Reputation: 1397
There are lots of demos online where slicing is carried out on blobs of the myblobcontainer/{Year}/{Month}/{Day}
format or similar (i.e. https://azure.microsoft.com/en-gb/documentation/articles/data-factory-scheduling-and-execution/).
Now clearly this will allow for really easy slicing of the data as the parameters for year month and day have really clearly been defined.
What I have though, are files that more like this:
myblobcontainer/log_20151231_144229.csv
which is clearly YYYYMMDD_HHMMSS
.
I want to process my files hourly, not re-process anything, and ideally not have to mess around too much restructuring my blobs.
Does anyone have any idea how I can "read" only these files that fall inside my hour slices?
Upvotes: 0
Views: 632
Reputation: 205
The folderPath in Blob Dataset can work as a path prefix. So you could set the folderPath like "$$Text.Format('myblobcontainer/log_{0:yyyyMMdd}', WindowStart)"
, and all files satisfy the prefix will be copied.
Upvotes: 1