Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-10-17

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:30 jeffreykegler rns: Converting from the ASF to a parse forest grammar is a reasonable thing to do, is my conclusion.  Re improving efficiency --
00:31 jeffreykegler you want to avoid full traversals -- an alternative approach would be to switch to the low-level Glade interface and set up a custom traverser, one which tracks your criteria (operator precedence and associativity) and does not descend the tree when the criteria are not met.
00:32 jeffreykegler Remember, in writing your own traverser, to memoize results so that, for each glade, the work is never done twice.
00:34 jeffreykegler Also, with regard to matters like accessibility of one symbol from another in a grammar, you want an implementation of Warshall's algorithm to do transitive closures.  (We'll also need one for the LUIF.)
00:35 jeffreykegler You'll wind up doing transitive closures all the time, so it's probably worth coding up in C, and then having an XS wrapper, Lua wrapper, etc., etc.
00:36 jeffreykegler Aria did a Javascript implementation which may be clearer to work from than my code in Libmarpa, which is blisteringly fast but relies on some data structures that are special to Libmarpa.
02:09 rns joined #marpa
02:11 rns jeffreykegler: re http://irclog.perlgeek.de/​marpa/2014-10-17#i_9522653 — thanks a lot, now I need some time to think it over.
03:11 CQ_ joined #marpa
03:42 ronsavage joined #marpa
09:57 ronsavage joined #marpa
10:47 sivoais joined #marpa
12:01 lwa joined #marpa
12:54 aredridel joined #marpa
18:08 jeffreykegler joined #marpa
18:08 jeffreykegler http://programmers.stackexchange.com/questions/1​35446/parser-combinator-that-looks-like-bnf?rq=1
18:09 jeffreykegler In the above link to Stackexchange, the OP asks whether a parser combinator can act like BNF.
18:10 jeffreykegler Now, given the right underlying algorithm, it could not only look, but perhaps also act like BNF. :-)
18:12 jeffreykegler Combinator parsing is just a way of thinking of parsing, one whose limits happen to match those of left parsing fairly well.
18:12 jeffreykegler But why insist on a limited underlying algorithm?
18:46 Aria YES.
18:46 Aria This is exactly what I've been thinking about.
18:46 Aria I want combinator parsers for non-left grammars.
18:46 Aria I want embedded SQL that doesn't stink ;-)
19:52 flaviu joined #marpa
22:35 flaviu joined #marpa
23:32 jeffreykegler joined #marpa

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary