Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-05-04

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

All times shown according to UTC.

Time Nick Message
02:23 jeffreykegler joined #marpa
07:46 vivtek joined #marpa
08:22 ilbot3 joined #marpa
08:22 Topic for #marpa is now Logs: http://irclog.perlgeek.de/marpa/today Nopaste: http://scsys.co.uk:8002/marpa Stable release: https://metacpan.org/pod/Marpa::R2 Developer release, 2014-02-14: https://metacpan.org/releas​e/JKEGL/Marpa-R2-2.079_015 Source: https://github.com/jeffreykegler/Marpa--R2
15:51 jeffreykegler joined #marpa
16:05 jeffreykegler1 joined #marpa
17:10 jeffreykegler joined #marpa
19:39 jdurand joined #marpa
19:43 jeffreykegler1 joined #marpa
19:49 jeffreykegler joined #marpa
20:10 jdurand Jeffrey, how do marpa_b_ambiguity_metric() and marpa_o_ambiguity_metric() differ ? Could that be a side effect of using ranking ? Thx.
21:49 ronsavage joined #marpa
22:24 jdurand Also: about marpa_v_arg_0(), marpa_v_arg_n(), marpa_v_result(), is it guaranteed that the number in return will always be >= 0 ?
22:36 jdurand FYI I ave revisited my C wrapper to Marpa, almost finished now, I am dealing on how to make the value phase user-friendly -; current API is at https://github.com/jddurand/MarpaX-Languages-​XML-AST-XS/blob/master/include/marpaWrapper.h, and implementation (not an easy thing to read) at https://github.com/jddurand/MarpaX-Language​s-XML-AST-XS/blob/master/src/marpaWrapper.c - totally untested, just compiles perfectly
23:07 jeffreykegler joined #marpa
23:09 jeffreykegler jdurand: re marpa_b_ambiguity_metric() and marpa_o_ambiguity_metric() -- the part you can rely on is the same for each.  In documenting these I was deliberately vague about return values for ambiguous parses.  All you should rely on is a return value > 1.
23:11 jeffreykegler The problem is that an exact count of the ambiguities is expensive, but that of the cheap metrics it is far from clear which is the best.  So I just left both of these so that they distinguish ambiguous from unambiguous cheaply, and left the rest for later (or never).
23:13 jdurand That's fine - but is is possible that one would return > 1 and the other not ? I.e. can I rrely on calling only one and not the other ?
23:14 jeffreykegler jdurand: Stack locations are guaranteed to be non-negative integers and the bottom of the stack is guaranteed to be zero.
23:14 jeffreykegler The documentation of this is here: http://jeffreykegler.github.io/Marpa-w​eb-site/libmarpa_api/latest/Maintainin​g-the-stack.html#Maintaining-the-stack
23:15 jeffreykegler jdurand: re marpa_b_ambiguity_metric() and marpa_o_ambiguity_metric() -- good question -- let me check.
23:18 jdurand Thx for the stack explanation. I am happy that my generic stack piece of code fits perfectly Marpa's design
23:18 jeffreykegler jdurand: OK -- I am back.  re marpa_b_ambiguity_metric() and marpa_o_ambiguity_metric(), it's a good thing you re-asked the way you did because ...
23:19 jeffreykegler after ranking an ambiguous parse might become unambiguous in which case marpa_b_ambiguity_metric() and marpa_o_ambiguity_metric() return different values.
23:20 jdurand Ok, that was what I suspect after reading marpa_o_rank(). Very good. Many thanks.
23:20 jeffreykegler I would have forgotten this little detail had you not reminded me. :-)
23:24 jdurand I can understand that marpa_o_high_rank_only_set() can make an ambiguous grammar unambiguous. But the notion of "null ranks high" escape in my mind: can it have the same effect?
23:30 jdurand And also about: marpa_o_high_rank_only_set(), what is the default value. 0 ?
23:30 jeffreykegler Yes, the "null variants" rank differently, and the low ranking null variants will be discarded in case of null ranks high.
23:32 jeffreykegler For marpa_o_high_rank_only(), the default value is 1 -- this does not seem to be documented, which I'll fix.
23:36 jdurand Thanks! I have updated my wrapper accordingly. Exciting next phase is test and its usuability in a real application
23:36 jeffreykegler Great.  Looking forward!
23:38 jdurand Yes. Going to bed, see you.
23:39 jdurand hmmm going to sleep - I do not know how to word that correctly in english, this was a brutal translation from french! -; ok AFK

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