PI Adapter for Structured Data Files
Create a PI Adapter that supports parsing data from simple file formats (CSV, XML, JSON, plain text) and is capable of writing data to EDS, PI, and OCS.
We have started developing a PI Adapter for Structured Data Files.
A problem we have with the current UFL interface is that it cannot read open files, which means that we can only read the content when a batch has finished and the file has closed. This of course means we can't get to "near real time" data ingestion into PI, instead it could be days. So it'd be great if you can consider this for the Adapter - surely we can't be the only ones with this use case!
The PI UFL interface supports serial port data entry, or virtual serial port data entry. We use this to send gas chromatograph lab data to PI for environmental monitoring.
If PI Adaptor for Structured Data Files will be replacing PI UFL interface in the future, then a serial port data transfer method needs to be included among the file formats within PI Adaptor for Structured Data Files, or else something additional needs to be created to support sending serial port data to PI.
Will there be continuing support from OSIsoft for some sort of a method (adaptor, interface, connector, etc.) to send serial port data to PI?
It would be nice if this PI Adapter had a way to run code before or after a file is processed. This would be good for resetting variables to some initial value or printing stuff to the log, but I'm sure that there are other use cases for this.
With the PI Interface For UFL and the PI Connector For UFL, you would have to get clever with the BOS and EOS functions. The latter is complicated further if you used SkipFile in your code, and both are complicated further if you have many different message sections.
Quite a few of the PI Adapter suggestions have the admin comment "Look for more updates on our progress in Q1 2021". Q1 2021 has passed. Any updates to share?
It would be great if this PI Adapter had much more native support for XML files. That way, we can browse and parse XML files like a tree as they were intended, rather than as a sequence of lines.
I still think that OSIsoft should consider leveraging Power Query, which is a language that already supports many types of data sources and not just text-based ones. See this suggestion: