Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • 14:41
    Travis dry-rb/dry-view (master) errored (638)
  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 14:36

    dry-bot on master

    [devtools] config sync (compare)

  • 11:48

    flash-gordon on v1.2.2

    (compare)

  • 11:47

    flash-gordon on master

    Update CHANGELOG (compare)

  • 11:36
    flash-gordon closed #376
  • 11:36

    flash-gordon on reject-string-keys

    (compare)

  • 11:36

    flash-gordon on master

    Reject non-symbol keys in schem… Test invalid exception message Merge pull request #378 from dr… (compare)

  • 11:36
    flash-gordon closed #378
  • 11:24
    flash-gordon synchronize #378
Tim Riley
@timriley
That’s why I was asking whether it was .Form schema, I was wondering if/how that worked with the new type specs
Ralf Schmitz Bongiolo
@mrbongiolo
@timriley so each accept a type as first argument?
I'm using Dry::Validation.Form with type_spec = true on the configs
I've tried your suggestions, but then I get this error: +nil+ is not a valid predicate name
Tim Riley
@timriley
ah, I probably got it wrong, sorry
Tim Riley
@timriley
It’s an interesting question though, the intersection of type specs and each
Ralf Schmitz Bongiolo
@mrbongiolo
Yes, I guess it makes sense to be able to coerce array values in the way that you expressed. Not sure if that would be doable or way to off of what each() proposes
Christian Meier
@mkristian
hi, I am looking for some way to share validation rules between server and javascript client. I wonder if dry-validation can export its 'rules' in one way or the other ?
Andy Holland
@AMHOL
@mkristian I've thought about this a lot in the past, the best solution I could think of was a custom verb like DESCRIBE to build the model representation on start-up, but I think just sticking to server-side validation is the better option
Having said that dry-validation builds an AST, so you could use that to create some representation of the validation predicates
Christian Meier
@mkristian
that sound interesting.
server side validation is easier but with javascript only clients not an option - anymore
Andy Holland
@AMHOL
Not sure what you mean, surely you send your data to the server at some point?
Christian Meier
@mkristian
yes, we do but client side validation is much more responsive. the AST idea I will look into it. as we already decided to have only 'simple' validation on the client and all those custom and more complex stuff stays on the server
Andy Holland
@AMHOL
Yeah, it would be nice in an ideal world, I just think the cost outweighs the benefits personally
Christian Meier
@mkristian
I personally are more on your side but it is not my decision I just need to find a solution :)
Fran Worley
@fran-worley
If you really don't want to write the js validations yourself you could just Ajax the data on change to the server and display any errors
Christian Meier
@mkristian
@fran-worley I want to have both and want to share the validation rules at some of them between the client and the server
Andy Holland
@AMHOL
Well if a client/my boss asked me to do that, my suggested solution would be let's leave that rabbit hole alone, have the user hit submit and display the errors in one hit, but I know that's not always an option :p
Christian Meier
@mkristian
well, I was hopping (not too much) that there is already some solution out there with dry-validation
but from hear I get the courage and question the client side validation approach tomorrow - let's see what response I get
Andy Holland
@AMHOL
Cool, well if you end up having to implement it and need help, feel free to post on https://discuss.dry-rb.org/, it will require some intimate knowledge of the library and would be an interesting project, so would be nice to have it documented
Christian Meier
@mkristian
sure will do
Andy Holland
@AMHOL
FYI you can return the ast like this:
require 'bundler/inline'
gemfile(true) { gem 'dry-validation' }

s = Dry::Validation.Schema(build: false) do
  required(:age) { int? & gt?(18) }
end

s.ast
# => [[:rule, [:age, [:and, [[:rule, [:age, [:predicate, [:key?, [[:name, :age], [:input, Undefined]]]]]], [:rule, [:age, [:and, [[:rule, [:age, [:key, [:age, [:predicate, [:int?, [[:input, Undefined]]]]]]]], [:rule, [:age, [:key, [:age, [:predicate, [:gt?, [[:num, 18], [:input, Undefined]]]]]]]]]]]]]]]]]
Christian Meier
@mkristian
cool
George Millo
@georgemillo
how do people test their dry-v schemas, if at all?
I started writing some custom RSpec matchers, then I realised that my spec files were practically identical to my schemas themselves:
# schema:
required(:name).filled
optional(:age).maybe

# spec:
expect(schema).to have_required_key(:name).filled
expect(schema).to have_optional_key(:age).maybe
which is better than no tests at all, but tests take time to write, and now I'm writing 2x as much code for very little extra benefit
what are people's thoughts on writing tests for their schemas? do you find it worth the effort?
George Millo
@georgemillo
(I'm thinking that if I'm going to create custom matchers, my time is probably better spent creating them for Reform validations, which of course use dry-v anyway, but the matchers themselves will have to be different)
Christopher Dennl-Ortega Arrieta
@cdennl
@georgemillo don't test the schema but the instance which uses the schema
George Millo
@georgemillo
@cdennl I thought as much. In my case that would be Reform
Christopher Dennl-Ortega Arrieta
@cdennl
the schema alone is pretty meaningless
@georgemillo yes
or the model / op which uses the schema
George Millo
@georgemillo
there's a gaping hole in the Reform ecosystem for some smart RSpec matchers that test the validations
Fran Worley
@fran-worley
@georgemillo I've started moving away from tests like that. I test the actual result. I.e if I pass in this input, I expect to get the following errors. I think you get a more reliable result that focuses more on what matters (when my object is valid vs invalid and why) than the exact code I used to get there. Less coupling on the exact implementation therefore less brittle tests etc.
Fran Worley
@fran-worley
Plus as soon as you get advanced edge cases that depend on injected objects (say current user) you either end up testing manually or having such a complex matcher that it slightly defeats the point of having such a clean implementation library like dry-validations ! (Just my personal opinion of course 😜)
Lucas Hosseini
@beauby
^— :+1:
George Millo
@georgemillo
@fran-worley yeah, my custom matchers work like you described (passing in input, seeing if its valid), rather than testing the actual code (not even sure how you'd do that, presumably there's some way to reflect on the Schema class to figure it out)
where I realise I've gone wrong through is testing things at such a low level, i.e. at the level of the schema itself rather than the place that uses it (e.g. a reform form)
Fran Worley
@fran-worley
@georgemillo Sounds really cool, might do a minitest equivalent and actually use it in Reform :)
George Millo
@georgemillo
I opened at issue over at Reform if anyone wants to weigh in on this further: trailblazer/reform#414
Hannes Nevalainen
@kwando
hmm, is there a way to check that error messages have been defined for all predicates? would be nice to just tell "something" to check if I have provided error messages for all custom predicates (don't like to blow up on Dry::Validation::MissingMessageError)
This doc has a link to a gem called dry-component but the link is dead
has this gem been renamed?
oh wait, Google suggests that this is the old name for dry-system
amirite?
Nikita Shilnikov
@flash-gordon
yep