Xtend M3 Error while using Customized Update Transaction

Hi All,

We are facing issue in XtendM3 when trying to use multiple table in XtendM3 MI program for Update Transaction. It shows error and the error file is attached.

[View:/cfs-file/__key/communityserver-discussions-components-files/127/EXTENSION_EXECUTE-EXT530MI-EXT530MI-AUTOWORLDCOM_TST-AJAIN19-20210517-170907.777893775.txt:320:240]

Thanks,

Akshay

Find more posts tagged with

Accepted answers

All comments

swatiy

Hi Akshay,

I see field does not exist in container error here. What exatcly is line no. 109 in your code?

UpdEmplLeave.main(UpdEmplLeave.xtendm3:109)

If you can share snapshot of this, may get more clarity.

Also please explain more about "not able to use multiple tables" ? You mean different tables or same table multiple times?

Thanks!

Legacy Contributor

Hi Swati,

I'm sharing snapshot.

Thanks

swatiy

Hi Akshay,

You are doing Closure before doing Read. This format seems is not correct.

You can refer Infor Xtend M3 document for reading table with selection, check the formats and change your code.

antonio-caria

You should put log (debug"messages in all code.

All messages should be translated when possible....

Avoid Code like Infor style (MAK):

if (mi.inData.get("CONO") == null) {
iCONO = program.LDAZD.CONO.toString();
} else {
iCONO = mi.inData.get("CONO").trim();
}

should be:

iCONO = mi.inData.get("CONO") == null ? program.LDAZD.CONO.toString() : mi.inData.get("CONO").trim();

benjamin-nowak

What's wrong with MAK style of code? You need to recognize that 99% of people using XM3 are legacy MAK Java (and some of us even MDM RPG) developers. If it's clear and anyone can read it, i say do it.

antonio-caria

MAK Code is not good Code (secure, simple, fast, standard, etc)

I can write a standard M3 MAK program in less than 30% of lines code without big changes.

Of course for RPG programmers is nice ....

benjamin-nowak

Antonio, I was part of the original transformation from RPG to Java. There was a distinct reason we retained the constructs and methods (built into the foundation classes) of RPG. Regardless of what language, code itself can be either good or bad, but when you look at resources available to translate and debug/trace.. you NEED to be in a format that is common. That is the point I was trying to make. You may be a wizard at Groovy, but if no-one else can understand what you are doing, I would not want it on my system.

Legacy Contributor

@antonio-caria / @benjamin-nowak

While I think it's perfectly fine to use ternary conditional operator, I don't believe we should ask everyone to start converting their if statements to this format. Specially when you're retrieving parameters and subparameters and transforming them with toString and trim which makes readability a bit harder sometimes.

MAK style had lots of advantages and of course a bit of shortcomings. Using the very format and structure we have had in MAK, we have been able to build XtendM3, which would not have been possible otherwise.

As for the shortcomings, XtendM3 lets you to use the Java and Groovy capability to a bigger extend and leverage a modern coding standard but being consistent is of course vital so that solutions can be maintained easily.

And as for the specific parameter retrieval here, I covered it in a workshop of what can be done to abstract this in a nice way, which is something I'll take into consideration for adding it to the framework so it's part of the API.

antonio-caria

Well Java is Java.

You can write good and bad code.

The code transformed from RPG into Java (like RIO sofrware) is not good Java and I can explain easly why - We call it JAVEX code.

I have worked with many transformation sotware in IBM, Oracle, etc and every has this kind of problem.

As You know, Infor has improved the code in last releases.

Now, for me I prefer 10 lines of code to 50 lines when the result is exactly the same - but that's my opinion.

Copyright © 2025 Infor. All rights reserved.