This version (2017/05/27 13:44) is a draft.
Approvals: 0/1

[06:05:41] <Anil> Hello - i am trying to build a rest api with data source interaction using vertx [06:06:00] <Anil> can some one help me in validating the approach ?

[07:40:04] * ChanServ sets mode: +o temporalfox [07:45:57] * ChanServ sets mode: +o temporalfox

[09:03:59] * ChanServ sets mode: +o temporalfox [10:44:46] * ChanServ sets mode: +o temporalfox

[11:11:16] * ChanServ sets mode: +o temporalfox [14:59:22] * ChanServ sets mode: +o temporalfox

[17:44:38] <aesteve> hello everyone ! :)

[17:45:36] <aesteve> temporalfox: I thought about your comment about @DataObject the other day, and I'm wondering if Groovy's AST transformations couldn't help

[17:45:53] <temporalfox> which comment ?

[17:46:12] <temporalfox> as DataObject ?

[17:46:20] <aesteve> yes that kind of stuff

[17:46:22] <temporalfox> I'm not a fan of AST transformation

[17:46:29] <temporalfox> although I find that very powerful

[17:46:46] <temporalfox> I think something like vertx sync could be done using AST transformation

[17:46:54] <temporalfox> by rewriting AST

[17:47:07] <aesteve> but it clinged to my mind, that by default in a Groovy environment, @DataObject could generate all the needed stuff (fromJson, toJson) thanks to an AST transformation

[17:48:02] <temporalfox> we do that using AST

[17:48:05] <temporalfox> it generates a Converter class

[17:48:12] <temporalfox> because such code is quite error prone

[17:48:23] <aesteve> oh that's what you were telling me too

[17:48:37] <aesteve> the “generates the boilerplate code” thing

[17:48:44] <aesteve> I'll have a look

[17:49:19] <temporalfox> for instance

[17:49:19] <temporalfox>

[17:49:28] <temporalfox> is generated from data object

[17:50:15] <aesteve> ok that's basically what I had in mind

[17:50:18] <aesteve> :)

[17:51:05] <aesteve> now I'm still thinking about what I could add to ease web development in groovy

[17:51:41] <temporalfox> support for groovy templates

[17:51:52] <aesteve> very good point

[17:52:03] <aesteve> but my PR is bugged and I don't know why

[17:53:48] <aesteve>

[17:54:30] <aesteve> there are problems in “enums.adoc” for example

[18:00:48] <temporalfox> ah I did not see you did a PR

[18:01:03] <temporalfox> why didn't we merge it ?

[18:08:09] <aesteve> i didn't send the PR yet

[18:08:21] <aesteve> the link is just the PR “preparation”

[18:08:46] <aesteve> but if you look carefully, there are modifications about enums.adoc I have no clue where they come from

[18:09:18] <aesteve> and second problem, the “create” method isn't codegen friendly

[18:14:53] <temporalfox> it is hard to see this problem from here

[18:15:19] <aesteve> yeah obviously :\

[18:15:50] <aesteve> for the enums I have really no clue, but for the codegen stuff, maybe you can give me an advice

[18:16:08] <aesteve> the problem is that there are different template engines in Groovy

[18:16:19] <aesteve> Markup, Simple, …

[18:17:18] <temporalfox> why are you using codegen enums with groovy ?

[18:17:24] <temporalfox> can I try to build it ?

[18:17:36] <aesteve> yes !

[18:17:46] <aesteve> that must be a build problem

[18:17:55] <aesteve> but I can't figure out what I did wrong

[18:18:02] <temporalfox> ok

[18:18:02] <temporalfox> I'm going to try now

[18:18:22] <temporalfox> should I clone this ?

[18:18:28] <aesteve> idd

[18:19:34] <temporalfox> and the branch groovy-template-engines2

[18:19:58] <aesteve> it's this branch, idd.

[18:21:59] <aesteve> you'll notice the io.vertx.ext.web.templ.GroovyTemplateEngine.create() is @GenIgnore'd

[18:22:00] <aesteve>

[18:22:18] <aesteve> and it's more a design, than a VertxGen question

[18:22:31] <temporalfox> so it builds

[18:22:46] <temporalfox> I see this

[18:22:49] <temporalfox> yes

[18:22:53] <temporalfox> so what is the problem ?

[18:23:12] <temporalfox> can't you have a

[18:23:12] <temporalfox> static HandlebarsTemplateEngine create()

[18:23:20] <aesteve> actually not

[18:23:23] <temporalfox> that does not use a groovy.text.TemplateEngine ?

[18:23:40] <aesteve> because it can be a GStringTemplateEngine, MarkupTemplateEngine

[18:23:53] <aesteve> and they're really different

[18:24:02] <temporalfox> so you need to make a

[18:24:13] <temporalfox> createGStringTemplateEngine and a createMarkupTemplateEngine static

[18:24:41] <temporalfox> which one is the latest from C[unknown:eacute]dric Champeau ?

[18:24:52] <temporalfox> it's MarkupTemplateEngine

[18:24:53] <temporalfox> I think

[18:25:20] <aesteve> idd

[18:26:06] <temporalfox> static GroovyTemplateEngine createMarkupTemplateEngine() {

[18:26:10] <temporalfox> return new GroovyTemplateEngineImpl(new MarkupTemplateEngine());

[18:26:33] <aesteve> I did that at some point

[18:26:44] <aesteve> but I wasn't happy with it

[18:26:48] <temporalfox> so what is the problem ?

[18:27:00] <aesteve>,%20groovy.text.markup.TemplateConfiguration,%20groovy.text.markup.TemplateResolver)

[18:27:13] <aesteve> there's templateconfiguration stuff for example

[18:27:31] <temporalfox> you should make the config

[18:27:36] <temporalfox> with a data object

[18:27:41] <aesteve> ok

[18:27:42] <temporalfox> like a MarkupConfig

[18:27:46] <aesteve> didn't think about this

[18:28:19] <temporalfox> you could also maybe

[18:28:20] <temporalfox> do a TemplateResolver

[18:28:25] <temporalfox> using vertx file system

[18:28:34] <aesteve> I'll have to

[18:28:49] <temporalfox> I can try to do these to see how it works

[18:28:53] <aesteve> the most painful stuff will be the tests

[18:29:24] <aesteve> actually, I'm not even sure Handlebars templateresolver is tested but it should (it's the same thing)

[18:29:35] <aesteve> inclusion / composition

[18:32:22] <aesteve> what do you wanna try ? templateresolver or markupconfig ?

[18:33:13] <temporalfox> markupconfig

[18:33:14] <temporalfox> first

[18:33:23] <temporalfox> does vertx uses locale stuff somewhere ?

[18:33:39] <temporalfox> I'm seing a Locale and don't know how it is handled in general

[18:33:55] <aesteve>

[18:34:24] <aesteve> something very useful from vertx-web

[18:35:02] <temporalfox> ok there is own locale class

[20:33:24] <temporalfox> I don't understand how the “somedir” works in tests

[20:33:43] <aesteve> let me see

[20:35:46] <aesteve>

[20:35:57] <aesteve> it's within the 'root' vertx-web project

[20:36:30] <aesteve> iirc it's for testing that templates can be loaded either from the classpath or the fs

[20:36:52] <aesteve> but tbch I tried to mimic what was done with handlebars

[21:29:12] <temporalfox> aesteve these tests don't pass for existing classes like XMLTemplateTest

[21:29:21] <temporalfox> Caused by: java.nio.file.NoSuchFileException: somedir/test-xml-template2.gtpl

[21:29:21] <temporalfox> at sun.nio.fs.UnixException.translateToIOException(

[21:29:21] <temporalfox> at sun.nio.fs.UnixException.rethrowAsIOException(

[21:29:21] <temporalfox> at sun.nio.fs.UnixException.rethrowAsIOException(

[21:29:23] <temporalfox> at sun.nio.fs.UnixFileSystemProvider.newByteChannel(

[21:29:28] <temporalfox> is it expected ?

[21:29:52] <aesteve> no, and they pass on my machine

[21:30:19] <aesteve> are the handlebars test broken, too ?

[21:33:21] <temporalfox> I haven't tried them

[21:35:16] <aesteve> mmh ok

[21:44:29] <temporalfox> perhaps it is because I'm trying it in the vertx-web-templ-groovy folder itself

[21:44:50] <aesteve> maybe

[21:45:09] <aesteve> i was about to ask if you were running them from maven or your IDE

[21:46:13] <aesteve> I'm abit lost since the project has been refactored into small pieces (template engines are now subprojects)

[21:46:49] <aesteve> actually I had to install IntelliJ since Eclipse doesn't “like” subprojects

[22:05:20] <aesteve> actually, the project doesn't compile anymore on my computer

[22:05:41] <aesteve> [ERROR] /Users/arnaud/git/vertx-web/vertx-web/src/main/java/io/vertx/ext/web/handler/sockjs/impl/[62,1] io.vertx.ext.web.handler.sockjs.impl.SockJSSession is not abstract and does not override abstract method end() in io.vertx.core.streams.WriteStream

[22:06:49] <temporalfox> yes

[22:07:01] <temporalfox> you need to rebase on latest master

[22:07:05] <temporalfox> I run from maven

[22:17:29] <aesteve> mvn clean test

[22:17:36] <aesteve> just gives me 1 test failure : SockJSProtocolTest.testProtocol:77→AsyncTestBase.assertEquals:431 Protocol tests failed expected:<0> but was:<2>

[22:17:46] <aesteve> but templates are OK

[22:19:33] <aesteve> got the same error within the vertx-template-engines dir

[22:19:37] <aesteve> … :(

[22:25:14] <aesteve> ok so it's the opposite from what I remembered, they're running from the IDE, but not from Maven

[22:25:18] <aesteve> I'll have a look

[22:25:24] <aesteve> god it's so painful

[22:35:59] <temporal_> :-)

[22:36:15] <temporal_> let's say what CI says

[22:36:21] <temporal_> let' see

[22:36:45] <temporal_> CI seems happy

[22:36:51] <aesteve> actually, the problem is I'm completely lost, I use to work with Gradle, Eclipse, and flat projects

[22:36:55] <temporal_> have you changed something that would make this test fail ?

[22:36:57] <temporal_> ok

[22:37:09] <aesteve> and I have to use Maven, Nested projects and IntelliJ

[22:37:44] <aesteve> atm it does no longer compile on my computer I have no clue why, “TemplateEngine” cannot be found

[22:38:13] <aesteve> just like if groovy-temp-engine wasn't a subproject of vertx-web

[22:38:24] <aesteve> that makes no sense to me :D

[22:39:04] <aesteve> I think the problem lies in maven, and maybe <artifactSet>

[22:39:04] <aesteve> <includes>

[22:39:04] <aesteve> <include>*:groovy</include>

[22:39:04] <aesteve> <include>org.apache.commons:commons-lang3</include>

[22:39:04] <aesteve> <include>org.antlr:*</include>

[22:39:04] <aesteve> <!– run in no-rhino mode (do not include rhino) –>

[22:39:04] <aesteve> </includes>

[22:40:48] <aesteve> since it's the shade plugin config, isn't it used to package resources when it runs the tests ?

[22:40:51] <aesteve> no idea ^^

[22:54:15] <aesteve> doesn't work either. No idea what I should do. It's the exact same thing as handlebars, idk why maven doesn't find the 'somedir' directory

[22:56:00] <temporal_> does handlebar sees it ?

[22:56:21] <aesteve> the handlebars test don't fail, they work just fine

[22:56:25] <aesteve> maven or IDE, both

[22:56:28] <temporal_> seems ok for me

[22:56:40] <aesteve> but mines are failing

[22:56:52] <aesteve> only with maven, IntelliJ has no problem

[22:56:53] <temporal_> maven shade plugin is only for packaging

[22:56:56] <temporal_> not testing

[22:57:07] <aesteve> yes and the test-jar dependency is in the POM

[22:59:24] <temporal_> yes

[22:59:28] <temporal_> let's find out!

[22:59:46] <temporal_> I think it's the resolver thing

[22:59:52] <temporal_> that does not load template from classpath

[23:00:03] <temporal_> has it ever worked ?

[23:01:31] <aesteve> it used to work in the very first version before the project was split into submodules

[23:01:38] <temporal_> ok

[23:02:55] <aesteve> so : return new GStringTemplateEngine(Utils.getClassLoader());

[23:03:12] <aesteve> mmh no actually

[23:03:27] <aesteve> the file is sent to TemplateEngine by TemplateHandler iirc

[23:04:03] <aesteve> the resolver will only be used by the template engine if another template is included (i think)

[23:04:15] <temporal_> ah in my case it fails

[23:04:15] <temporal_> because I changed some logic in tes tinit

[23:04:24] <temporal_> so perhaps we don't see same failure

[23:05:15] <temporal_> now it passes

[23:05:21] <temporal_> ah no

[23:05:25] <temporal_> I mean yes

[23:05:27] <temporal_> but one does not

[23:05:31] <temporal_> testTemplateOnClasspath

[23:05:34] <temporal_> like you

[23:05:52] <temporal_> actually it is

[23:05:53] <temporal_> testGetGroovyEngine

[23:06:23] <temporal_> it's because I changed the test

[23:06:42] <temporal_> because the create() I'm doing does not use a template engine as argument

[23:06:49] <temporal_> so the test doe snot init it anymore

[23:06:50] <aesteve> yes that makes sense

[23:07:15] <aesteve> but the classpath thing works ??

[23:07:23] <temporal_> it seems to

[23:07:31] <temporal_> for XmlTemplateTest at least

[23:07:54] <temporal_> let me push that somewhere once it works fully

[23:08:04] <aesteve> I'd be glad to

[23:08:34] <aesteve> because the line failing is : String rawTpl = Utils.readFileToString(vertx, loc);

[23:10:00] <temporal_> try that

[23:10:01] <temporal_>

[23:10:15] <temporal_> it works from cmd line

[23:13:09] <aesteve> same issue

[23:13:23] <aesteve> Caused by: io.vertx.core.file.FileSystemException: java.nio.file.NoSuchFileException: somedir/test-gstring-template2.gtpl

[23:13:35] <temporal_> are you running from CLI ?

[23:13:42] <temporal_> in which directory are you ?

[23:13:49] <temporal_> I'm running from the groovy dir

[23:14:35] <aesteve> same

[23:14:51] <aesteve> macbook-air-de-arnaud:vertx-template-engines arnaud$ pwd

[23:15:36] <temporal_> juliens-mbp-2:vertx-web-templ-groovy julien$ pwd

[23:15:40] <aesteve> ## /tmp/vertx-web/vertx-template-engines

[23:15:49] <temporal_> “/Users/julien/java/vertx-web-aesteve/vertx-template-engines/vertx-web-templ-groovy”

[23:15:49] <aesteve> irc doesnt like '/'

[23:17:03] <aesteve> same within the groovy dir

[23:17:08] <aesteve> same OS, same project

[23:17:24] <temporal_> have you tried from my fork ?

[23:17:31] <aesteve> yes the /tmp

[23:17:34] <aesteve> is your fork

[23:19:10] *** ChanServ sets mode: +o temporalfox

[23:21:10] <aesteve> yes that's your fork, on the groovy-template-engines2 branch

[23:21:39] <aesteve> I can see the MarkupConfig class

[23:23:25] <aesteve> the important thing is that it works for you and the CI

[23:23:27] <aesteve> :D

[23:27:49] <temporalfox> it should work for you

[23:28:44] <aesteve> what's really really weird

[23:29:01] <aesteve> is that IntelliJ has absolutely no issue running the tests

[23:29:08] <aesteve> but maven does

[23:30:53] <temporalfox> what is your maven version ?

[23:31:08] <temporalfox> I'm using v

[23:31:09] <temporalfox> 3.2.5

[23:31:39] <aesteve> 3.3.3

[23:33:07] <aesteve> but even if it was a version issue, handlebars or thymeleaf would fail, too

[23:34:54] <aesteve> I _must_ have done something wrong :) but what ? that's crazy