Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-07-05

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

All times shown according to UTC.

Time Nick Message
00:06 ronsavage joined #marpa
00:12 ronsavage jk: There's a Perl module, but not on CPAN: https://github.com/rns/Marp​aX-Languages-CommonMark-AST. I reached it via http://savage.net.au/Marpa.html. The trick is to not search with 'markdown', but just use 'mark'.
01:53 Aria John Gruber's original markdown is in perl, isn't it?
01:53 Aria And of course there's pandoc.
01:53 Aria If you can wade in to the Haskell.
01:59 ronsavage Yeah, a MetaCPAN search for markdown shows many modules. I believe the above (without checking) is more recent, and also tries to cope with the ambiguities in the original spec.
01:59 ronsavage I looked at pandoc, but was not allowed to register on their google group in order to ask questions. Nasty.
02:36 Aria Yeah. Commonmark is decent.
02:44 jeffreykegler joined #marpa
02:52 ronsavage joined #marpa
02:58 jeffreykegler re http://irclog.perlgeek.de/m​arpa/2015-07-05#i_10850088 -- https://github.com/rns/Marp​aX-Languages-CommonMark-AST
02:59 jeffreykegler I remember discussion of this.  I'm not sure it's complete.
03:05 ceridwen ronsavage, Why weren't you allowed to register on their Google group?
03:55 ronsavage ceridwen: No reason given. IIRC there was a button to click to register, which was inactivated. I don't remember clearly, but I think a few years later I re-visited the group briefly and it looked like the button was clickable, but did not bother trying it. By then, I'd decided to use HTML and or POD for docs, POD being easily convertible to HTML.
03:56 ceridwen That's bizarre.
04:25 ceridwen jeffreykegler (and anyone else who's interested in parsing theory): http://staff.um.edu.mt/mros1/SHEL​F/graham_harrison_ruzzo_1980.pdf
04:26 ceridwen This is a paper that has contains a modified version of CYK that works on grammars not in Chomsky Normal Form, and they relate that algorithm to Earley in one of the later sections.
04:26 ceridwen Also interesting is their use of bit vectors.  I
04:27 ceridwen I've been looking for information on vectorizing parser calculations, mostly to determine if it's possible, and the answer is, "Yes," but with a lot of caveats.
04:27 Aria Oooh.
05:46 rns joined #marpa
05:48 rns jeffreykegler: re http://irclog.perlgeek.de/m​arpa/2015-07-04#i_10850057 -- I'm using cmark from https://github.com/jgm/cmark -- reference C implementation of CommonMark. And yes, https://github.com/rns/Marp​aX-Languages-CommonMark-AST is far from complete.
05:50 rns the latest version is 0.20.0 -- https://github.com/jgm/cmark/releases -- builds from source just fine.
05:58 ronsavage joined #marpa
06:09 rns left #marpa
06:40 ronsavage For other modules, I've been asked to stop shipping Build.PL. Any opinions. I've shipped both Build.PL and Makefile.PL for almost all my modules for years, but with Module::Build no longer in code, it matters to some people.
06:41 ronsavage When I say 'other modules', it's because I'm finally ready to announce MarpaX::Languages::Lua::Parser. I guess I should chop Build.PL, but I thought I'd ask first.
07:04 rns joined #marpa
07:08 rns s/no longer in code/no longer in the Perl core/ ?
07:10 ronsavage MarpaX::Languages::Lua::Parser is up-to-date on github. I'll release it in a few days.
07:15 rns left #marpa
14:27 jeffreykegler joined #marpa
14:30 LLamaRider joined #marpa
15:01 jeffreykegler ronsavage: re http://irclog.perlgeek.de/m​arpa/2015-07-05#i_10850485 -- MarpaX::Languages::Lua::Parser
15:01 jeffreykegler Thanks!
15:02 jeffreykegler re http://irclog.perlgeek.de/m​arpa/2015-07-05#i_10850293 -- Graham, Harrison & Russo, 1980
15:02 jeffreykegler I read it at one point.
15:04 jeffreykegler Note that, with Kollos, and the binarization of rules and elimination of nullables, Libmarpa rules and
15:04 jeffreykegler Earley parse items fit easily into fixed-length bit vectors.
15:05 jeffreykegler Aria: perhaps Graham, Harrison & Russo, 1980 -- http://staff.um.edu.mt/mros1/SHEL​F/graham_harrison_ruzzo_1980.pdf -- is one for the collection?
15:05 Aria Was just adding it a moment ago, actually!
15:06 jeffreykegler I thought you might be anticipating me
15:06 jeffreykegler Thanks
15:06 Aria Ooh. Tell me more about how you're thinking Earley items can lay into bit vectors.
15:06 jeffreykegler Also, isn't Commonmark just a standard, not an implementation?
15:06 Aria Yeah, but it's a standard one can implement cleanly and fully.
15:07 Aria Since the edge cases have actually been specified.
15:07 Aria I've been noodling bit vectors for earley items for a while, and kept not finding clever ways to lay both the rule and the origin set number in.
15:10 jeffreykegler Origin is the hard one, and may have to simply be a full length word
15:11 jeffreykegler Also hard are the source links, which are necessary for practical parsing
15:12 jeffreykegler And which actually require more space
18:09 Aria Yeah :/
20:45 Aria https://www.npmjs.com/package/chart-parsers
22:16 ronsavage joined #marpa
22:48 ronsavage rns: Yes, no longer in recent Perl cores.
22:49 ronsavage JK: Re http://irclog.perlgeek.de/m​arpa/2015-07-05#i_10851333. I think Commonmark has been criticized as still having ambiguities. I'll have to search for the link.
22:51 ronsavage (1) http://spec.commonmark.org/ (2) http://blog.codinghorror.com/standa​rd-markdown-is-now-common-markdown/.
23:21 jeffreykegler joined #marpa
23:41 ceridwen joined #marpa
23:41 ceridwen joined #marpa

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