Perl 6 - the future is here, just unevenly distributed

IRC log for #testml, 2010-08-09

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

All times shown according to UTC.

Time Nick Message
03:27 ashleydev joined #testml
05:43 ashleydev_ joined #testml
15:26 ashleydev joined #testml
15:27 ashleydev_ joined #testml
16:15 ashleydev joined #testml
18:02 ashleydev joined #testml
18:26 ingy patch: I am going to add .OK and .HAS(...) support to TestML pm5/6 today
18:26 ingy I think you know that == ... and .EQ(...) are the same
18:26 ingy I am adding ~~ for .HAS
18:27 ingy but nothing for .OK yet...
18:27 ingy maybe ? or ?? someday...
18:27 ingy undecided
18:29 patch ingy: cool, i liked that addition to the spec
18:30 ingy patch: do you blog in esperanto?
18:30 patch ?? is sort of nice because the other operators are two char repetitions as well
18:30 patch and it stands out well
18:31 ingy nod. it's just a little weird and unnecessary
18:32 ingy we can wait on it
18:32 ingy I'm even a bit worried about ~~
18:32 ingy because HAS is not smart
18:33 ingy it's implemented using .index()
18:33 ingy that's it
18:33 patch TestML is also not perl
18:34 ingy ~~ should maybe be .SMART (in the .Future)
18:34 ingy sure, but P6 is the only prior art for ~~
18:34 patch yeah
18:35 patch i think the perl smartmatch operator might be a little too magical for testing
18:36 ingy totally
18:36 ingy testing (especially acmeist testing) needs to be very straightforward
18:37 patch agreed
19:13 ingy I had an idea for pegex....
19:14 ingy we were talking about debug syntax
19:14 ingy how about <:debug rule1> debug this rule
19:15 ingy <::debug rule1> debug this and below
19:16 ingy <:debug! rule1> don't debug this
19:16 patch that syntax could lead well to other <:xxx rule> options
19:16 ingy <::debug! rule1> don't debug below
19:16 ingy patch++ # yes! you got it!
19:17 ingy xxx is just an option in the grammar node hash
19:17 ingy patch: I was thinking of <:ws rule1>
19:18 ingy for perl6ish whitespaceing stuffs
19:19 ingy I think we can have support for a lot of p6 things
19:19 ingy but make them into explicit options
19:19 ingy <::ltm rule1>
19:19 ingy :)
19:19 patch i was just thinking that, about the whitespace and ltm :)
19:20 ingy \o/
19:23 ingy I've recently been thinking of ' ? ', ' * ', and ' + ' as synonyms for <ws>? <ws>* and <ws>+
19:24 ingy but the both allow for redefinition of <ws>
19:24 ingy *they
19:24 patch with or w/o the quotes?
19:24 ingy without
19:25 ingy rule: <name> <COLON> + <expression>
19:26 ingy hmmm
19:26 ingy looks weird
19:26 ingy rule: <name> <COLON> <ws>+ <expression>
19:26 patch that would make the whitespace in the grammar significant
19:26 patch i think that first example should make the colon optional
19:26 ingy well it is to some degree
19:27 patch sorry, not optional
19:27 ingy <foo>? != <foo> ?
19:27 patch 1 or more
19:27 ingy the latter is syntax error
19:28 ingy wait... you think the colon should be optional in pegex?
19:28 ingy or something else?
19:28 patch no...
19:29 patch i think <COLON> ? should make a colon optional as well as <COLON>?, but at the very least i don't think it should perform a completely different action
19:29 ingy nod
19:29 ingy currently <COLON> ? is an error
19:29 ingy which is fine
19:29 ingy strict up front
19:30 ingy looser later
19:30 ingy be back in 2hr
19:30 patch i do think that <COLON>? is better style
21:53 ashleydev joined #testml
22:06 ingy I think that pegex needs a builtin default receiver class, that builds a general purpose AST, based on the grammar.
22:08 ingy er, ww
23:58 ashleydev_ joined #testml

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