Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-08-27

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

All times shown according to UTC.

Time Nick Message
00:01 jeffreykegler I just switched from desktop to laptop which might (or might not) be the cause of any timestamp confusion.
00:02 ronsavage1 Re this code: http://scsys.co.uk:8002/418977 Why does the last line of the program fail on the double-quotes?
00:02 jeffreykegler Re github: pull requests turn out to be a very, very handy way of dealing with changes -- it took me a while to get comfortable with them, but now they're my favorite.
00:09 jeffreykegler Didn't see your error message, but my guess is that it complained about what's a lexeme.
00:10 jeffreykegler Lexemes have to be a boundary between G1 and L0 -- not on a G1 LHS and never on an L0 RHS.  <generic_id> is on an L0 RHS.
00:10 jeffreykegler Does that sound right?
00:18 ronsavage1 Sorry, should have put the error msg in __DATA__.
00:18 ronsavage1 Exception: Error in SLIF parse: No lexeme found at line 1, column 11
00:18 ronsavage1 * String before error: rankdir =\s
00:18 ronsavage1 * The error was at line 1, column 11, and at character 0x0022 '"', ...
00:18 ronsavage1 * here: "LR"
00:22 jeffreykegler Add the trace_terminals named argument to the recognizer constructor.
00:22 ronsavage1 Forget the time thing. The time I quoted is on the backlog, not pidgin.
00:22 jeffreykegler "trace_terminals => 1" might be enough.
00:23 jeffreykegler "trace_terminals => 99" is very verbose, and will tell you exactly what it found, accepted, rejected, discard, etc.
00:26 ronsavage1 Output from 1 and 2: http://scsys.co.uk:8002/418981
00:34 jeffreykegler The problem may be here "assignment_statement::= generic_id equals_literal generic_id"
00:35 jeffreykegler According to the trace, you're not expecting a <double_quoted_id>, and therefore using LATM won't accept one.
00:36 jeffreykegler You might want "assignment_statement ::= generic_id equals_literal generic_id_token"
00:40 jeffreykegler ronsavage1: AFK, but I'll be back online soon.
00:41 ronsavage joined #marpa
00:53 ronsavage Thanx Jeffrey. Your suggestion worked immediately.
00:58 ronsavage1 joined #marpa
01:06 ronsavage1 Re the new text '(The LR-regular, or LRR, grammars.)'. I take it you mean Marpa /does/ parse LRRs.
01:21 jeffreykegler joined #marpa
01:26 ronsavage1 OK. I've added the intro text to the Marpa page.
01:33 jeffreykegler joined #marpa
01:37 jeffreykegler ronsavage: Would it be clearer if I said "The LR-regular (aka LRR) grammars."?
01:40 Aria I usually just avoid abbreviating, especially something with weird particles like "LR".
01:40 Aria Does anyone actually call them "LRR"?
01:45 jeffreykegler A google search brings up lots of appropriate hits on "LRR grammar".
01:45 jeffreykegler My sense is that the trend is toward LRR, which is more orthogonal with other grammar class names (LL, LR, LALR, GLR, etc.)
01:45 jeffreykegler LR-regular (again just my sense of these things) is the older name, back from when they were unfamiliar.
01:45 Aria Hm.
01:46 Aria LRR, 451k; LR-regular, 1.1m hits.
01:46 jeffreykegler Joop in the title of his paper only claims O(n) for LR(k), because he was worried nobody would know what he was talking about if he said LR-regular, much less LRR.
01:46 Aria Aye.
01:46 jeffreykegler Aria: both hit counts are O(c) :-)
01:46 Aria Haha.
01:47 Aria I guess not everyone had the wikipedia list handy when they wrote ;-)
01:47 jeffreykegler As we have discovered a lot of parsing experts still very closely to the familiar, and don't even consult Wikipedia.
01:48 jeffreykegler I use "LRR" because it's shorter, and I wind up repeating it a lot.
01:50 jeffreykegler In fact, my target readership for the latest Marpa writeup I assume will go into it not knowing what LR-regular or LRR is, and come out of reading it was a rough sense that LRR ~ "just about everything of practical interest, plus some"
01:51 jeffreykegler * was -> with
01:59 * Aria nods.
01:59 Aria That makes sense.
01:59 ronsavage1 Re: "Would it be clearer if I said "The LR-regular (aka LRR) grammars."?" No, I was referring to the problem is the 'unless', and want to emphasize which clause (before/after 'unless') the (...) refers to.
02:01 ronsavage1 There seems to be something missing from 'experts still very closely to the familiar'
02:04 jeffreykegler "experts *STICK* very closely to the familiar"
02:05 jeffreykegler I seem to have graduated from typos to spoonerisms recently -- my errors are all correctly spelled words. :-)
02:07 ronsavage1 And your omitted words have no visible errors either! ROFL.
02:17 jeffreykegler "All unambiguous grammars, unless they have unbounded middle recursions which are unmarked." was intended to mean the same as
02:20 jeffreykegler "All unambiguous grammars, as long as they do not contain any unmarked and unbounded middle recursions."
02:21 jeffreykegler which in turn may be better written as
02:21 jeffreykegler "All unambiguous grammars, as long as they are free of unmarked middle recursions."
03:06 ronsavage joined #marpa
03:09 ronsavage OK. Latest version uploaded, with that 1 change.
03:14 lucs ronsavage: Are you seeing the private messages I'm sending you?
04:03 ronsavage I've updated http://savage.net.au/Marpa.html to claify I'm not the author of Marpa. Thanx lucs.
04:12 jeffreykegler joined #marpa
04:13 jeffreykegler Re authorship, I didn't have a problem with the old page.
04:14 jeffreykegler Compare Perl.  How many Perl discussions mention don't mention Larry prominently?  A lot of them.
04:14 ronsavage With him it's assumed. But for Marpa beginners, better to preempt any misunderstanding.
04:15 jeffreykegler If you're comfortable with the change, that's fine.
04:16 jeffreykegler But, to be clear, I don't see any need to mention me prominently when talking about Marpa.
04:16 jeffreykegler Folks curious about where it came from will find out fast enough.
04:17 * hobbs parses LRR as Loading Ready Run :)
04:17 ronsavage1 joined #marpa
04:19 jeffreykegler hobbs: the naming of class of problems (P, NP, LRR, SLR, GLR, LL, etc. etc.) is quite the zoo, so much so the definitive page on them is called the Complexity Zoo.
04:20 ronsavage1 Are there any wikipedia pages I could add to the new text?
04:21 jeffreykegler I'm not sure if that would not be distracting.
04:21 ronsavage1 And way back to you offer re providing links to tutorials, yes please?
04:21 jeffreykegler Right, yes, you're correct.  I owe you tutorials.
04:22 ronsavage1 Nevertheless, it might give people at least an idea of what to look for, unless you think the Wikipedia text is deficient.
04:22 jeffreykegler Back to wikipedia links on complexity classes: it's my feeling, given that I'm targeting someone new to Marpa, that I already talk too much about classes of grammars.
04:23 ronsavage1 OK.
04:23 ronsavage1 AFK
04:24 jeffreykegler And I want them clicking thru to Marpa tutorials, not Wikipedia articles on grammar classes.
04:32 rns joined #marpa
04:36 jeffreykegler joined #marpa
07:24 ronsavage joined #marpa
07:26 ronsavage Re: "I don't see any need to mention me prominently": OK. All mentions, if any, will be sub-prominent.
07:30 ronsavage But seriously, remember that saying they (allegedly) have in the US Army: Any order that can be misunderstood has been misunderstood. That's what I'm thinking of.
08:22 ronsavage1 joined #marpa
08:30 ronsavage joined #marpa
08:54 shadowpaste joined #marpa
15:20 ilbot3 joined #marpa
15:20 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
17:47 jeffreykegler joined #marpa
20:40 jeffreykegler joined #marpa
23:56 ronsavage1 joined #marpa

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