[11:00:35] hi [11:01:15] hey [11:01:49] any ariya? [11:19:08] hello folks [11:19:10] sorry I'm late [11:20:20] anyone is still here? [11:21:30] no worries [11:21:31] hello [11:21:44] hi jlast [11:22:01] will be back in a minute [11:23:32] jlast: your https://github.com/jquery/esprima/pull/1248 looks really good! [11:23:42] let me see if I can spend some time looking at the test failures [11:25:34] JSCS reports coding style errors, but it should be easy for you to fix them [11:25:55] and make sure to rebase off master, to exclude failing downstream JSCS tests [11:29:17] thanks ariya - just got back [11:29:41] i actually hunted down the test errors this morning [11:30:12] (most of them were from not eval'ing `var source = "foo"` [11:30:58] looks like it is easy to fix :-) [11:31:15] we could also skip it for the browser tests [11:31:28] yup, they're all fixed. one of them was source-element which was empty and the other was the invalid-yield [11:31:28] I'm not sure those tests are highly critical (for the browser-based tests) [11:31:39] sure [11:32:29] I'm dropping some notes in the PR so that it's more digestable. To be clear, I'll do a code quality pass before i consider it ready per se... [11:33:42] sounds good [11:34:08] on a general topic, I got the final PR to close our ES6 support. yay! [11:34:25] once it's reviewed and landed, we can say that we cover ES6 more or less (notwithstanding defects) [11:34:30] wow [11:34:44] by the way, it's crazy how many more tests es6 adds [11:34:50] I have prepared Release 2.5 umbrella ticket: https://github.com/jquery/esprima/issues/1246 [11:34:51] and i'm sure it's true for complexity as well. [11:35:13] argh; I switched networks and didn't realize I was disconnected [11:35:17] what'd I miss? [11:35:32] gibson042: we just started :-) [11:35:40] gibson042: and I got your email, ready to write something soon [11:35:48] thanks for that reminder [11:36:02] np [11:37:07] jlast: let me know when you feel confident about your PR so that we can land it asap [11:37:23] will do. [11:37:33] of course, assuming that your karma runner doesn't reveal any browser-specific issues :-) [11:37:38] also, do you have any opinions on the parseFunction work? [11:37:51] jlast: yeah, that one is bit tough after the yield complication [11:37:53] haha [11:38:06] what do you mean? [11:38:12] is the yield work still pending? [11:38:17] jlast: it's complete [11:38:28] jlast: but the idea unifying both happened before that [11:38:40] and the two functions looked extremely identical back then [11:39:24] I don't have any strong opinion, let me think about this a bit more [11:39:36] more likely it's OK to refactor it as you propose in that PR [11:40:52] alright - happy to try a different approach too. also, i meant to ask if there's another ticket you think i could look at [11:41:24] let me check [11:41:48] https://github.com/jquery/esprima/issues/1159, if mike is busy and can't take that [11:41:53] https://github.com/jquery/esprima/issues/1215 [11:42:08] coveralls.io is flaky, hence the experiment with other coverage service [11:42:18] #1159 would be nice w/ a little more context [11:42:33] https://github.com/jquery/esprima/issues/1048 if you want [11:42:52] it will be fun since you need to walk through the grammar, feel free to poke me if you need some hints [11:43:02] coverage is interesting, but i've got a decent amount of tools work on the horizon :P [11:43:13] https://github.com/jquery/esprima/issues/1006 if you fancy extending the ast node [11:43:25] var let looks like a lot of fun [11:43:31] for this, you may need to complete https://github.com/estree/estree/issues/6 as well [11:43:37] like `var let = 2;` should work? [11:43:54] see also https://github.com/dherman/estree/commit/70ba4ab1fd8de48ca9dfd94a79820d3fe65d4009 [11:44:09] jlast: only in non strict mode, yes [11:44:22] yes, ofcourse [11:45:00] https://github.com/jquery/esprima/issues/1164 is good if you want to understand fancy feature [11:45:01] Alright, i'll see what I can do with these. I'll try and use the issues to answer some questions [11:45:06] probably you can start by writing test cases [11:45:21] we have over 1200 tests already, hopefully we'll reach 1337 at some point and stop there! [11:45:35] yea, that's what i was thinking. haha at that point every test you add you have to remove one [11:45:51] ^ test cases and pseudo expected output [11:46:21] jlast: I think those tickets should keep you busy for a while :-) [11:46:32] by the way, i wrote two "intro" blog posts on esprima internals as i was starting http://jasonlaster.github.io/ [11:46:58] my style is to keep a journal as i'm going, and i don't spend too much time on post-production value... so take it w/ a grain of salt [11:47:30] someday, it will be useful to fold the content into esprima.org [11:47:37] I also plan to add some videos there [11:47:38] I looked for some other "esprima getting started material" and couldn't find any. Are there other good resources? [11:48:18] there is [11:48:23] not specific to esprima, though [11:48:26] let me check [11:49:41] https://www.youtube.com/watch?v=XA8_hZfVecI [11:49:55] would be nice to collect those videos in one place [11:50:27] https://www.youtube.com/watch?v=UqTlToUYK1E is another good one [11:50:58] yea, you and mike and one recently too [11:51:32] that's a bit more practical :-) [11:51:39] one day I also like to explain parseBinaryExpression [11:52:02] the rest of the parser is recursive descent, but that one is stack based, for performance and to reduce recursion [11:52:50] in all cases, I need to leave soon [11:53:05] jlast: glad to have you helping us with this! [11:53:14] don't hesitate to open issues/PR and discuss more stuff [11:53:18] nice talking. Happy to. [11:53:30] thanks for all the help so far [11:53:42] thanks for your participation! [11:53:45] see you next week folks [11:53:47] * ariya waves