Perl 6 - the future is here, just unevenly distributed

IRC log for #metacpan, 2014-12-12

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

All times shown according to UTC.

Time Nick Message
00:41 neilb_ joined #metacpan
01:00 vroom joined #metacpan
01:38 vroom joined #metacpan
01:52 vroom joined #metacpan
02:15 klapperl joined #metacpan
03:27 vroom joined #metacpan
03:37 vroom joined #metacpan
09:08 punter joined #metacpan
09:11 neilb joined #metacpan
10:01 neilb_ joined #metacpan
10:30 ChoiKyuSang joined #metacpan
10:59 neilb joined #metacpan
12:40 vroom joined #metacpan
12:49 neilb joined #metacpan
14:18 neilb joined #metacpan
14:47 cjm joined #metacpan
16:03 Ptolemarch joined #metacpan
16:14 Ptolemarch joined #metacpan
16:48 oiami joined #metacpan
17:18 Ptolemarch joined #metacpan
17:33 neilb joined #metacpan
17:38 neilb_ joined #metacpan
18:35 chmrr joined #metacpan
19:05 chmrr We've started sticking additional metadata into RT extension meta.yml files (https://metacpan.org/source/ALEXMV/RT-Extension-SLA-1.03/META.yml#L31 ) -- what's the right way to run a metacpan query over that data?
19:05 dipsy [ META.yml - metacpan.org ]
19:06 chmrr I'm not terribly experienced in elasticsearch, so clearly I'm doing something wrong here: https://chmrr.net/nopaste/2014-12-12D06i415D (sent to /v0/release )
19:22 Ptolemarch joined #metacpan
19:32 trs chmrr: unfortunately the metadata isn't indexed/analyzed by ES
19:32 trs so it's not searchable.
19:32 trs sorry to rain on your parade.
19:33 chmrr Bleh.  Is there any alternate place to shove data that _is_ indexed?
19:33 mst I'm not sure I understand why you can't put the dependency in normally?
19:34 chmrr RT.pm isn't in @INC, and we don't want cpan to try to install it.
19:35 trs chmrr: the ES "mapping" for an index type shows you what fields are indexed/analyzed and
19:35 trs http://api.metacpan.org/v0/release/_mapping
19:36 trs I don't think there's anywhere else sensical you could shove it.
19:36 trs other than the normal prereq list
19:36 chmrr Is "x_tyoe" in that a typo?
19:37 trs seems like it!
19:39 trs though I'm not sure how it got there.
19:39 trs ^ ranguard ?
19:42 trs chmrr: basically custom metadata stuff is unlikely to ever be indexed by ES, since it's unbounded.
19:43 trs this is yet another reason to wish metacpan used Pg.
20:00 chmrr Hm.  Well, I guess we can use metacpan to get the metadata, and iterate through it ourselves.
20:15 vroom joined #metacpan
20:16 ranguard trs: ^ I just sys admin here, not got a clue about ES really (just learning atm) :)
20:23 trs ranguard: damn, sorry, I guess I meant rwstauner ?
20:25 trs chmrr: perhaps you could pull a cute trick: put a real RT dep in META.* but omit it when Makefile.PL generates MYMETA.*
20:26 oalders trs: Pg is actually possibly on the horizon
20:26 chmrr Hm -- possibly.  Except that doesn't suffice for putting a maximum version
20:26 trs \o/
20:26 chmrr Which data is going to end up in pg?
20:26 trs chmrr: there is x_breaks, I think? which ether advocates?
20:27 oalders it's on riveter's to-do list, if it makes sense for her work
20:28 ranguard trs: probably :)
20:29 trs chmrr: https://metacpan.org/pod/CPAN::Meta::Spec#Version-Ranges ?
20:29 dipsy [ CPAN::Meta::Spec - specification for CPAN distribution metadata - metacpan.org ]
20:29 oalders chmrr: probably most of the data we currently have (except files) would end up in Pg
20:31 trs chmrr: or maybe instead of metacpan, use grep.cpan.me? cpangrep file:META.yml '(x_requires_rt|x_rt_too_new):'
20:32 vroom joined #metacpan
20:34 chmrr Mmm.  I do worry that the x.y.z version format will complicate the normal dependency codepaths.  But the version ranges stuff is interesting
20:35 trs I believe they just expand like perl's
20:36 chmrr They apparently have to have a leading v on them: https://metacpan.org/pod/CPAN::Meta::Spec#Dotted-integer-versions
20:36 dipsy [ CPAN::Meta::Spec - specification for CPAN distribution metadata - metacpan.org ]
20:37 trs a
20:37 trs ah
20:38 trs I imagine 4.2.10rc1 might cause havoc too, though unsure.
20:38 chmrr Yup.  Mmmm, vstrings.  Always such fun.
20:39 mst oalders: if you start doing Pg, please please let me get involved in the schema design
20:39 trs but it could be translated to v4.2.10_01
20:39 mst I've wanted that since before metacpan.org first went live
20:39 oalders mst: i'll take you up on that for sure
20:40 mst \o/
20:50 vroom joined #metacpan
20:51 oiami left #metacpan
20:58 punter joined #metacpan
21:10 vroom joined #metacpan
21:18 chmrr Is it possible to query for releases containing a particular file, and get back the release information, not the file's information?
21:19 chmrr The "releases that include a cpanfile" example is most of the way there, but returns files, not releases, so it's an additional query for each result to get the release metadata.
21:20 oalders chmrr: you're looking at https://github.com/CPAN-API/metacpan-examples/blob/master/scripts/endpoints/file/2-get-dists-with-cpanfile.pl ?
21:20 dipsy [ metacpan-examples/2-get-dists-with-cpanfile.pl at master · CPAN-API/metacpan-examples · GitHub ]
21:20 chmrr I was looking at the link off of explorer.metacpan.org
21:20 chmrr But it looks ~equivalent to that
21:22 oalders chmrr: https://github.com/CPAN-API/cpan-api/blob/master/docs/API-docs.md#joins
21:22 dipsy [ cpan-api/API-docs.md at master · CPAN-API/cpan-api · GitHub ]
21:22 oalders it looks like you might be able to try joining
21:22 chmrr Yeah.  But the joins looks like for imposing limits on the joined table, not getting results from it.
21:22 chmrr And I don't see a relationship from Release -> File on https://github.com/CPAN-API/cpan-api/blob/master/lib/MetaCPAN/Server/Controller/Release.pm but I was hoping I was just wrong
21:22 dipsy [ cpan-api/Release.pm at master · CPAN-API/cpan-api · GitHub ]
21:23 chmrr There is one the other way, File -> Release, https://github.com/CPAN-API/cpan-api/blob/master/lib/MetaCPAN/Server/Controller/File.pm but I don't know that that allows me to get release data in the resultset
21:23 dipsy [ cpan-api/File.pm at master · CPAN-API/cpan-api · GitHub ]
21:23 oalders i've actually never messed with the joins, so i can't say much about them
21:23 chmrr *nod*
21:24 oalders but making multiple requests to get what you actually want wouldn't be unusual
21:24 oalders you can use a caching useragent to speed things up
21:24 chmrr Makes sense.
21:49 vroom joined #metacpan
22:15 vroom joined #metacpan
22:41 vroom joined #metacpan
23:03 Ptolemarch joined #metacpan
23:30 punter joined #metacpan

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