FME Server 2020.2 - Filecopy used in Data Download App not using Zip file structure as set in Desktop

Hi @mlufkin​ , I'm sorry to hear you've run into issues with this after your upgrade! I did some testing using FME 2020.2.4 to see if I could replicate this issue, but it worked for me. In my testing, I used a single File Copy writer writing out one file and a second writer. It created the folder structure I wanted (I only removed the FILECOPY_1 parts in the output zip folder). These were my settings: How are the writer(s) and Data Download output folder structure configured in your case?

Thanks @mlufkin​! I do see the same behaviour too. If I set the path to a blank value, the file is saved to FILECOPY_1 anyway. It also looks like the Path in Zip File isn't being updated properly when re-publishing a workspace. For example, if you were to change GRIDS to something else, that new name would not be used and GRIDS would still be created in the zip folder. We'll investigate this one more on our end. Thanks for raising this issue! I'll post an update here when we know more. Sorry for this trouble!

I have the same issue now, as we migrated to FME(R) 2021.2.5.0 (20220331 - Build 21816 - WIN64) . I was able to resolve this when publishing manually. When I have removed the "Path in Zip File" parameter and have published the workspace with FME Desktop, the generated ZIP file has no subfolders. But we usually publish the workspaces with a python script and the FME server REST api, see https://docs.safe.com/fme/html/FME_REST/apidoc/v3/index.html#!/user

When I transfer the FME files onto another system and publish them with my scripts, the Path to Zip File parameter is back. Are there any hidden parameters in the REST api I could use?

bye, Nicole

Hi @siennaatsafe​ 

Thanks a lot for your hint!

We have tried to update the settings with a call in Python code but it does not seem to update.

url = '{}/fmerest/v3/repositories/{}/items/{}/detailedservices'.format(fme_server_url, STORME_REPOSITORY_NAME, filename)

payload = "{'detailedServices': {'dataDownload': {'zipLayout': ['']}}}"

r = requests.put(url, data=payload, headers=put_headers_form)

We are getting a Http Status OK, but when we are doing a Get-Request there is no change in the settings.

Whats wrong with it?

Thanks for your help!

kind regards, Nicole

{
  "jobSubmitter": {
    "reader": "SHAPEFILE_1",
    "enabled": true
  },
  "dataDownload": {
    "reader": "SHAPEFILE_1",
    "zipLayout": {},
    "writers": [
      "GENERIC_1"
    ],
    "enabled": true
  },
  "kmlNetworkLink": {
    "enabled": false
  },
  "dataStreaming": {
    "enabled": false
  }
}

Hi @siennaatsafe​ , 

Thanks a lot for the help!

I'm working with Nicole on the project. We finally managed to update the zipLayout  with the following code: 

            url = '{}/fmerest/v3/repositories/{}/items/{}/detailedservices'.format(fme_server_url, STORME_REPOSITORY_NAME, filename)
            rGet = requests.get(url, headers=get_headers)
            detailedservices = rGet.json()
            detailedservices['dataDownload']['zipLayout']= {"TEXTLINE_1": "Response.txt", "TEXTLINE_2": "Response.txt"}
            payload = json.dumps(detailedservices)
            rPut = requests.put(url, data=payload, headers=put_headers_form)
            logging.info(filename + " zipLayout updated for TEXTLINE_1 and TEXTLINE_2")

It was tricky to figure out what the ZipLayout needed to look like. Our experience is that it is only updated if the name in the ZipLayout is the same as the Writer's name. So we specified it for each writer. I hope this helps if anyone else struggles with the same problem.

Best wishes, 

Shari

Excellent - thanks!