[jira] [Updated] (SPARK-4819) Remove Guava's "Optional" from public API
[
https://issues.apache.org/jira/browse/SPARK-4819?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen updated SPARK-4819:
-
Attachment: SPARK_4819_null_do_not_merge.patch
Here was what I tried when replacing Optional with use of null. It just went
well out of hand even after figuring out an ugly workaround for varargs. It
ended up being a good change in some ways (correctly expressing the bound for
Java RDDs, fixing up a Scala/Java Long problem, removing some fake class tags)
but this ended up far too unwieldy
> Remove Guava's "Optional" from public API
> -
>
> Key: SPARK-4819
> URL: https://issues.apache.org/jira/browse/SPARK-4819
> Project: Spark
> Issue Type: Sub-task
> Components: Spark Core
>Affects Versions: 1.2.0
>Reporter: Marcelo Vanzin
>Assignee: Sean Owen
> Attachments: SPARK_4819_null_do_not_merge.patch
>
>
> Filing this mostly so this isn't forgotten. Spark currently exposes Guava
> types in its public API (the {{Optional}} class is used in the Java
> bindings). This makes it hard to properly hide Guava from user applications,
> and makes mixing different Guava versions with Spark a little sketchy (even
> if things should work, since those classes are pretty simple in general).
> Since this changes the public API, it has to be done in a release that allows
> such breakages. But it would be nice to at least have a transition plan for
> deprecating the affected APIs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Updated] (SPARK-4819) Remove Guava's "Optional" from public API
[
https://issues.apache.org/jira/browse/SPARK-4819?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen updated SPARK-4819:
-
Assignee: Sean Owen (was: Reynold Xin)
I'm going to give this a try. The game plan is roughly to replace Optional
with T, where the value can be null. This is the closest equivalent in the Java
7 JDK. Of course, Java 8 would be nice here too, and would be pleased to see
deciding to require Java 8 for 2.x, but that's a different question.
One catch: It's simple to use {{v.orNull}} on an instance of Scala's
{{Option[V]}} to get the value or null, for use with Java. It's like the mirror
to {{Option(v)}}. However this only works if Scala knows type V is a reference
type. So this means adding ">: Null" bounds to a lot of types in the Java API.
Which isn't a bad idea, in that Java RDDs can only contain objects and not
primitives anyway.
Comments?
> Remove Guava's "Optional" from public API
> -
>
> Key: SPARK-4819
> URL: https://issues.apache.org/jira/browse/SPARK-4819
> Project: Spark
> Issue Type: Sub-task
> Components: Spark Core
>Affects Versions: 1.2.0
>Reporter: Marcelo Vanzin
>Assignee: Sean Owen
>
> Filing this mostly so this isn't forgotten. Spark currently exposes Guava
> types in its public API (the {{Optional}} class is used in the Java
> bindings). This makes it hard to properly hide Guava from user applications,
> and makes mixing different Guava versions with Spark a little sketchy (even
> if things should work, since those classes are pretty simple in general).
> Since this changes the public API, it has to be done in a release that allows
> such breakages. But it would be nice to at least have a transition plan for
> deprecating the affected APIs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Updated] (SPARK-4819) Remove Guava's "Optional" from public API
[
https://issues.apache.org/jira/browse/SPARK-4819?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Reynold Xin updated SPARK-4819:
---
Issue Type: Sub-task (was: Task)
Parent: SPARK-11806
> Remove Guava's "Optional" from public API
> -
>
> Key: SPARK-4819
> URL: https://issues.apache.org/jira/browse/SPARK-4819
> Project: Spark
> Issue Type: Sub-task
> Components: Spark Core
>Affects Versions: 1.2.0
>Reporter: Marcelo Vanzin
>
> Filing this mostly so this isn't forgotten. Spark currently exposes Guava
> types in its public API (the {{Optional}} class is used in the Java
> bindings). This makes it hard to properly hide Guava from user applications,
> and makes mixing different Guava versions with Spark a little sketchy (even
> if things should work, since those classes are pretty simple in general).
> Since this changes the public API, it has to be done in a release that allows
> such breakages. But it would be nice to at least have a transition plan for
> deprecating the affected APIs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Updated] (SPARK-4819) Remove Guava's "Optional" from public API
[
https://issues.apache.org/jira/browse/SPARK-4819?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen updated SPARK-4819:
-
Target Version/s: 2+
> Remove Guava's "Optional" from public API
> -
>
> Key: SPARK-4819
> URL: https://issues.apache.org/jira/browse/SPARK-4819
> Project: Spark
> Issue Type: Task
> Components: Spark Core
>Affects Versions: 1.2.0
>Reporter: Marcelo Vanzin
>
> Filing this mostly so this isn't forgotten. Spark currently exposes Guava
> types in its public API (the {{Optional}} class is used in the Java
> bindings). This makes it hard to properly hide Guava from user applications,
> and makes mixing different Guava versions with Spark a little sketchy (even
> if things should work, since those classes are pretty simple in general).
> Since this changes the public API, it has to be done in a release that allows
> such breakages. But it would be nice to at least have a transition plan for
> deprecating the affected APIs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
