Usage of addResultFile() in User Defined Java Class

[dave-csc]

Hello,

I'm trying to add the result files functionality in a custom transform made with User defined Java Class.

To do so, I added this code in the processRow() method (f is the java.io.File object related to the file I just processed):

    addResultFile(new ResultFile(
                    ResultFile.FILE_TYPE_GENERAL,
                    org.apache.hop.core.vfs.HopVfs.getFileObject(f.getPath()),
                    getPipelineMeta().getName(),
                    getTransformName()));

However, when I execute the code I have no errors (even the files are correctly created), but a Get files from result in a subsequent pipeline can't find the files that I "added" to the files result. What am I doing wrong?

Thanks for your support :)

[geocine]

files from result (the Get files from result transform) only reads the previous pipeline result that the workflow action passes in (pipeline.getPreviousResult()), not the current pipeline's runtime state.

[dave-csc]

Hi @geocine,

that's exactly what I meant for "subsequent pipeline": the scenario is that the UDJC is included in a (first) pipeline, and generates some files. A second pipeline is linked to the first in a parent workflow: I would expect to read the generated files from the previous pipeline there, but I have no entries in Get files from result (and also no errors in the workflow log, even the files are correctly created)

[geocine]

have you tried passing the filename out as a row and adding it to the result in the parent pipeline? then in the parent pipeline you can add Files to result after the mapping step and point it to the full path field.

[hansva]

@dave-csc could you provide what you currently have? Or make a quick sample based on what you have. Then I'll take a look to see if I can figure out where it's going wrong

[dave-csc]

Hi @hansva,

I try to post this file that contains the UDJC: _map-blob-to-file.hpl.txt

It's a sort of "service" pipeline that should be used inside a pipeline to convert the BLOB fields in a database into the corresponding files: the mapping inputs are the file path (usually a constant), the file name (a different one for each record) and the blob as Binary type.

This pipeline is actually used in the first pipeline to generate the files from a photo database (Oracle) and should add the processed files into the Result files area. But when I set up a second pipeline that reads the files from that results, there are no entries...

Feel free to ask if you need other info... Thanks :)