Exporters

Contents:

Exporters are for use with loaders that have an unload() method, and are used to deliver time series to the loaders’ input for unloading (generating SHEF text from the time series).

Loaders can unload time series without exporters, but the time series to unload must be available on the SHEF parser’s (and thus the loader’s) input device in the format used by the loader. If the time series are available in this format, the using run_shef_parser --loader <loader-spec> --unload will suffice. However, in most contexts, some extra functionality is needed to retrieve the desired time series (with the desired time window) from the data store, and format it appropriately for the loader.

Workflow Using an Exporter

Instead of running from a command line, exporters are used from within Python - either in scripts or in interactive sessions. The workflow generally follows the following process:

Import the exporter
Instantiate the exporter with its specific parameters
Specify or connect to a data store (if not already done in previous step)
Set the exporter’s time window by assigning Optional[datetime.dateime] objects to its start_time and end_time properties
Set the exporter’s output to a file or other ouptut using the exporter’s set_output(...) method
Call the exporter’s export(...) method, passing a time series identifier (or a group identifier if the exporter supports groups)
Repeat steps 3-6 as necessary
Delete the exporter (possibly by exiting the script). The exporter should release any resources is has.

A work flow might look like:

from datetime import datetime, timedelta
from shef.exporters.abc_exporter import AbcExporter
exporter = AbcExporter(<abc-data-store>, <user-credentials>)
exporter.end_time = datetime.now()
exporter.start_time = exporter.end_time - timedelta(days=7)
exporter.set_output("/path/to/output_file.shef")
with open("time_series_to_output.txt") as f:
   for time_series in [line.strip() for line in f]:
      exporter.export(time_series)
exit()