[solved] Schedule a report and make it available to end users in Aris Document Storage ?

Hello Michel,

Today I was hitting the same issue  of uploading report output directly to ADS, not willing to have an intermediate step where the output is stored in a file on the server.

I succeeded in doing this and am happy to share with the ARIS Community. Fair to say is that I've been inspired by the file ADSUploader.js which can be found in the 'Reports/Dashboard data' folder. The folder structure (Dashboarding/<database name>/) is inspired by the ADSUploader.js. You can change this and all hard-coded variables to your needs.

All required code is inside the function. One can copy/paste the function into ones code and then call it like shown in the first code snippet.

var xmlOutput = Context.createXMLOutputObject("Models.xml", "Models");
// assemble your XML document...
uploadXmlOutputToAds(xmlOutput, "DHAN", "Models.xml", "All models as xml", "All models as xml description.");

The upload function:

// Upload an in-memory XML document directly to ADS.
// params:
//   outputObject  : XML outputObject.
//   adsFolderName : Last part of the folder name in ADS where the document is stored.
//                   Complete foldername: /Dashboarding/<database name>/<adsFolderName>
//   adsFileName   : Name of the file in ADS. Include extension.
//   documentTitle : Title of the document as shown in ADS metadata.
//   documentDescription : Description of the document as shown in ADS metadata. Optional.
//
// When the folder in ADS doesn't exist, it will be created.
function uploadXmlOutputToAds(outputObject, adsFolderName, adsFileName, documentTitle, documentDescription) {
    var ads = Context.getComponent("ADS");
    var db  = ArisData.getActiveDatabase().Name(Context.getSelectedLanguage());
    var repository = ads.getADSRepository("portal");
    var rootFolder = repository.getRootFolder();
    var folderName = "Dashboarding\\" + db + "\\" + adsFolderName;
    var adsFolder = repository.getFolder(rootFolder, folderName);
    if (!adsFolder) {
        adsFodler = repository.createFolder(rootFolder, folderName);
    }
    var documentInfo = repository.createDocumentMetaInfo(documentTitle, adsFileName, documentDescription);
    var outputByteArray = outputObject.getDocumentContent();
    var document = repository.createAndOverwriteExistingDocument(
                        adsFolder, documentInfo,
                        new java.io.ByteArrayInputStream(outputByteArray) )
}

A screenshot of the result in ADS is attached.

Best regards,
Harm

The report creates its output and writes it (to the desktop client if executed manually , or on the server I guess if scheduled) . This is done when calling writeReport()

No, calling writeReport() writes a file whose content previously only existed in Server RAM to a temporary output directory (on the server) that is specifically created for an instance of your report execution. It does nothing more - no file is passed to the client while the report is running.

After your report concludes there are two conditions for a file to be transfered to the client:

• the file name must be registered as an output file (the selected file is registered by default, you may manually add more file names e.g. with Context.addOutputFileName(...))
• the file must actually exist (obviously not even the automatically registered selected file can be transfered if you never wrote such a file to the output directory, or if you wiped it from the output directory before the report concludes)

how to delete the unused file (the one created in setp one above) ?

The file created in step one can be deleted with

Context.deleteFile(Context.getSelectedFile(), Constants.LOCATION_OUTPUT);

In addition to deleting the file I also recommend that if you really don't want to transmit your file to any potential client you create your file with a different filename than the one provided by Context.getSelectedFile() (and adjust the first parameter of the Context.deleteFile(...) call). As I've written before, this file name is registered by default for output file transfer, therefore if you instead use something like "businessReport." + your file extension without registering this made up file name as an output file it would never be transfered to begin with. Also make sure to set the selected file with Context.setSelectedFile("noRealName.noRealExtension") to something that's not "businessReport." + your file extension. This prevents users from acquiring the file even if they were to manually set their output file name to "businessReport." + your file extension.

how to avoid altogetther the creation of this temporary file ? Is there a function of the OutputObject  that could be called to get the byte stream needed by the ADS functions ?

That I don't know. The ARIS OutputObject is most likely backed by different libraries depending on the output format (e.g. Apache POI if you create a docx Document, ...) and I doubt there is a universal method to tell all those different libraries to output into a buffer instead of writing a file to the filesystem.