crystal-lang/crystal

The Crystal programming language | http://crystal-lang.org | Fund Crystal's development: http://is.gd/X7PRtI | Docs: http://crystal-lang.org/docs/ | API: http://crystal-lang.org/api/

From IRC (bridge bot)

@FromIRC

<mfiano> That project seems to be structured well and very easy to read

<mfiano> (and nicely documented)

George Dietrich

@Blacksmoke16

i try :smiley:

From IRC (bridge bot)

@FromIRC

<mfiano> at least as a first real project I've seen

Jason Waldrip

@jwaldrip

Getting some weird behavior...

class Psykube::Generator::Podable < ::Psykube::Generator
  alias Resource = Pyrite::Api::Apps::V1::Deployment | Pyrite::Api::Batch::V1beta1::CronJob | Pyrite::Api::Batch::V1::Job | Pyrite::Api::Apps::V1::StatefulSet | Pyrite::Api::Apps::V1::DaemonSet | Pyrite::Api::Core::V1::Pod

  protected def result : Resource
    case manifest.type
    when "Deployment"
      Deployment.result(self)
    when "Cron"
      CronJob.result(self)
    when "Job"
      Job.result(self)
    when "StatefulSet"
      StatefulSet.result(self)
    when "DaemonSet"
      DaemonSet.result(self)
    when "Pod"
      Pod.result(self)
    else
      raise "Invalid type: `#{manifest.type}`"
    end
  end
end

given that... I am getting a strange error

George Dietrich

@Blacksmoke16

manifest.type should be an enum :)

Jason Waldrip

@jwaldrip

❯ shards build && mv bin/psykube `which psykube` --error-trace
Dependencies are satisfied
Building: psykube
Error target psykube failed to compile:
Showing last frame. Use --error-trace for full trace.

In src/cli/commands/concerns/kubectl.cr:139:20

 139 | spec = podable.spec
                      ^---
Error: undefined method 'spec' for Pyrite::Api::Admissionregistration::V1::MutatingWebhookConfiguration (compile-time type is Pyrite::Kubernetes::Object+)

George Dietrich

@Blacksmoke16

got a link to the code?

Jason Waldrip

@jwaldrip

ya, let me push

Jason Waldrip

@jwaldrip

https://github.com/psykube/psykube/blob/pyrite-update/src/cli/commands/concerns/kubectl.cr#L138

George Dietrich

@Blacksmoke16

does that spec method exist on all those types in the union?

Jason Waldrip

@jwaldrip

Let me double check

George Dietrich

@Blacksmoke16

it seems the type of that union is being collapsed into their parent Pyrite::Kubernetes::Object+ type which doesn't have that method

https://github.com/crystal-lang/crystal/issues/2661#issuecomment-877286921 is prob similar to what's going on ^ if i had to guess

Jason Waldrip

@jwaldrip

Which is strange... the return type method ensures the proper type is returned

or at least it should

George Dietrich

@Blacksmoke16

but you're doing a .as(Psykube::Generator::Podable::Resource)

where Resource is an alias to a union of other types

Jason Waldrip

@jwaldrip

so really I would have to case over each type in the union to get what I want

George Dietrich

@Blacksmoke16

or make use of overloads for specific types

Jason Waldrip

@jwaldrip

ya, let me see what I can do

Alex Kampa

@alex-kampa

Now here's a bit of a surprise:

puts [1, 2, 3, 4, 5].reduce(0) { |acc, i| acc = acc + 1 if i.to_s =~ /3/ }

Error: undefined method '+' for Nil (compile-time type is (Int32 | Nil))

George Dietrich

@Blacksmoke16

i think you need some parans

parens*

i think the if is making the type of acc possibly nil, because the return value of the block is used to set the acc

.reduce(0) { |acc, i| i.to_s =~ /3/ ? acc + i : acc }

is 3 the expected value here?

From IRC (bridge bot)

@FromIRC

<mfiano> Can someone ELI5 the ~> version operator?

George Dietrich

@Blacksmoke16

From IRC (bridge bot)

@FromIRC

<mfiano> I read that

<mfiano> Maybe I'm dumb

George Dietrich

@Blacksmoke16

afaik its the same as the ^ in other langs

From IRC (bridge bot)

@FromIRC

<riza> mfiano: the idea with ~> is that it allows for upgrades within the specified point but no higher

George Dietrich

@Blacksmoke16

just a shortcut to say allow versions within this range

From IRC (bridge bot)

@FromIRC

<riza> so if you ~> x.y.z, you leave room for z to increment but not y

George Dietrich

@Blacksmoke16

From IRC (bridge bot)

@FromIRC

<riza> and if you ~> x.y, y can increment, z can be anythnig and x cannot change

<mfiano> Oh very good. I don't see why you would want to leave room for y or x to upgrade...how can you know anything other than a patch won't break your code?

<riza> its used under the assumption that the project you're referencing follow SemVer

Alex Kampa

@alex-kampa

Thanks @Blacksmoke16 your solution works, even though

puts [1, 2, 3, 4, 5].reduce(0) { |acc, i| i.to_s =~ /3/ ? acc + i : acc }

is a bit harder to read than

puts [1, 2, 3, 4, 5].reduce(0) { |acc, i| acc = acc + 1 if i.to_s =~ /3/ }

but ok I guess that's the price to pay for a compiled language ... and shows that the transition from Ruby to Crystal is not always that easy.

From IRC (bridge bot)

@FromIRC

<riza> mfiano: if the project is following SemVer, then you know that anything which is a breaking change is a major version increment

<mfiano> Ok I see

George Dietrich

@Blacksmoke16

@alex-kampa whats the end goal of this code?

i noticed i had acc + i instead of acc + 1, so are all you wanting to do is count the number of time 3 is in the array?

Alex Kampa

@alex-kampa

counting elements in an array that match a certain regex

at this stage just playing around with various ways to achieve that

George Dietrich

@Blacksmoke16

puts [1, 2, 3, 4, 5].count { |i| i.to_s.matches? /3/ }

or in this specific case puts [1, 2, 3, 4, 5].count 3

Where communities thrive

People

Repo info

Activity