, klyjikoo wrote: > > Thanks to Hans, Consider this example : You have indeed offered an example of and LL(2) grammar that is equivalent to an LL(1) grammar. However both grammars (if the derivation is correct) accept the same

Re: An example LL(K) language that is not LL(K-1) ?

Fri, 5 Feb 2010 22:32:14 +0000 (UTC)

On 2010-02-02, klyjikoo wrote: > >> I don't think your assumption that any LL(k) can be transformed into >> an LL(k-1) is correct. The 'k' in LL(k) is assumed to be the supremum >> of lookahead symbols that you need in order to parse your input. So,

Re: An example LL(K) language that is not LL(K-1) ?

Fri, 5 Feb 2010 22:31:52 +0000 (UTC)

Although, I'm not an LL-language expert, I believe the trick to LL(k) languages that aren't LL(k-1) is to have a (center) recursive rule that needs k tokens to disambiguate. Something like: S := A A := B C B := b A d // 1 form of recursion B := b a A a d // other form of recursion

Re: Detailed algorithms on inline optimization

anton@mips.complang.tuwien.ac.at (Anton Ertl) — Wed, 3 Feb 2010 04:09:50 +0000 (UTC)

anton@mips.complang.tuwien.ac.at (Anton Ertl) writes: >Kaz Kylheku writes: >>On 2010-01-25, Anton Ertl wrote: >>> Chris Dodd writes: >>>>unsigned fib(unsigned n) { return n < 2 ? n : fib(n-2) + fib(n-1); }

Re: An example LL(K) language that is not LL(K-1) ?

Wed, 3 Feb 2010 04:09:32 +0000 (UTC)

> I don't think your assumption that any LL(k) can be transformed into > an LL(k-1) is correct. The 'k' in LL(k) is assumed to be the supremum > of lookahead symbols that you need in order to parse your input. So, > suppose you have an LL(2) grammar, then you cannot convert it to an