Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-06-20

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

All times shown according to UTC.

Time Nick Message
00:09 ceridwen joined #marpa
00:58 ronsavage savage.net.au/Perl-modules/MarpaX-Languages-Lua-Parser-1.00.tgz is available. I have not read the docs 'one-last-time', so I haven't released it to CPAN. Also, it's on github at https://github.com/ronsavage/MarpaX-Languages-Lua-Parser
01:48 ilbot3 joined #marpa
01:48 Topic for #marpa is now Start here: http://savage.net.au/Marpa.html - Pastebin: http://scsys.co.uk:8002/marpa - Jeffrey's Marpa site: http://jeffreykegler.github.io/Marpa-web-site/ - IRC log: http://irclog.perlgeek.de/marpa/today
03:11 CQ_ joined #marpa
03:17 jeffreykegler joined #marpa
03:17 jeffreykegler ronsavage: Thanks!
03:18 jeffreykegler Re the question "How do I interpret the output?"
03:39 jeffreykegler joined #marpa
04:09 idiosyncrat joined #marpa
04:10 idiosyncrat ronsavage: Sorry.  Pidgin ate the rest of that question
04:11 idiosyncrat Are you talking about the "tokens list" output?
04:11 idiosyncrat And did my original have a parse tree as output?  Or do I mis-remember?
04:12 idiosyncrat Anyway, as the basis for a better answer to that question, we may want to note that the module is not really intended to be useful in itself ...
04:12 idiosyncrat so much as saving a lot of work for someone who *does* have a very specific output in mind.
07:46 lwa joined #marpa
07:56 lucs joined #marpa
08:10 lwa joined #marpa
08:36 ronsavage joined #marpa
08:40 ronsavage jk: Firstly, I was definitely hoping someone would help with the output's interpretation. And yes, the arrayref I return with the output_tokens() method is what I was referring to. You output a deeply nested set of arrayrefs which I unpack into a list (the arrayref). I will add another method which returns the original tree before my unpacking, but since the default structure is not much use, I will also return a cleaned-up tree.
08:40 ronsavage And yes, perhaps I should add another FAQ about it's intended use!
12:51 koo6 joined #marpa
15:36 idiosyncrat joined #marpa
15:38 idiosyncrat ronsavage: Yes, my original "name,value" tree is verbose, but it is general --
15:38 idiosyncrat an application could traverse it to produce the output it wants.
15:39 idiosyncrat I think that is how our "tokens output" works, in fact.
15:39 idiosyncrat Also, an application could simply hack the grammar with its own semantics.
15:40 idiosyncrat I note, in output_tokens, shift and unshift are used, presumably to preserve the order.
15:40 idiosyncrat These potential require moving every element of the array (Perl has some clever optimizations which sometimes avoid this), and can be much slower.
15:41 idiosyncrat Faster is to use push and pop.  If this produces a list in the wrong order, use reverse.  Even with the reverse this can be much faster.
15:42 idiosyncrat Caveat: I've not measured recently, so you might want to test that push/pop is still as much faster as I think it is.
18:08 idiosyncrat joined #marpa
20:21 lwa1 joined #marpa
20:33 hobbs joined #marpa
20:34 lwa joined #marpa
20:34 sadmac joined #marpa
20:38 shadowpaste joined #marpa
20:41 sivoais_ joined #marpa
21:02 Cheery joined #marpa
21:12 sivoais joined #marpa
21:17 lucs_ joined #marpa
21:34 aredridel joined #marpa
21:36 CQ joined #marpa
21:48 ceridwen joined #marpa
21:48 ceridwen joined #marpa
22:40 idiosyncrat joined #marpa
23:02 koo6 joined #marpa
23:05 btyler joined #marpa
23:20 koo6 joined #marpa
23:27 ernimril joined #marpa
23:34 idiosyncrat joined #marpa

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