I would like to write batch triggers against a file drop to a UFL interface. The batch should start with the first timepoint from a file and end with the last timepoint, and use a portion of the file name as the “Product” in the batch. More specifically, if possible, I would like to trigger a batch with three sub-batches against files that will always be sent to the interface in the same order.
For example, if I have the 3 files:
I would want a batch with sub-batch names operation1, operation2, and operation3, and “FileName” would be captured in a pi tag as well.
In the UFL documentation I see that you can use PFN to count the file name as a line read, but as far as I can tell you cannot easily associate a timestamp with it unless perhaps it is in the file name? I would like to know if there is a way to avoid putting the timestamp in the file name, as I would like to keep the names as short as possible since they will be manually saved by users.
Unfortunately the data sets we are using do not have any points in the realm of “time elapsed” or “sample number” that would assist in easily allowing triggers to be written against the values provided in the file. These files will also be dropped into the UFL several hours after the first data point in the file is collected.
I may be asking for too much here, but I thought I would see if anyone has made similar attempts.