topic Re: Auto Create Folder in QlikView

Auto Create Folder

KHSDM — Tue, 25 Sep 2018 09:26:25 GMT

Dear all,

I notice that there is this new behavior in 12.20SR6. QlikView will auto create folder if it does not exist. Previously there is no such behavior. Job will fail if the folder does not exist. Now QlikView created many unnecessary folders due to the typo in the path and messed the whole folder structure.

Anyone else facing the same issue?

Re: Auto Create Folder

KHSDM — Wed, 03 Apr 2019 09:20:19 GMT

BUMP this thread... any idea if there is any setting that I can change to keep the old behaviour where it won't create the folders automatically?

Re: Auto Create Folder

Brett_Bleess — Wed, 15 May 2019 20:45:37 GMT

To which folder are you referring here, need further details in order to have a look at things.

Regards,
Brett

Re: Auto Create Folder

KHSDM — Thu, 16 May 2019 03:05:24 GMT

Hi Brett,

I'm referring to the folder when we use the "STORE" function in the script to store the QVD. Previously, the script will failed if the folder does not exist. But now, it will automatically create the folder if it does not exist.

Re: Auto Create Folder

Vegar — Thu, 16 May 2019 04:19:23 GMT

I have not seen such an setting in QlikView nor Qlik Sense. I guess you will need to manual check before storing a file.

Sub storeqvd (_table, _folder)
if FileTime('$(_folder)')
Store [$(_table)] into [$(_folder)/$(_table).qvd];
Else
Trace $(_folder) does not exist;
End if
End sub

MyTable:
Load *
From source;
Call storeqvd('MyTable', 'C:/MyFolder/')

Re: Auto Create Folder

Brett_Bleess — Tue, 25 Jun 2019 21:07:23 GMT

Vegar is on the only track available, here is the Help on things:

https://help.qlik.com/en-US/qlikview/November2018/Subsystems/Client/Content/QV_QlikView/Scripting/ScriptRegularStatements/Store.htm

It basically seems to indicate the path must be valid, i.e. the folder has to exist etc., so the only way I would see to automate things would be some sort of IF statement in the load script along with a variable etc. along the lines of Vegar's solution I think.

Regards,
Brett