nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Generate Report From Compare & Merge

medini_1047650437

Hi Jan，

Using Compare & Merge, we can get the difference between two projects.

Is there such a script: generate a report based on the compare&merge.

In fact, they want to be able to see the differences of two versions' Safety Requirement in a table rather than only through the compare view in medini.

Thanks in advance.

Sincerely.

Qianqian Liang

Find more posts tagged with

medini

medini legacy

Accepted answers

All comments

medini_1024169437

Qianqian, that's technically possible and we added some support to the tool back in 2019 or 2020.

However, I don't have a sample script that I can post here.

You have to put a script into the target side of the compare. You will find it then in the "Execute" menu of the compare tree root node.

The top level diff node is passed as selection into the script so you can go on from there.

You can use "Debug -> What is this (in debug mode) to check what attributes and references each node has.

Hope that helps you to get started,

Jan

medini_1053080245

Hi Jan,

Something I can't get is how to generate a report with my script. I understand that I can create a contextual script to get data from the compare/merge browser but I will only be able to have result through the console? Or is there something I miss?

In addtion, it seems that in the help support, there is also a possibility to have a table like the image below for the FMEA. My idea would be to have the requirement table with all the Safety requirements that have been modified. But i have only the difference details

medini_1024169437

Hi Romain,

sorry for the delay. About the first point: your script indeed has to handle the output, you could wrote something to the console, to the clipboard or write an Excel file.

You don't have access to BIRT though, so you cant write a BIRT report from that script (I assume that was fundamentally your question).

What you see in that picture is the result of an experimental option which is in the tool since a while, check the compare/merge preferences.

Regards

Jan

medini_1053080245

Ok i will check again this feature because it is checked to yes and i don't have the window table (only the "difference details") for the FSR table for example.

About the fact to get the output into an Excel file, can you give me some details (or give me the link to a dedicated forum discussion) about the way to generate Excel from a Script? I know only how to parse/import but not the opposite.

medini_1024169437

Romain,

I think you have to double click the table diff node to get this one open, hope it works for you.

For Excel output, I think John Paul had the same question here and Eckhardt pointed himt to other threads that do not really explain it technically.

I need to dig out some sample code. In short: you can use the very same excel.js library for import and export.

Jan

medini_1024169437

Romain, I found a simple text script that can open a excel file for writing and set some cell values.

Simplest API, no creation/duplication of worksheets, styling etc. But maybe a good start for you.

Happy Scripting

Jan

medini_1053080245

Thx i am testing your scripts.

What is the commands to:

Dupplicate a file that I have as a template?

The idea is to have a template file (Excel) into my project and dupplicate it to a new file (to edit it) as many times as I need.

medini_1024169437

Good idea. You could use the appendToFile util offered by helper.js, see https://medini.freshdesk.com/a/solutions/articles/1000273630

Jan

medini_1053080245

thx yes it works fine

medini_1053080245

Other questions now: the finder function seems not working when I switch to compare/merge perspective. The script is working (at least to get data with the finder) but when I switch to compare/merge perspective it does not work.

Is there a different way to do when we are in this perspective?

medini_1024169437

That sounds odd to me, Romain, that should be perspective independent, could you explain what does not work?

Jan

medini_1053080245

Global.getFinder(selection[0]).findByType(Metamodel.compare.Match).asArray();

nothing happen in the console...

medini_1053080245

According to https://medini.freshdesk.com/support/discussions/topics/1000092973/page/1?url_locale=, i have added this to know what is the issue (because no error message when I am in the diff perspective).

try {
 callSomeCode();
} catch (e) {
 alert("Failed with" + e);
}

It seems that the finder is not known in this perspective (error = finder not defined). How can i use finder here

medini_1024169437

Romain, there are two things to consider here:

there is no "finder" variable by default because in a diff/merge context you deal with two projects, sozrce and target
the approach to use Global.getFinder was absolutely correct, but selection[0] is the root of the diff model (which does not belong to a project itself)
instead uf left and right of the root diff node to acces sthe two projects and create finder for that

Here some sample coee that should explain it better I think:

/**
 * Sample code to establish finder instances in a diff script context
 */
try {
 if (selection) {
 // use this finder to find inside the diff tree
 var diffFinder = Global.getFinder(selection[0]);
 console.log("{0}", selection[0].left);
 console.log("{0}", selection[0].right);
 // use this finder to find in target project
 var leftFinder = Global.getFinder(selection[0].left);
 var firstPart = leftFinder.findByType(Metamodel.sysml.SysMLPart).first();
 console.log("{0}", firstPart.name);
 }
} catch (e) {
 alert("Failed with" + e);
}

medini_1024169437

No I read you message again and saw that you werent looking for a project finder but instead wanted to search the diff tree for matches, sorry.

Anyway, that worked for me, I think you just forgot to log the findings:

 ...
 diffFinder.findByType(Metamodel.compare.Match).forEach(function (match) {
 console.log("Found match {0}", match);
 });

Happy Scripting

Jan

medini_1053080245

Thx a lot Jan.

medini_1063548292

Hi All,

I made an initial skeleton for the BIRT report.

This template is standalone = not depending on the analyze_data_set_library or anything else.

Execute the Custom report on any model browser element.

Visually I targeted primarily the Excel output format but it will also work in other formats.

Happy reporting,

Marian

medini_1024169437

Thanks Marian for posting,

could you give some quick user guidance? I was trying it and figured that I had to execute some project diff first and gthen your report grabs the diff viewer and reports the diff contenbt, correct?

Happy Scripting

Jan

medini_1063548292

HI Jan,

Yes you got it right! And sorry for no guidance. I have created it ad-hoc as I needed to debug something with the Diff/Merge .

So here it goes...

The basic idea of the report is that it grabs the Difference viewer.
Then it will grab the result of the last merge and processing the Match objects and populate them in a Scripted DataSet

Note: this was the only way how to it as the Match is not an EMF class so one cannot use OCL DS on a diff/merge objects

Current report version simply prints basic information on the results into the Table,
The Reported Table is organized in Groups: Addition, Deletion, Update and Unmatched cases based on the content of 'Left', 'Right' object

Currently It is not depended on the analyze library since it does not use any of the metaclass DataSets (yet).

If one would like to create a specific diff/merge report on the results of a FMEDA worksheet or system model ,... then the analyze library would be needed or the OCL DataSets have to be utilized in the report.

PS: I think the same can be also done in M2doc however that would require scripted services as it requires processing of non-EMF objects. However, I think the BIRT is still better way since it allows generation of an Excel output, and better handling f the Tables

Happy reporting,

Marian

medini_1024169437

Super explanation Marian,

one open point remains, you write "... require scripted services as it requires processing of non-EMF objects ..." , which objects do you mean? THe diff tree is EMF, so...

Jan

medini_1063548292

Hi Jan,

Sure I will comment.. I stand corrected .

I took a wrong assumption that those are non EMF objects because I failed to create plain OCL DataSets in BIRT.

Later I learned ( some wise JM man told me) that compare classes are EMF objects and only the BIRT UI does not proposes the compare package suggestions.

As a confirmation I create skeleton for m2doc. Attached to the comment.

Long story short the AQL types works with compare::Match, compare::Diff, etc

PS: Hold your hats as this time the memory blow is a real deal. I tried to add some HEAP cleanup at the beginning and at the end of the report but still did not manage to fully get rid HEAP build up. It climbed up to 21 GB ! for ~ 90k EMF compare package objects.

I think with some filtering and better structured template data organization would help to improve the memory performance

Happy reporting,

Marian

medini_1024169437

Awesome, JM man is happy

Jan