Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2016-12-24

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 notviki m: dd [ &infix:<->.candidates.elems, &infix:<−>.candidates.elems ]
00:00 camelia rakudo-moar b306de: OUTPUT«[20, 2]␤»
00:00 AlexDaniel :O
00:00 AlexDaniel right
00:00 AlexDaniel so there's just no Range − something for −
00:01 AlexDaniel but what about other 18 cases?
00:01 notviki They all work out, except for Date
00:02 notviki m: dd Date.today - 42
00:02 camelia rakudo-moar b306de: OUTPUT«Date.new(2016,11,12)␤»
00:02 notviki m: dd Date.today − 42
00:02 camelia rakudo-moar b306de: OUTPUT«Cannot resolve caller Numeric(Date: ); none of these signatures match:␤    (Mu:U \v: *%_)␤  in block <unit> at <tmp> line 1␤␤»
00:03 AlexDaniel notviki: interesting, right? I've never thought that it would be so hard to support ∞ in addition to Inf and − in addition to -
00:04 samcv 10
00:04 samcv oops
00:04 AlexDaniel well, 10 is the right answer
00:04 samcv good!
00:05 dalek rakudo/nom: 91af128 | (Zoffix Znet)++ | src/core/ (2 files):
00:05 dalek rakudo/nom: Fix infix:<U+2212> and prefix:<U+2212>
00:05 dalek rakudo/nom:
00:05 dalek rakudo/nom: Currently, U+2212 infix op exists solely as a numerical op, but we
00:05 dalek rakudo/nom: use it for stuff like Ranges and Dates too. For Ranges, it treats
00:05 dalek rakudo/nom: the Range as a Cool. For Dates multi resolution fails.
00:05 dalek rakudo/nom:
00:05 dalek rakudo/nom: Fix by defining a generic op that just forwards everything to texas version.
00:05 dalek rakudo/nom: Also do the same to prefix:<U+2212> to preemptively avoid future bugs.
00:05 dalek rakudo/nom: Currently it's defined for a single arg, but we teach it to just pass along
00:05 dalek rakudo/nom: all the args for prefix:<->
00:05 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/91af128243
00:06 notviki AlexDaniel: the ∞ isn't bad, but − is just insane. It's all over the place.
00:07 dalek roast: 7a62089 | (Zoffix Znet)++ | S03-operators/u2212-minus.t:
00:07 dalek roast: Add separate file for U+2212 minus test
00:07 dalek roast:
00:07 dalek roast: Currently include tests for infix:<−> and prefix:<−> test, but other
00:07 dalek roast: things involving − may be well moved here.
00:07 dalek roast: review: https://github.com/perl6/roast/commit/7a6208902c
00:08 dalek rakudo/nom: 43ebc63 | (Zoffix Znet)++ | t/spectest.data:
00:08 dalek rakudo/nom: Add S03-operators/u2212-minus.t to test files to run
00:08 dalek rakudo/nom:
00:08 dalek rakudo/nom: Tests for stuff involving U-2212 minus
00:08 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/43ebc63a0a
00:09 samcv notviki++
00:09 AlexDaniel notviki: perhaps there's a way to run every test that has - in it with −
00:10 AlexDaniel and same with every other non-texas operator
00:10 notviki would be nice
00:44 ugexe when multiple threads encounter a locked lock, is it safe to assume that that they will be allowed entry in the order they reach the lock?
00:54 RabidGravy dunno
00:54 RabidGravy I'm not going to look in the code to find out
01:10 dalek roast: 0fec90b | samcv++ | S15-nfg/GraphemeBreakPropertyTest-gen.p6:
01:10 dalek roast: Clean up the Grapheme Break Test script
01:10 dalek roast: review: https://github.com/perl6/roast/commit/0fec90b2d8
01:10 dalek roast: 9920a6e | samcv++ | S15-nfg/GraphemeBreak (2 files):
01:10 dalek roast: Change the name of the output file to be the same as the source test
01:10 dalek roast:
01:10 dalek roast: The file we generate the tests from is GraphemeBreakTest.txt
01:10 dalek roast: Let's call it the same.
01:10 dalek roast: review: https://github.com/perl6/roast/commit/9920a6e32d
01:10 dalek roast: 605d412 | samcv++ | S15-nfg/GraphemeBreakTest.t:
01:10 dalek roast: Add in the new test file
01:10 dalek roast: review: https://github.com/perl6/roast/commit/605d412b1a
01:14 samcv https://github.com/rakudo/rakudo/pull/969 renamed the test, if somebody can merge
01:16 samcv once the merge is done i'll remove the old testname
01:20 samcv almost done with my fun script that will fudge all your tests for you in a file
01:23 dalek rakudo/nom: 286e7af | samcv++ | t/spectest.data:
01:23 dalek rakudo/nom: Update spectest.data to use the renamed GraphemeBreakTest.t
01:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/286e7afe57
01:23 dalek rakudo/nom: aa00700 | (Zoffix Znet)++ | t/spectest.data:
01:23 dalek rakudo/nom: Merge pull request #969 from samcv/GraphemeBreakTest2
01:23 dalek rakudo/nom:
01:23 dalek rakudo/nom: Update spectest.data to use the renamed GraphemeBreakTest.t
01:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/aa007004e1
01:23 notviki samcv: done
01:23 samcv sweet
01:23 dalek rakudo/nom: 266a168 | (Zoffix Znet)++ | t/spectest.data:
01:23 dalek rakudo/nom: Align test category tag
01:23 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/266a168b86
01:24 dalek roast: c058839 | samcv++ | S15-nfg/GraphemeBreakPropertyTest.t:
01:24 dalek roast: Remove old GraphemeBreakPropertyTest.t
01:24 dalek roast:
01:24 dalek roast: Now that the new name is in spectest.data for Rakudo.
01:24 dalek roast: review: https://github.com/perl6/roast/commit/c058839d92
01:48 samcv hmm so far my super-fudger.p6 will double fudge tests and doesn't pay attention to things already fudged… but it's working very nicely otherwise. just need to fix that part
01:48 notviki super fudger? What does it do?
01:49 samcv it runs all the tests and fudges any that don't passs
01:49 samcv that's how i fudged 57 of the 750 or so tests for the test file i just added today
01:49 samcv would be cool if it could unfudge tests too
01:50 notviki IIRC we have that.
01:50 samcv atm it's kind of basic, only works for single line tests
01:50 samcv do we
01:50 notviki the second one I mean, the unfudger
01:50 samcv oh
01:50 samcv ok
01:50 notviki tools/autounfudge.pl maybe?
01:50 samcv there is no tools folder?
01:50 samcv or in rakudo?
01:50 notviki in rakudo repo
01:51 samcv kk
01:55 samcv ok nice i got it working :)
01:57 samcv and not double fudging tests and such
01:58 samcv should I add it to roast? i mean you can specify
01:58 samcv Usage:
01:58 samcv ./super-fudger.p6 <test-file> <backend> <fudge-type> <message>
01:58 samcv it's implementation agnostic
01:58 notviki Sure, why not.
01:59 samcv could be useful especially the autogenerated tests. since I need to regenerate 4 more tests for unicode 9
02:00 dalek roast: 7833fb5 | samcv++ | super-fudger.p6:
02:00 dalek roast: Add super-fudger.p6, to fudge ALL failing tests in a .t file
02:00 dalek roast:
02:00 dalek roast: At the moment only works on tests that can be tested as stand-alone
02:00 dalek roast: lines. If the tests are already fudged it will not try and re-fudge
02:00 dalek roast: them.
02:00 dalek roast: review: https://github.com/perl6/roast/commit/7833fb57ac
02:02 samcv have to check out one of the tests S15-nfg/test-gen.p6 generates. cause it fails 498/500 now. passed all 500 before
02:04 samcv unless something huge changed. agh
02:04 samcv can't be right
02:09 samcv ok if i generate them again from unicode 8.0, they still fail... wtf
02:14 samcv .tell jnthn when I try and generate with roast/S15-nfg/test-gen.p6 only one of the tests (mass-chars.t) actually seem to generate more than 2 passing tests, the rest all fail with either UCD 8.0 or UCD 9.0
02:14 yoleaux2 samcv: I'll pass your message to jnthn.
02:36 dalek roast: 3916198 | samcv++ | S15-unicode-information/uniprop.t:
02:36 dalek roast: uniprop: add tests for Math and Grapheme_Extend
02:36 dalek roast: review: https://github.com/perl6/roast/commit/391619895a
02:36 dalek roast: d889fd2 | samcv++ | S15-unicode-information/uniprop.t:
02:36 dalek roast: uniprop: add tests for Hyphen property, and more rigorous Dash tests
02:36 dalek roast: review: https://github.com/perl6/roast/commit/d889fd2ee2
02:37 samcv hmm maybe for the Emoji ZWJ sequences we can still use uniprop, but just call some other nqp routine to verify it or something
02:38 samcv have a multi for that specific property, so you can check the property of a string... idk. or have its own routine... but that seems a bit odd
02:41 notviki hm... either I'm too wasted or we have a bug in DateTime.in-timezone :/
02:46 notviki # Failed test 'DateTime - Duration = DateTime'
02:46 notviki # at t/spec/S32-temporal/DateTime.t line 818
02:46 notviki # expected: DateTime.new(2015,12,25,3,6,7.77,:timezone(50400))
02:46 notviki #      got: DateTime.new(2015,12,24,8,6,7.77,:timezone(-18000))
02:46 samcv notviki, regression?
02:46 notviki The test failure... no, that's a new test, but it's very simple code and I expected the test to pass..
02:46 notviki ohhh...
02:46 notviki the former hypothesis held true :}
02:46 samcv :P
02:48 kalkin- joined #perl6-dev
02:54 travis-ci joined #perl6-dev
02:54 travis-ci Rakudo build failed. Zoffix Znet 'Align test category tag'
02:54 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186467955 https://github.com/rakudo/rakudo/compare/aa007004e159...266a168b86c6
02:54 travis-ci left #perl6-dev
02:54 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
02:55 samcv weird it looks like it only failed on one of the platforms
02:56 samcv t/04-nativecall/14-rw-attrs.t ............ Failed 16/16 subtests
02:56 samcv but was fine on the other…
02:56 notviki flapper :(
02:57 samcv notviki, you can get travis to retry a test if it fails
02:57 notviki It failed only on one run on Linux. The bottom ones are JVM builds
02:57 notviki (they're not reported in the channel)
02:57 samcv oh
02:57 samcv yeah
02:57 samcv i know that
02:57 notviki samcv: how do we get travis to retry a test if it fails?
02:58 samcv just put travis_retry in front of the command to run :D
02:58 notviki :o
03:01 dalek rakudo/nom: 9f07e1a | (Zoffix Znet)++ | .travis.yml:
03:01 dalek rakudo/nom: Add travis_retry samcv++
03:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9f07e1a56a
03:01 dalek rakudo/nom: d25cfff | (Zoffix Znet)++ | src/core/DateTime.pm:
03:01 dalek rakudo/nom: Remove trailing whitespace
03:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d25cfff863
03:03 dalek rakudo/nom: 6b850ba | (Zoffix Znet)++ | src/core/DateTime.pm:
03:03 dalek rakudo/nom: Implement infix +/- for DateTime/Duration
03:03 dalek rakudo/nom:
03:03 dalek rakudo/nom: Per discussion[^1], add + and - infix ops to handle:
03:03 dalek rakudo/nom: DateTime - DateTime = Duration; DateTime - Duration = DateTime;
03:03 dalek rakudo/nom: DateTime + Duration = DateTime; Duration + DateTime = DateTime;
03:03 dalek rakudo/nom:
03:03 dalek rakudo/nom: [1] https://irclog.perlgeek.de/perl6/2016-12-22#i_13786280
03:03 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6b850babd5
03:03 travis-ci joined #perl6-dev
03:03 travis-ci Rakudo build failed. Zoffix Znet 'Add travis_retry samcv++'
03:03 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186475753 https://github.com/rakudo/rakudo/compare/266a168b86c6...9f07e1a56a0f
03:03 travis-ci left #perl6-dev
03:03 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
03:03 dalek roast: 0a65ed7 | (Zoffix Znet)++ | S32-temporal/DateTime.t:
03:03 dalek roast: Remove trailing whitespace
03:03 dalek roast: review: https://github.com/perl6/roast/commit/0a65ed75c7
03:05 dalek roast: 2188e7e | (Zoffix Znet)++ | S32-temporal/DateTime.t:
03:05 dalek roast: Test +/- infixes on DateTime/Duration
03:05 dalek roast:
03:05 dalek roast: Implemented in https://github.com/rakudo/rakudo/commit/6b850babd5
03:05 dalek roast: Discussion: https://irclog.perlgeek.de/perl6/2016-12-22#i_13786280
03:05 dalek roast: review: https://github.com/perl6/roast/commit/2188e7ef52
03:06 travis-ci joined #perl6-dev
03:06 travis-ci Rakudo build failed. Zoffix Znet 'Remove trailing whitespace'
03:06 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186475774 https://github.com/rakudo/rakudo/compare/9f07e1a56a0f...d25cfff8636c
03:06 travis-ci left #perl6-dev
03:06 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
03:08 notviki samcv: any idea how to fix it?
03:10 travis-ci joined #perl6-dev
03:10 travis-ci Rakudo build failed. Zoffix Znet 'Implement infix +/- for DateTime/Duration
03:10 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186475919 https://github.com/rakudo/rakudo/compare/d25cfff8636c...6b850babd5e8
03:10 travis-ci left #perl6-dev
03:10 dalek rakudo/nom: a2cb929 | (Zoffix Znet)++ | .travis.yml:
03:10 dalek rakudo/nom: Try to make travis_retry work
03:10 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a2cb92944f
03:10 buggable [travis build above] ☠ Did not recognize some failures. Check results manually.
03:12 notviki on second thought....
03:13 dalek rakudo/nom: a9272bf | (Zoffix Znet)++ | .travis.yml:
03:13 dalek rakudo/nom: Remove travis_retry
03:13 dalek rakudo/nom:
03:13 dalek rakudo/nom: Because it'll mute flopping tests. The failures *do* represent problems,
03:13 dalek rakudo/nom: as opposed to failing due to some reason not under our control.
03:13 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a9272bf14d
03:20 samcv uhm hi i'm back now
03:20 samcv let me see
03:21 samcv notviki, don't put it in quotes
03:21 samcv just travis_retry whatever here
03:21 samcv not travis_retry "whatever here"
03:22 samcv at least that's how I did it and it worked for the highlighter
03:22 samcv or probably put it ALL in quotes
03:22 samcv not just the second part
03:22 samcv one of the two
03:24 samcv notviki,, uhm could using GCC 4.6 have anything to do with tests flapping? or probably not. I think not but. do we know why it is flapping?
03:26 samcv would we want to build with gcc-5 if we had the option to?
03:28 * notviki has no idea
03:30 samcv i had to get a later gcc for the docs travis because it didn't support the newer C++ standard
04:14 travis-ci joined #perl6-dev
04:14 travis-ci Rakudo build passed. Zoffix Znet 'Try to make travis_retry work'
04:14 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186476643 https://github.com/rakudo/rakudo/compare/6b850babd5e8...a2cb92944fbb
04:14 travis-ci left #perl6-dev
04:29 samcv notviki, just got the person who makes the font I use for IRC have different lengths of dash for − minus sign than hyphen :) so i am happy
04:31 moritz samcv++ # spreading the Unicode love
04:32 samcv this is the font if you're curious https://camo.githubusercontent.com/7c01c7f3233553bfbd1133f1e48bbbbe6789823a/68747470733a2f2f6c757a2e6c752f746577692f746577692e706e67 https://github.com/lucy/tewi-font
04:32 samcv first link is a picture btw
04:42 travis-ci joined #perl6-dev
04:42 travis-ci Rakudo build passed. Zoffix Znet 'Remove travis_retry
04:42 travis-ci https://travis-ci.org/rakudo/rakudo/builds/186476894 https://github.com/rakudo/rakudo/compare/a2cb92944fbb...a9272bf14d43
04:42 travis-ci left #perl6-dev
04:48 samcv woo. got it actually building with gcc-5. gotta figure out how to make it not do that on mac or whatever but might be nice to have it use ubuntu 14.04.5 not like 12.whatever
05:31 pyrimidine joined #perl6-dev
06:18 pyrimidine joined #perl6-dev
06:45 dalek roast: 359680b | samcv++ | S15-unicode-information/uniprop.t:
06:45 dalek roast: uniprop: add tests for 8 more properties
06:45 dalek roast:
06:45 dalek roast: Extender, Grapheme_Base, Join_Control, Grapheme_Link
06:45 dalek roast: General_Category, Joining_Group, Joining_Type, Sentence_Break
06:45 dalek roast: review: https://github.com/perl6/roast/commit/359680bc45
07:00 dalek roast: b6f2fa4 | samcv++ | S15-unicode-information/uniprop.t:
07:00 dalek roast: uniprop: Fix error in one of the tests
07:00 dalek roast: review: https://github.com/perl6/roast/commit/b6f2fa4500
07:12 samcv notviki, I have my changed .travis.yml that uses newer versions of GCC and other things if it is wanted
07:12 samcv working properly on Linux and MacOS platforms
07:12 samcv and the JVM ones are failing for the same reason as they were failing before
07:21 samcv actually why not go to GCC6 or something i guess
07:55 dalek roast: dd0c572 | samcv++ | S15-unicode-information/uniprop.t:
07:55 dalek roast: uniprop: add tests for 10 more properties
07:55 dalek roast:
07:55 dalek roast: Deprecated White_Space Ideographic Radical Bidi_Mirrored
07:55 dalek roast: Variation_Selector ID_Continue Sentence_Terminal ISO_Comment
07:55 dalek roast: Emoji_Presentation
07:55 dalek roast: review: https://github.com/perl6/roast/commit/dd0c572dc7
08:19 pyrimidine joined #perl6-dev
08:21 samcv almost half way through the properties
08:22 samcv trying to get a full 1/2 of them done by tonight
08:37 pyrimidine joined #perl6-dev
08:39 pyrimidi_ joined #perl6-dev
08:57 lizmat good *, #perl6-dev!
08:57 lizmat Files=1157, Tests=53982, 188 wallclock secs (10.77 usr  4.48 sys + 1124.51 cusr 121.50 csys = 1261.26 CPU)
09:37 RabidGravy joined #perl6-dev
10:11 [Tux] o/
10:11 [Tux] This is Rakudo version 2016.12-84-ga9272bf14 built on MoarVM version 2016.12-19-ge2ffc358
10:11 [Tux] csv-ip5xs        3.065
10:11 [Tux] test            13.410
10:11 [Tux] test-t           5.514
10:11 [Tux] csv-parser      14.489
10:12 samcv [Tux], help
10:13 [Tux] 42
10:13 samcv XD
10:13 samcv those numbers look smaller
10:13 timotimo huggable: speed
10:13 huggable timotimo, http://tux.nl/Talks/CSV6/speed4.html
10:13 timotimo buggable: speed
10:13 buggable timotimo, ▄▄▅▆▅▅▅▆▄▅▅▄▅▄▆▅▅▅▅▄▄▄▄▄▄▄▄▄▄█▃▃▃▄▃▃▃▃▃▃▃▃▃▃▃▃▂▂▂▁ data for 2016-12-02–2016-12-24; variance: 5.514s–7.592s
10:13 timotimo look at that dip!
10:15 lizmat PSA: I've written a blog post that can be used for today's Advent calendar slot
10:16 lizmat those of you with access, please let me know of any comments / suggestions
10:16 lizmat I will publish this in 45 minutes unless ab6tract appears and posts his planned advent post
10:20 dalek rakudo/nom: 4ff2fb2 | samcv++ | / (2 files):
10:20 dalek rakudo/nom: uniprop: set ISO_Comment as Str type
10:20 dalek rakudo/nom:
10:20 dalek rakudo/nom: Since Unicode 5.2.0 this value is required to return an empty string.
10:20 dalek rakudo/nom: Make sure we do this.
10:20 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4ff2fb245a
10:20 dalek rakudo/nom: 742d1b9 | lizmat++ | / (2 files):
10:20 dalek rakudo/nom: Merge pull request #971 from samcv/ISO_Comment
10:20 dalek rakudo/nom:
10:20 dalek rakudo/nom: uniprop: set ISO_Comment as Str type
10:20 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/742d1b908d
10:20 samcv woot
10:26 dalek roast: 701771e | samcv++ | S15-unicode-information/uniprop.t:
10:26 dalek roast: uniprop: add tests for Decomposition_Type, NFC_Quick_Check
10:26 dalek roast: review: https://github.com/perl6/roast/commit/701771e68c
10:26 dalek roast: 8dedf60 | samcv++ | S15-unicode-information/uniprop.t:
10:26 dalek roast: uniprop: unfudge ISO_Comment property
10:26 dalek roast:
10:26 dalek roast: rakudo.moar fixed in rakudo/rakudo@4ff2fb2
10:26 dalek roast: review: https://github.com/perl6/roast/commit/8dedf60065
10:33 pyrimidine joined #perl6-dev
10:52 lizmat https://perl6advent.wordpress.com/2016/12/24/day-24-one-year-on/  # last post
11:09 samcv don't say last :(
11:11 lizmat of this year  :-)
11:11 samcv perl 6 advent for 100 years
11:22 AlexDaniel joined #perl6-dev
11:39 pyrimidine joined #perl6-dev
11:43 timotimo yew please
11:43 timotimo yes*
11:53 FROGGS joined #perl6-dev
12:41 pyrimidine joined #perl6-dev
13:00 MasterDukeLaptop joined #perl6-dev
13:08 notviki samcv: I've no opinion on that travis change. FWIW, gcc version on my box is gcc version 4.8.4 ....
13:10 samcv we could always keep the gcc at 4 though. but bumping to ubuntu 14.04 would be more in line with what people have, so hmm
13:10 samcv yeah gcc shouldn't matter as much imo
13:10 samcv but the other things on the system will be 2+?? years newer? not sure
13:11 notviki but I can argue either way: I'm not testing on gcc 6, but travis would :)
13:13 notviki What does that change from "language perl" to "language" java" give?
13:13 * notviki is opposed to that change on emotional level :)
13:15 samcv it makes sure java is all setup properly
13:15 samcv before we had it for a newer perl version. but with 14.04 we don't need that
13:15 samcv and this saves us from jvm stuff.
13:15 samcv without that it didn't have some of the java things needed to build in the path even if installed
13:17 samcv 14.04 was released
13:17 samcv April 17, 2014 April 26, 2012 for precise
13:17 samcv so 2 years newer
13:17 samcv hehe notviki
13:18 notviki lizmat: FWIW, there's a typo in first sentence "thing that are"
13:20 lizmat oops, notviki++
13:21 lizmat notviki: can you live with my advent post ?
13:22 lizmat if not, you need to nudge ab6tract a bit mor  :-)
13:22 lizmat *more
13:23 lizmat afk for the rest of the day&
13:27 notviki lizmat++ good post
13:27 notviki lizmat: yeah, I can live. I just felt bad for ab6stract because it was I who told them they can delay their post :}
13:28 dalek rakudo/nom: f0aa65e | samcv++ | / (2 files):
13:28 dalek rakudo/nom: Fix bug in makeUNIPROP.p6 causing it not to generate 2 props
13:28 dalek rakudo/nom:
13:28 dalek rakudo/nom: I forgot to flush the code generator function at the end, so it
13:28 dalek rakudo/nom: ended up missing the last two properties:
13:28 dalek rakudo/nom: 'Full_Composition_Exclusion' and 'Canonical_Combining_Class'
13:28 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f0aa65e5eb
13:28 dalek rakudo/nom: 4ff317e | (Zoffix Znet)++ | / (2 files):
13:28 dalek rakudo/nom: Merge pull request #972 from samcv/uniprop-make-fix
13:28 dalek rakudo/nom:
13:28 dalek rakudo/nom: Fix bug in makeUNIPROP.p6 causing it not to generate 2 props
13:28 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4ff317e173
13:29 dalek rakudo/nom: 4b6f4e6 | samcv++ | .travis.yml:
13:29 dalek rakudo/nom: Travis CI: Use Ubuntu 14.04 instead of 12.04 and GCC 6 instead of 4
13:29 dalek rakudo/nom:
13:29 dalek rakudo/nom: Use a much newer Ubuntu version and use a much newer GCC.
13:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4b6f4e6411
13:29 dalek rakudo/nom: 175e3d1 | (Zoffix Znet)++ | .travis.yml:
13:29 dalek rakudo/nom: Merge pull request #970 from samcv/gcc-5
13:29 dalek rakudo/nom:
13:29 dalek rakudo/nom: Travis CI: build on Ubuntu 14.04 instead of 12.04 and use GCC 6
13:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/175e3d1bd5
13:32 samcv sweet :D
13:44 dalek roast: e8b513a | samcv++ | S15-unicode-information/uniprop.t:
13:44 dalek roast: uniprop: add tests for two more properties
13:44 dalek roast:
13:44 dalek roast: Full_Composition_Exclusion Changes_When_NFKC_Casefolded
13:44 dalek roast: review: https://github.com/perl6/roast/commit/e8b513a8f2
13:50 RabidGravy joined #perl6-dev
13:52 dalek roast: a529332 | (Zoffix Znet)++ | S32-str/parse-base.t:
13:52 dalek roast: Greatly simplify parse-base tests
13:52 dalek roast:
13:52 dalek roast: Do not duplicate tests for method and sub forms, instead use a loop
13:52 dalek roast: that loops over the sub and the method we get from .^lookup
13:52 dalek roast: review: https://github.com/perl6/roast/commit/a529332b74
14:24 Ven joined #perl6-dev
14:46 moritz samcv: if you send a signed Contributor License Agreement to TPF (http://www.perlfoundation.org/contributor_license_agreement), we can give you direct commit access to Rakudo
14:49 nine samcv: send the CLA via email. We need your awesome productivity :)
14:54 * notviki nods
15:20 pyrimidi_ joined #perl6-dev
15:51 pyrimidine joined #perl6-dev
15:56 pyrimidi_ joined #perl6-dev
16:28 dalek rakudo/nom: 3282813 | (Zoffix Znet)++ | src/core/Str.pm:
16:28 dalek rakudo/nom: Allow missing whole part in parse-base
16:28 dalek rakudo/nom:
16:28 dalek rakudo/nom: Omiting zero on fractional numbers (e.g. -.5) is a common notation in
16:28 dalek rakudo/nom: North America; makes sense to allow it.
16:28 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3282813e45
16:29 dalek roast: 897cc99 | (Zoffix Znet)++ | S32-str/parse-base.t:
16:29 dalek roast: Test parse-base can take a fractional without a whole part
16:29 dalek roast:
16:29 dalek roast: Rakudo commit: https://github.com/rakudo/rakudo/commit/3282813e45
16:29 dalek roast: review: https://github.com/perl6/roast/commit/897cc99452
17:30 Ven joined #perl6-dev
17:40 * japhb wonders how much his original parsing code has been changed over the years that that is even necessary ...
17:42 notviki ?
17:42 notviki That code didn't exist 2 months ago...
18:01 japhb notviki: My code (originally in Str.Numeric, now forming the guts of the largest val() variant) handled almost every way of describing numbers that Perl 6 could represent.  But importantly, it did it with a bunch of internal helper routines that handled subsets of the problem.  So rather than extending parse-base bit by bit, why not just expose what you need from those internal routines (in R::I, perhaps) and use that?
18:02 japhb (And by the way, it's not ego pushing this, it's "This is already a solved problem.")
18:08 notviki I don't forsee any further extensions to parse-base
18:08 japhb Famous last words ... from the bot that brought you support of unicode minus.  ;-)
18:09 notviki Last I looked at already-existing parsing code is was a tangled mess of overengineering and bugs
18:10 notviki m: :16(':36<SEE>').say
18:10 camelia rakudo-moar 328281: OUTPUT«36806␤»
18:10 notviki m: :16('-FF').say; # can't do a negative
18:10 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: base-16 number must begin with valid digits or '.' in ':16<⏏-FF>' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:11 notviki m: :2('0xE').say
18:11 camelia rakudo-moar 328281: OUTPUT«14␤»
18:12 japhb As I recall, in the original speculation, you are supposed to do the negative outside the :base() if you are using that form.  Str.Numeric itself should handle it.
18:12 japhb And as for the base mixing, I think that was a side effect of some refactor done over the years.  :-/
18:14 notviki m: say :16<0eff>
18:14 camelia rakudo-moar 328281: OUTPUT«3839␤»
18:14 notviki m: say :16<0bff>
18:14 camelia rakudo-moar 328281: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Cannot convert string to number: malformed base-2 number in '⏏ff' (indicated by ⏏)␤at <tmp>:1␤------> say :16<0bff>⏏<EOL>␤»
18:15 notviki m: say ':16<0b11>'.Numeric
18:15 camelia rakudo-moar 328281: OUTPUT«2833␤»
18:17 notviki The problem with all of these is they feel like they've been designed by imagining the user simply hardcoding the numbers.
18:18 notviki So if I want to ask the user for a number and a base that number is in, I then have to combine the two variables in some contorted form
18:18 japhb Str.Numeric's use case is converting Strs -- and by the time it gets to .Numeric, it may as well be hardcoded.
18:18 * notviki doesn't follow
18:19 japhb m: say ':16("SEE")'.Numeric
18:19 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: malformed ':16' style radix number, expecting '<' or '[' after the base in ':16⏏(\"SEE\")' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:19 notviki If the data is coming from an external source, why would it "may as well be hardcoded"?
18:19 japhb m: say ':16<SEE>'.Numeric
18:19 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: base-16 number must begin with valid digits or '.' in ':16<⏏SEE>' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:20 notviki I have to manually construct the ':16<SEE>' string and then deconstruct it again on errors to make sane errors
18:20 japhb m: my $input = 'SEE'; dd ":16<$input>".Numeric
18:20 camelia rakudo-moar 328281: OUTPUT«Failure.new(exception => X::Str::Numeric.new(source => ":16<SEE>", pos => 4, reason => "base-16 number must begin with valid digits or '.'"), backtrace => Backtrace.new)␤»
18:20 japhb Ah, I see what you're getting at.
18:21 notviki But your base is also hardcoded in that line.
18:21 notviki m: my $base = 16; my $input = 'SEE'; say ":$base<$input>".Numeric
18:21 camelia rakudo-moar 328281: OUTPUT«Type Int does not support associative indexing.␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:21 notviki m: my $base = 16; my $input = 'SEE'; say ":$base\<$input>".Numeric
18:21 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: base-16 number must begin with valid digits or '.' in ':16<⏏SEE>' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:22 notviki m: my $base = "16\b\b"; my $input = 'SEE'; say ":$base\<$input>".Numeric
18:22 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: malformed ':16' style radix number, expecting '<' or '[' after the base in ':16⏏\b\b<SEE>' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:22 notviki m: my $base = "\b16\b\b"; my $input = 'SEE'; say ":$base\<$input>".Numeric
18:22 camelia rakudo-moar 328281: OUTPUT«Cannot convert string to number: radix (in decimal) expected after ':' in ':⏏\b16\b\b<SEE>' (indicated by ⏏)␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
18:22 japhb So I see your point re: fixing up the exception.
18:23 notviki But you don't see my point that $input.parse-base($base) is more userfriendly than ":$base\<$input>".Numeric?
18:23 japhb Not entirely sure what point you're trying to make about the \b's though
18:23 notviki That the input base has significant affect of the error message.
18:24 notviki parse-base will only ever say that base looks wrong, but with the interpolation method you can get any confusing error.
18:24 japhb notviki: No, I understand the point you're making.  My point *originally* is that the guts of parse-base should not be a reimplementation of parts of Str.Numeric.  They should use the same code.
18:24 japhb (Or, I guess, val() nowadays.)
18:25 notviki but those parts aren't easily detachable. I see they close over a lot of previously computed values.
18:25 japhb That sounds different than the original, hmmm.
18:25 * japhb goes to look at the newest code
18:27 notviki all of them close over $eos
18:28 notviki Actually I don't see any part to take for parse-base. The actual code that handles radix numbers parses the :16<..> form. And it doesn't handle signs, so....
18:30 notviki and the failure mode is handled by the sub closing over $eos and $pos and twiddling it while working and then the code outside of the sub checks those variables.
18:31 japhb Sigh.  I concede the discussion.
18:32 notviki oh and it closes over $radix too
18:34 notviki I get your point about not repeating code, but currently parse-base is simple code that fits entirely on my screen and 70% of it is error reporting.
20:09 samcv moritz++ nine++ will do that!
20:10 samcv should I sign and scan?
20:10 notviki that's how I did mine
20:12 MasterDukeLaptop joined #perl6-dev
20:13 samcv ok will use this pdf here http://blob.perl.org/perlweb/foundation/legal/licenses/cla.pdf
20:34 MasterDukeLaptop if anybody is interested, here is a sorted list of the names (and their counts) find_symbol is called with during a rakudo compile
20:34 MasterDukeLaptop https://gist.github.com/MasterDuke17/a3b0018f427dc15c36d7f4c4b884946b
20:34 MasterDukeLaptop the top is Parameter, called 40919 times
20:35 notviki s: &infix:<??>
20:35 SourceBaby notviki, Something's wrong: ␤ERR: [31m===[0mSORRY![31m===[0m Error while compiling -e␤Undeclared routine:␤    infix:<??> used at line 6. Did you mean 'infix:<∖>', 'infix:<lt>', 'infix:<eq>', 'infix:<o>', 'infix:<~|>'?␤␤
20:36 notviki 1168 times... I guess it does that whenever it encouters a ternary? Tries to check for an infix first.
20:36 notviki m: sub infix:<??> ($,$) { say "hi" }; 42 ?? 55
20:36 camelia rakudo-moar 328281: OUTPUT«hi␤»
20:37 notviki m: sub infix:<??> ($,$) { say "hi" }; sub infix:<!!> ($,$) { say "hi" }; 42 ?? 55 !! 45
20:37 camelia rakudo-moar 328281: OUTPUT«hi␤hi␤»
20:37 notviki :o
20:37 notviki oh, it's the ?/: thing it cries about
20:37 notviki m: sub infix:<?> ($,$) { say "hi" }; sub infix:<:> ($,$) { say "hi" }; 42 ? 55 : 45
20:37 camelia rakudo-moar 328281: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unsupported use of ? and : for the ternary conditional operator; in Perl 6 please use ?? and !!␤at <tmp>:1␤------> ; sub infix:<:> ($,$) { say "hi" }; 42 ?⏏ 55 : 45␤»
20:39 MasterDukeLaptop btw, that's why i was trying to figure out the caller, but callframe(1).code.name doesn't work that early in the compile
21:12 b2gills m: my str @a = <a b c>; @a.sort.say; # lizmat I think you broke this with one of your optimizations
21:12 camelia rakudo-moar 328281: OUTPUT«compare requires a concrete string, but got null␤  in block <unit> at <tmp> line 1␤␤»
21:14 MasterDukeLaptop bisectable6: my str @a = <a b c>; @a.sort.say;
21:14 bisectable6 MasterDukeLaptop, Bisecting by output (old=2015.12 new=3282813) because on both starting points the exit code is 1
21:15 bisectable6 MasterDukeLaptop, bisect log: https://gist.github.com/2a5521e52c37ce568525a34ee2f750af
21:15 bisectable6 MasterDukeLaptop, (2016-03-17) https://github.com/rakudo/rakudo/commit/eed4ebb0c0f5c7169c9f4536f4bd13f405d8f3cd
21:17 b2gills star-m: my str @a = <a b c>; @a.sort.say; say $*PERL.compiler.version
21:17 camelia star-m 2016.10: OUTPUT«(a b c)␤v2016.10␤»
21:22 MasterDukeLaptop committable6: all my str @a = <a b c>; @a.sort.say
21:22 committable6 MasterDukeLaptop, https://gist.github.com/385390d5777568711d011fa69c01aadc
21:23 MasterDukeLaptop bisectable6: old=2016.12 my str @a = <a b c>; @a.sort.say;
21:23 bisectable6 MasterDukeLaptop, Bisecting by exit code (old=2016.12 new=3282813). Old exit code: 0
21:24 notviki committable6: 4c2c9c2 my str @a = <a b c>; @a.sort.say;
21:24 committable6 notviki, ¦«4c2c9c2»: ===SORRY!===␤native string arrays not yet implemented. Sorry.  «exit code = 1»
21:24 bisectable6 MasterDukeLaptop, bisect log: https://gist.github.com/c702aeeeff31f3ffa66f998d3546af9c
21:24 bisectable6 MasterDukeLaptop, (2016-12-23) https://github.com/rakudo/rakudo/commit/4b2cea078f595613ff0f89a9bf1462d4a4295fb9
21:24 samcv moritz, i have a PDF i scanned. What email to send the CLA to?
21:29 samcv or nine
21:43 b2gills samcv: http://www.perlfoundation.org/contact_us has a trademark(at)perlfoundation.org mail address listed
21:45 samcv thanks
21:57 btyler joined #perl6-dev
21:59 samcv awesome. CLA sent
22:02 notviki samcv++
22:02 samcv :) glad to be here!
22:02 samcv can I get a fancy irc hostmask too :3
22:03 dalek rakudo/nom: 350fdc1 | (Zoffix Znet)++ | src/core/Main.pm:
22:03 dalek rakudo/nom: Remove trailing whitespace
22:03 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/350fdc126b
22:03 notviki I think if you join #freenode and ask they'll give one
22:04 samcv do the people here have to tell them yes?
22:04 notviki no idea, I'm unaffiliated :)
22:04 * notviki summons mst
22:05 notviki hm, my Matt Signal isn't working today :)
22:05 notviki I guess they'll tell you what is needed on #freenode
22:13 samcv joined #perl6-dev
22:14 samcv ok back
22:14 samcv hope nobody said anything important
22:14 samcv joined #perl6-dev
22:14 samcv agh
22:15 samcv well that solved my lag issues by restarting ZNC
22:21 dalek rakudo/nom: 38ec79c | (Zoffix Znet)++ | src/core/Main.pm:
22:21 dalek rakudo/nom: Make sub MAIN usage ignore *% param
22:21 dalek rakudo/nom:
22:21 dalek rakudo/nom: The code assumes a param always has a name and when it tries to get
22:21 dalek rakudo/nom: the name of *% param, it crashes due to attempting to treat a null as
22:21 dalek rakudo/nom: a string.
22:21 dalek rakudo/nom:
22:21 dalek rakudo/nom: Fix by foregoing any nameless named params when printing usage.
22:21 dalek rakudo/nom:
22:21 dalek rakudo/nom: Fixes RT#127621: https://rt.perl.org/Ticket/Display.html?id=127621
22:21 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/38ec79c623
22:21 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=127621
22:22 dalek roast: a26b352 | (Zoffix Znet)++ | S06-other/main-usage.t:
22:22 dalek roast: Test *% in sub MAIN does not cause a crash
22:22 dalek roast:
22:22 dalek roast: RT#127621: https://rt.perl.org/Ticket/Display.html?id=127621
22:22 dalek roast: Rakudo fix: https://github.com/rakudo/rakudo/commit/38ec79c623
22:22 dalek roast: review: https://github.com/perl6/roast/commit/a26b352434
22:22 synopsebot6 Link:  https://rt.perl.org/rt3//Public/Bug/Display.html?id=127621
22:59 dalek roast: 1c498de | samcv++ | S15-unicode-information/uniprop.t:
22:59 dalek roast: uniprop: add tests for the last three NF*_Quick_Check props
22:59 dalek roast:
22:59 dalek roast: NFD_Quick_Check NFKC_Quick_Check NFKD_Quick_Check
22:59 dalek roast: We now cover all properties in the 'Enumerated Values' category.
22:59 dalek roast: review: https://github.com/perl6/roast/commit/1c498de88f
23:18 dalek roast: ed3f324 | samcv++ | S15-unicode-information/uniprop.t:
23:18 dalek roast: uniprop: add tests for 9 more properties
23:18 dalek roast:
23:18 dalek roast: Cased, Changes_When_Casefolded, Changes_When_Casemapped,
23:18 dalek roast: Changes_When_Lowercased, Changes_When_Uppercased, Changes_When_Titlecased, Uppercase,
23:18 dalek roast:  Terminal_Punctuation, Bidi_Control
23:18 dalek roast: review: https://github.com/perl6/roast/commit/ed3f324210
23:35 notviki t/spec/S15-unicode-information/uniprop.t is failing
23:35 notviki I think you've missed a plan bump?
23:36 notviki "Parse errors: Bad plan.  You planned 143 tests but ran 145."

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary