Updated pvdata

[shroffk]

Have to wait till PR#18 is merged before this can be merged.

[georgweiss]

Does not seem to build on Java 10 (several issues).

As for technical matters, I have insufficient experience with the code to say anything clever. However, I do have some comments in regards to VImage related classes:

1. Is the VImageType enum supposed to cover all relevant permutations of properties like colour space, number of bytes per pixel and others? If so, why not use a class that encapsulates these properties?
2. In VImgeDataType, why is there a need for a string type? Plus, the methods in the class are tightly dependent on the order in which the values appear in the source code. As an alternative, one could maybe consider an integer (or short) property in the enum in order to not depend on the ordinal, e.g. pvBoolean(0), pvByte(1), pvShort(2) etc.

[shroffk]

Does not seem to build on Java 10 (several issues).

The master itself is not building on jdk10 as yet, have expanded the scope of the
PR 17 #17

Is the VImageType enum supposed to cover all relevant permutations of properties like colour space, number of bytes per pixel and others? If so, why not use a class that encapsulates these properties?

I think the list is intended to be fixed to what area detector can handle.

In VImgeDataType, why is there a need for a string type? Plus, the methods in the class are tightly dependent on the order in which the values appear in the source code. As an alternative, one could maybe consider an integer (or short) property in the enum in order to not depend on the ordinal, e.g. pvBoolean(0), pvByte(1), pvShort(2) etc.

The class was based on the org.epics.pvdata.ScalarType , I wanted to keep it as similar as possible so that client code dealing with Types would remain the same.
I will ask anyone if it makes sense to keep the pvString, else we could consider to remove it.

Now that I look at it in detail, I agree with you that the heavy use of ordinal is rather unfortunate.
Ideally I would feel that the enum should be able to tell you if the type is a primitive, an integer, a unsigned value, etc....without having to check the ordinal.
However, this class was designed to match org.epics.pvdata.ScalarType I would like to keep them the same...we should consider fixing the org.epics.pvdata.ScalarType but that would require checking up with the other developers. This can be a discussion topic for the Aug14th meeting.

[georgweiss]

OK, I admit I have not inspected the commits in detail, but I think that we should try to spread the word that any new code should be Java 10 compliant. My impression is that Javadoc stuff that needs most attention.

[georgweiss]

I was under the impression that my fixes for Java 9+ were merged already to master. Can you please clarify?

[shroffk]

@georgweiss I too was under the impression that all your PR's have been merged.

[shroffk]

@georgweiss it seems like the javadoc issues are now with https://github.com/epics-base/exampleJava I will request Marty to have a look at it

From Marty

I have been spending all my time on C++ code so I did not pay attention to this request until now.
I have checked out branch updated_pvdata and will start looking at it.

[mrkraimer]

I am at

mrk> java -version
openjdk version "1.8.0_181"
OpenJDK Runtime Environment (build 1.8.0_181-b13)
OpenJDK 64-Bit Server VM (build 25.181-b13, mixed mode)
mrk> pwd
/home/epicsv4/masterJava/exampleJava
mrk> cat .git/config 
[core]
	repositoryformatversion = 0
	filemode = true
	bare = false
	logallrefupdates = true
[remote "origin"]
	url = https://github.com/epics-base/exampleJava.git
	fetch = +refs/heads/*:refs/remotes/origin/*
[branch "master"]
	remote = origin
	merge = refs/heads/master
mrk> 

Then

mrk> mvn package
...
Building index for all classes...
Generating /home/epicsv4/masterJava/exampleJava/serviceAPI/target/apidocs/allclasses-frame.html...
Generating /home/epicsv4/masterJava/exampleJava/serviceAPI/target/apidocs/allclasses-noframe.html...
Generating /home/epicsv4/masterJava/exampleJava/serviceAPI/target/apidocs/index.html...
Generating /home/epicsv4/masterJava/exampleJava/serviceAPI/target/apidocs/help-doc.html...
[INFO] Building jar: /home/epicsv4/masterJava/exampleJava/serviceAPI/target/epics-examples-serviceAPI-4.3.0-SNAPSHOT-javadoc.jar
[INFO] ------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO] 
[INFO] exampleJava ........................................ SUCCESS [  1.573 s]
[INFO] exampleJava - arrayPerformance ..................... SUCCESS [  2.922 s]
[INFO] exampleJava - database ............................. SUCCESS [  1.395 s]
[INFO] exampleJava - exampleClient ........................ SUCCESS [  1.677 s]
[INFO] exampleJava - exampleLink .......................... SUCCESS [  1.541 s]
[INFO] exampleJava - helloPutGet .......................... SUCCESS [  1.152 s]
[INFO] exampleJava - helloRPC ............................. SUCCESS [  1.201 s]
[INFO] exampleJava - powerSupply .......................... SUCCESS [  1.232 s]
[INFO] exampleJava - pvDatabaseRPC ........................ SUCCESS [  1.568 s]
[INFO] exampleJava - serviceAPI ........................... SUCCESS [  1.108 s]
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 16.983 s
[INFO] Finished at: 2018-08-14T15:15:39-04:00
[INFO] Final Memory: 30M/644M
[INFO] ------------------------------------------------------------------------

So what am I expected to do in response to

it seems like the javadoc issues are now with https://github.com/epics-base/exampleJava I will request Marty to have a look at it

[shroffk]

@mrkraimer I wanted you to look at the code changes here and let us know if you see any glaring problems.
If you look at the pvData code you will notice the use of the PVNumberArray interface instead of the PVScalarArray...the PVNumberArray does the transformation, gives access to the array elements without requiring copy, etc...

@georgweiss
Can we merge this and open a new issue ticket about the VImageType and its use of ordinals. We will not release epicsCoreJava until we resolve this.

[mrkraimer]

Look OK to me.
Go ahead and merge.

[georgweiss]

Go ahead and merge. /Georg From: Kunal Shroff <notifications@github.com> Reply-To: epics-base/epicsCoreJava <reply@reply.github.com> Date: Monday, 20 August 2018 at 21:03 To: epics-base/epicsCoreJava <epicsCoreJava@noreply.github.com> Cc: Georg Weiss <georg.weiss@esss.se>, Mention <mention@noreply.github.com> Subject: Re: [epics-base/epicsCoreJava] Updated pvdata (#19) @mrkraimer<https://github.com/mrkraimer> I wanted you to look at the code changes here and let us know if you see any glaring problems. If you look at the pvData code you will notice the use of the PVNumberArray interface instead of the PVScalarArray...the PVNumberArray does the transformation, gives access to the array elements without requiring copy, etc... @georgweiss<https://github.com/georgweiss> Can we merge this and open a new issue ticket about the VImageType and its use of ordinals. We will not release epicsCoreJava until we resolve this. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#19 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/Ah9CEFELSfJEqO7Trr8ng6JmpJ3GzfcFks5uSwfjgaJpZM4VmpQR>.