Use simulacrum for typeclass boilerplate

[rklaehn]

This PR hides the simulacrum dependency from the pom and uses the typeclass annotation on all typeclasses. It also removes some manually added code that is now generated by simulacrum.

Review by @mpilquist or @milessabin if possible. It would be good if @johnynek could look at the generated code with an eye on algebra/algebird compatibility.

[rklaehn]

@mpilquist I did as you suggested, and sbt validate runs locally. So let's see if travis agrees.

[rklaehn]

Here is the POM:

<?xml version='1.0' encoding='UTF-8'?>
<project xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://maven.apache.org/POM/4.0.0">
    <modelVersion>4.0.0</modelVersion>
    <groupId>org.spire-math</groupId>
    <artifactId>cats-kernel_2.11</artifactId>
    <packaging>jar</packaging>
    <description>kernel</description>
    <url>https://github.com/non/cats</url>
    <version>0.4.0-SNAPSHOT</version>
    <licenses>
        <license>
            <name>MIT</name>
            <url>http://opensource.org/licenses/MIT</url>
            <distribution>repo</distribution>
        </license>
    </licenses>
    <name>kernel</name>
    <organization>
        <name>org.spire-math</name>
        <url>https://github.com/non/cats</url>
    </organization>
    <scm>
        <url>https://github.com/non/cats</url>
        <connection>scm:git:git@github.com:non/cats.git</connection>
    </scm>
    <developers>
        <developer>
            <id>non</id>
            <name>Erik Osheim</name>
            <url>http://github.com/non/</url>
        </developer>
    </developers>
    <properties>
        <info.apiURL>https://non.github.io/cats/api/</info.apiURL>
    </properties>
    <dependencies>
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>scala-library</artifactId>
            <version>2.11.7</version>
        </dependency>
        <dependency>
            <groupId>org.scoverage</groupId>
            <artifactId>scalac-scoverage-runtime_2.11</artifactId>
            <version>1.1.0</version>
            <scope>provided</scope>
        </dependency>
        <dependency>
            <groupId>org.scoverage</groupId>
            <artifactId>scalac-scoverage-plugin_2.11</artifactId>
            <version>1.1.0</version>
            <scope>provided</scope>
        </dependency>
        <dependency>
            <groupId>com.github.mpilquist</groupId>
            <artifactId>simulacrum_2.11</artifactId>
            <version>0.5.0</version>
            <scope>provided</scope>
        </dependency>
    </dependencies>
</project>

[mpilquist]

OK cool. We can remove the reference from the POM by using the pomPostProcess setting. I'm away from a computer now but I can post an example later. Alternatively, take a look at the build settings class in scodec-build.

On Jan 16, 2016, at 12:05 PM, Rüdiger Klaehn notifications@github.com wrote:

Here is the POM:

4.0.0 org.spire-math cats-kernel_2.11 jar kernel https://github.com/non/cats 0.4.0-SNAPSHOT MIT http://opensource.org/licenses/MIT repo kernel org.spire-math https://github.com/non/cats https://github.com/non/cats scm:git:git@github.com:non/cats.git non Erik Osheim http://github.com/non/ https://non.github.io/cats/api/ org.scala-lang scala-library 2.11.7 org.scoverage scalac-scoverage-runtime_2.11 1.1.0 provided org.scoverage scalac-scoverage-plugin_2.11 1.1.0 provided com.github.mpilquist simulacrum_2.11 0.5.0 provided — Reply to this email directly or view it on GitHub.

[rklaehn]

The .pom after removing all provided deps (including the very weird scoverage deps). @johnynek does this look OK to you? I think once we switch over to typeclassic there will be one additional dependency.

<?xml version='1.0' encoding='UTF-8'?>
<project xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://maven.apache.org/POM/4.0.0">
    <modelVersion>4.0.0</modelVersion>
    <groupId>org.spire-math</groupId>
    <artifactId>cats-kernel_2.11</artifactId>
    <packaging>jar</packaging>
    <description>kernel</description>
    <url>https://github.com/non/cats</url>
    <version>0.4.0-SNAPSHOT</version>
    <licenses>
        <license>
            <name>MIT</name>
            <url>http://opensource.org/licenses/MIT</url>
            <distribution>repo</distribution>
        </license>
    </licenses>
    <name>kernel</name>
    <organization>
        <name>org.spire-math</name>
        <url>https://github.com/non/cats</url>
    </organization>
    <scm>
        <url>https://github.com/non/cats</url>
        <connection>scm:git:git@github.com:non/cats.git</connection>
    </scm>
    <developers>
        <developer>
            <id>non</id>
            <name>Erik Osheim</name>
            <url>http://github.com/non/</url>
        </developer>
    </developers>
    <properties>
        <info.apiURL>https://non.github.io/cats/api/</info.apiURL>
    </properties>
    <dependencies>
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>scala-library</artifactId>
            <version>2.11.7</version>
        </dependency>
    </dependencies>
</project>

[ceedubs]

@rklaehn thanks a bunch for this work!

What's the current status of this? @mpilquist, @johnynek, @milessabin does it look good to you?

[mpilquist]

👍

[johnynek]

👍 I guess. This PR is huge and I am not confident that I see all that is going on,

To be clear: I'm happy to use simulacrum here as long as consumers of the library see 0 additions deps to use the code. I understand the provided here was a phantom, and not actually needed at runtime. Is that right?

[rklaehn]

@johnynek yes, the provided dependency on simulacrum is not needed at runtime. That is why it is removed from the .pom.

[ceedubs]

👍

@rklaehn sorry there are merge conflicts. If you fix them and leave a comment that you have updated the PR, I'll try to merge before more conflicts pop up.

[johnynek]

@rklaehn Thanks for the work on this!

[rklaehn]

@ceedubs the merge conflict was because mima was configured on topic/kernel. It should be fixed now.

I am doing a talk about spire and the typelevel ecosystem at my local scala user group in munich on 2016-03-08. It would be great if algebra and spire on master would depend on cats-kernel at that time, so I could report something positive... :-)

[ceedubs]

Thanks @rklaehn!