Version 0.13 Updates

Version 0.13.0

Readers

Implemented the “readers” interface to allow better extraction for specific data types. The original read_bufr() interface was kept for backwards compatibility and works as before. (#88)

New code should use the readers, which can be activated via the reader keyword argument in the read_bufr() function. The following readers are available:

generic: extracts arbitrary BUFR keys with a hierarchical collector. It has the same functionality as the original read_bufr() function, with the flat=False option (set by default).

flat: extracts whole BUFR messages/subsets as flat records. It has the same functionality as the original read_bufr() function, with the flat=True option.

The following readers are experimental and may change in future releases:

synop: extracts synop-like data from BUFR using pre-defined parameters. A parameter is a high-level concept in pdbufr. It was introduced to overcome the problem that the same quantity can be encoded in BUFR in multiple ways. When using parameters like “t2m” we do not need to know the actual encoding, but the desired value is automatically extracted for us. Another advantage is that we can easily extract the observation periods, levels and units for each parameter, which is simply not possible with the generic reader.

temp: extracts temp-like data from BUFR using pre-defined parameters.