Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-06-06

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

All times shown according to UTC.

Time Nick Message
00:00 ronsavage I've been thinking about the '()' issue. Perhaps the () could mean optional capture of the whole rhs? in that case omitting them becomes a choice. This may lead to a problem defining/recognizing the end of a rule, so I'll make a suggestion. I realize syntax will change from SLIF but I'll use SLIF for this.
00:01 ronsavage Currently, a raw \n: makes the end of some rules and the beginning of others. Here, the \n is in the source of the grammar, and the /unquoted/ : precedes some pre-defined words. So perhaps all rules could start with an unquoted char, either the current : or a ! say. With the [:!] in column 1, so to speak, after the \n, rule termination (separation) is clear, and () are left with their newly proposed purpose.
01:49 ilbot3 joined #marpa
01:49 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
03:02 jeffreykegler joined #marpa
03:02 jeffreykegler ronsavage: reading your last, I think you might have a common mis-impression.
03:03 jeffreykegler Newlines in the SLIF have no special significance as whitespace.
03:04 jeffreykegler That is, if you run your grammar through fmt utility, so that your line alignment is ruined ...
03:04 jeffreykegler it will become unreadable, but the SLIF will parse it just fine.
03:04 jeffreykegler This is also the case in Lua, and therefore will continue to be the case with Kollos's LUIF
03:06 jeffreykegler My apologies if I misread your comments.
03:35 jeffreykegler re Kollos progress
03:36 jeffreykegler https://github.com/jeffreykegler/kollos/blob/master/luif_proto.pl
03:36 jeffreykegler I just finished a utility which breaks Lua into lexemes and puts it back together.
03:36 jeffreykegler Whitespace is included, so it "round trips"
03:37 jeffreykegler It successfully round-trips every '*.lua' file in the kollos directory, and therefore all the test files in the Lua 5.1 distribution.
03:38 jeffreykegler This is a first step toward the LUIF which, in addition to being a SLIF replacement, is a Lua superset/preprocessor
03:39 jeffreykegler That is, it allows you to change the grammar using Lua, and for that purpose will simply pass through all "pure Lua" code.
03:40 jeffreykegler One interesting thing I discovered about Lua -- the format of its numeric constants is not terribly well-defined.
03:41 jeffreykegler Lua passes the task on to the C library -- strtod() or strtoul(), so what's allowed depends on the system on which it runs.
03:42 jeffreykegler This is not documented in the Lua book, or in its reference manual -- I learned this by reading the code, and confirmed it from the Lua mailing list.
03:43 jeffreykegler The LUIF, I think, will just support the formats supported by C90 strtod() and strtoul()
04:07 jeffreykegler joined #marpa
04:18 jeffreykegler joined #marpa
06:37 lwa joined #marpa
09:32 ronsavage joined #marpa
09:35 ronsavage jk: You have correctly interpreted my comments. I was trying to clarify in my own mind how to parse a rule without  accidently concatenating it with the next rule, and described things in a way which clearly put too much emphasis on \n as being significant. Still, I hope my point about usage of () remains meaningful.
10:33 ernimril left #marpa
10:34 ernimril joined #marpa
13:16 sirdancealot joined #marpa
17:11 RichardMichaels joined #marpa
22:20 jeffreykegler joined #marpa
22:22 jeffreykegler https://github.com/jeffreykegler/kollos/blob/master/luif_proto.pl
22:22 jeffreykegler Did a touch up on my Lua "round trip" tokenizer.
22:23 jeffreykegler Some of you might find the way I handled Lua singleline and long comments, and long strings, interesting.
22:36 btyler joined #marpa
22:49 ronsavage joined #marpa

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