Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-04-01

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

All times shown according to UTC.

Time Nick Message
01:47 ilbot3 joined #marpa
01:47 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
02:39 ronsavage joined #marpa
13:39 LLamaRider joined #marpa
14:02 LLamaRider joined #marpa
15:00 LLamaRider joined #marpa
15:12 jeffreykegler joined #marpa
18:54 LLamalessRider joined #marpa
19:22 jdurand joined #marpa
19:31 jdurand Jeffrey, I would like to pause the SLIF at the very beginning of the input, without using an event nor a pause before - is that possible? It seems to only way to start SLIF is to do at least one read() - is that correct ?
20:32 LLamalessRider How is that different from not starting the parsing at all?
20:37 jdurand LlamalessRider: if you can $slr->terminals_expected() before the first read(), Marpa yells
20:37 jdurand I do not need any read at all, I just need terminals_expected(), progress(), lexeme_alternative() and lexeme_complete()
20:38 jdurand but clearly the first read() is doing more than reading. It is instanciating some length, and SLIF will refuse to work without the first read, any SLIF method related to recognizer -;
20:39 jdurand [I meant: "if you call" in the first reply - sorry]
21:09 LLamalessRider right, so there might be a need to call the initialization the first read is calling internally. I see now
21:10 LLamalessRider there should be some API that could be exposed a level higher, let's see what Jeffrey suggests
23:27 jeffreykegler joined #marpa
23:30 jeffreykegler jdurand: re http://irclog.perlgeek.de/​marpa/2014-04-01#i_8525290 -- not sure I understand the question 100%, but I won't let that stop me from trying to answer it :-)
23:30 jeffreykegler If fact, I'll answer in two different ways.
23:31 jeffreykegler First off, I take it you're still making heavy use of events?  If you could switch away from that, the question may become less pressing.
23:31 jeffreykegler Second, to answer more directly ...
23:33 jeffreykegler You can go down to the thin interface, which is documented.  This eliminates one layer of Perl processing.  Note that that still leaves you switching back and forth from Perl to C with every event, though.
23:33 jeffreykegler Third, [ I know I said two, but O(2) = O(3) ] ...
23:34 jeffreykegler I'm open to enhancements which speed up this application, which is an important one.  But I don't think the upper layers are yet stream-lined enough for us to be ready for that.
23:36 jeffreykegler Btw, this time of day my DSL service often gets flaky, so if I suddenly go silent, that's why.
23:52 jeffreykegler1 joined #marpa
23:59 jeffreykegler1 Btw, a mini-essay on compiler/transpiler design, which may or may not be relevant to anything ...

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