Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-03-14

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

All times shown according to UTC.

Time Nick Message
00:17 jeffreykegler joined #marpa
02:47 ilbot3 joined #marpa
02:47 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
04:48 jeffreykegler1 joined #marpa
05:20 jeffreykegler joined #marpa
05:36 jeffreykegler joined #marpa
06:44 jdurand joined #marpa
06:45 jdurand Re: http://irclog.perlgeek.de/​marpa/2014-03-13#i_8428118
06:45 jdurand It works if the /initial/ string length is at least as long as the longest token
06:45 jdurand but it does not in the fake input is changed on-the-fly. I.e. I start with a fake input of length 1
06:46 jdurand and expand it on demand to the longest token during lexing lifetime
06:46 jdurand So: given that arguments to Marpa are pointers to a SCALAR
06:47 jdurand would you mind to make the check on input length, that is in slr->g1_lexeme_complete(), not using a constant value, probably determined the very firsttime (read() call)
06:47 jdurand but dynamic, so that input length can change during lexing lifetime ?
06:48 jdurand "but it does not if the fake input is expanded on-the-fly"
06:49 jeffreykegler Can you switch to a token stream model, instead of a character stream model?
06:49 jeffreykegler So that every lexeme_complete has the same, conventional, length?
06:50 jeffreykegler That is, the traditional parsing model, is a token stream, where length is usually not stated,
06:51 jdurand Ah... damned, I suppose you mean just hardocing lexeme_compete() to have arguments 0 for position and 1 for length
06:51 jeffreykegler but can be regarded as 1
06:51 jdurand "lexeme_complete"
06:51 jeffreykegler Right
06:51 jdurand yes it works...
06:51 jdurand jusdt tested right now
06:51 jdurand so this is ok
06:52 jdurand I am testing a new model of parsing with Marpa: I do a initial read() to "initialize" the SLIF
06:52 jdurand and then I loop on terminals_expected()
06:52 jdurand and is forcing lexeme_complete() to always be at position 0 and length 1
06:53 jdurand One question though:
06:53 jdurand what is the fastest: terminals_expected() or pausing before on all lexemes ?
06:53 jeffreykegler pausing before would be faster is my guess
06:54 jeffreykegler Also, you can just attempt each alternative, and react if it is rejected.
06:54 jeffreykegler That can also be faster than checking terminals_expected()
06:55 jdurand Very good. Will continue my tests.... Have ot go, many thanks!!!
06:55 jdurand to go
06:55 jeffreykegler Good night!
17:24 jeffreykegler joined #marpa

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