Disqus - Latest Comments for etorreborre

Re: Type-safe Strings with Scala Macros

etorreborre — Thu, 13 Jul 2017 13:29:08 -0000

Note: there is a nice library now for such use cases: http://co.ntextu.al.

Re: Modern Functional Programming: Part 2

etorreborre — Wed, 16 Nov 2016 15:39:30 -0000

I think we agree. I just didn't want to give the impression that denotational semantics was just running natural transformations.

Re: Modern Functional Programming: Part 2

etorreborre — Wed, 16 Nov 2016 15:24:34 -0000

John, I wonder if you are not too enthusiastic in your blog post. I re-read the part on denotational semantics: "Denotational semantics provide an unprecedented ability to reason about programs in a composable and modular fashion". My understanding is that this subject was historically not obvious at all! It was precisely because some programs could not be given some obvious modular denotational semantics that monads and monad transformers were invented. Just to quote an article (http://citeseerx.ist.psu.ed... "For example, it has long been recognized that traditional denotational semantics [24] is not suitable for compiler generation for a number of crucial reasons [16], among which is the lack of modularity and extensibility".

Re: Typelevel.scala | EDSLs as functions

etorreborre — Fri, 04 Nov 2016 05:58:15 -0000

Can't wait for part 2!

Re: Modern Functional Programming: Part 2

etorreborre — Wed, 28 Sep 2016 01:59:47 -0000

Hi John, I'm still reading your article and noticed a typo in the `transfer` function. It should have `to: To[Account]` as a parameter.

Re: Scalaz-Stream: Feed `Process` through the given effectful `Channel` concurrently

etorreborre — Wed, 09 Apr 2014 03:34:24 -0000

How would you go if you had some of your task having to wait on the result on previous tasks before running. That is, everything that is before Task 4 can run concurrently, then Task 4 must run, then, based on the result of Task 4, we decide if Tasks 5 to 10 need to run (and if they do, they run concurrently). At the moment, with my implementation I can do that because I just pipe a Process1[A, Task[A]] (let's call it execute1) onto my Process[Task, A] then call .sequence(n) on that. execute1 is a function using receive1 and calling itself recursively so it can take decisions based on the previous As. Of course, it is also side effecting in the sense that it starts the tasks right away. I'm not sure if 1. what I wrote is understandable, 2. if there's a better way of doing this using mergeN. I'm probably going to ask the question on the mailing list...

Re: Scalaz-Stream: Feed `Process` through the given effectful `Channel` concurrently

etorreborre — Tue, 08 Apr 2014 22:16:48 -0000

Hi Eugene, here is what I used to achieve concurrency: https://gist.github.com/eto.... It is slightly different from what you did. What do you think?

Re: Entities and Values

etorreborre — Wed, 05 May 2010 00:27:33 -0000

> It is also important that an entity can be constructed by passing nulls to facilitate partial domain model instantiations in unit tests.

I'm not sure about this recommendation. Most of time I pass mocks or real entities to construct aggregates. I found very valuable to have a flexible set of test fixtures providing entities/values for different contexts that you evolve with my production code.

Re: Running Scala specs tests in Maven with JUnit 4

etorreborre — Sun, 26 Jul 2009 20:05:59 -0000

Hi Mark,

Thanks for your post.

2 tips for making your specification even more concise (with specs.1.6.0-SNAPSHOT or 1.5.1 SNAPSHOT):

This will make your specification class a JUnit test:
class BasicCassandra extends SpecificationWithJUnit { ... }

It is still runnable in the console with:

scala -cp <...> run BasicCassandra

And you can write (if you like it):

cass.connect().before
cass.close().after

Cheers,

Eric.

Re: Scala and EasyMock

etorreborre — Mon, 20 Jul 2009 00:05:39 -0000

1. Right now I don't have the bandwidth to write a full Easymock trait, tough that would be a nice addition to have.

2. Yes, you can access the JMock functionalities without using specs:

http://code.google.com/p/sp...

There is also a trait for Mockito which is actually my favorite mock library (I didn't provide a reusable trait outside of specs yet but there's very little to do for that and I'll certainly do it)

Re: Scala and EasyMock

etorreborre — Sun, 19 Jul 2009 22:45:17 -0000

Hi, you can get an even better syntax by using the "Manifest" undocumented feature of Scala.

useMock[ToMock] { mock =>
val toTest = new ToTest(mock)
toTest.doSomethingWithToMock()//should call method1() and method2() on mocked
}

See how this was done for specs and jmock for example: http://code.google.com/p/sp...

Cheers,

Eric.