Problem after assemble job

[numnational]

Hy all.

I want to redirecting 2 or more files who are "tagged" by "Switch Meta Data" in different directions.

I tag "Emag" and "Original". After this all is OK, i can redirecting to 2 files where i want with the 2 different tag.

My problem is : the first pdf is without treatment so it is rapid and the second is with several treatment, so it is slowly. But the first file must waiting the second, so i use for this an assemble job. After this assemble job, i found a folder, an inside my 2 files. But the tag create by "Switch Meta Data"is lost. One time i have "Emag" on the 2 files and other time i have "Original" on the 2 files... What happenning ? Can i keep my original tag ????

The flow is here (of course the flow is empty any complex treatment, only for testing) : https://www.dropbox.com/s/rqy2yoq707y4f ... sflow?dl=0

Thanks.

[sander]

In your supplied flow Merge metadata is set to No.

Try Yes

Edit: Sorry, this can be found in your assemble job.

[numnational]

Thanks for your answer but sorry don't work... I have always the same problem .

In switch documentation i can see about merge metadata in "Assemble Job" : Private data and External datasets will all be retained. If multiple jobs have private data or external datasets with the same tag, a dataset or value will be arbitrarily chosen.

So, is it impossible what i want ?????

[freddyp]

"Assemble job" by definition has to merge the properties and metadata of multiple jobs. In the flow in your Dropbox I see that you route the job in two directions after the assemble using a condition. May I assume that you do not want to route the folder, but the files inside the folder? If so, your "new" job files will inherit the properties of the job folder and they will behave in the same way, which is not what you want. Correct?

[numnational]

Yes you understand what i want, this is correct and i don't want this...

Maiby can I create one folder by file ? or have you a idea ?

[freddyp]

The quick and dirty solution is to do a "Pack job" before "Assemble job" and "Unpack job" after "Job dismantler" (which comes after "Assemble job" to take the assembled packed jobs back out of the folder).

"Pack job" keeps all the properties and metadata of a job, so by packing the job first you get a sort of ZIP file with everything in it. After "Unpack job" all the information is still there.

[freddyp]

I found this use case interesting and generic enough so I am turning a solution for this problem into a free app on the Enfocus Appstore. The packing and unpacking will work, but I find it is too much processing for this type of problem. The app will have the following properties:

• Group identifier
• Number of files
• Timer interval (for checking if a group is complete)
• Orphan timeout

The "Group identifier" and "Number of files" will allow the use of variables of course.

The behavior is that it will hold jobs until a group is complete and then simply let the members of the group pass without assembling them, thus ensuring that the properties and metadata of each job are kept and at the same time ensuring that the jobs are released very close together. Orphaned jobs (jobs that belong to a group that does not reach the required number of files after the timeout) go to the Error connection.

Anything else it should do?

I am thinking of calling it "Hold job group". Any better suggestions?

Expect publication on the Enfocus Appstore by the beginning of next week. I will post it here when it is available.

[numnational]

OK Freddy great news I waiting this app next week thanks lot !

I have another flow who make what i want but i have a lot of rename, so it's not clean as i want.

Thanks.

[freddyp]

New insight as a result of another question that came up in a different context, and that invalidates the need for an app: set "Strip unique name" to "No" in "Assemble job". The result is still a folder but the files inside keep their Switch unique ID. When you follow on with "Job dismantler" the files inside the folder become jobs again and Switch will still recognize them based on their unique ID and the link with all the properties and metadata is still there.

The app would save you one step ("Job dismantler") but if you can get to the same result with standard functionality, that is better of course.

I had never had a use case for this which is why I did not immediately think of doing it that way. And you can solve your problem now, you do not have to wait till the beginning of next week

[numnational]

Hi Freddy, i was in long weekend

Sorry but it doesn't work. Of course i try with "no" in 'strip unique name" in Assemble job, before than your submit, but don't work.

You can check my flow here if you want :
https://www.dropbox.com/s/81u3c89p15vb9 ... sflow?dl=0

Maiby i make something wrong... ????

Thanks...

[freddyp]

Indeed, you are right. Well, then I will publish the app anyhow.

[numnational]

Thanks

[numnational]

Hy Fredd, any news about the new tools ? I 'm waiting for working

[freddyp]

I was out of the office for more than 2 weeks. I am slowly catching up with all the things I have to do. The code is finished, but I still have to write the documentation before it can be published as an app. Tomorrow? At the latest on Thursday.

[numnational]

Thank you Freddy, i'm waiting no problem