Feature/pulse schedule 3to4 #33
Conversation
DaanVanVugt
left a comment
DaanVanVugt
left a comment
There was a problem hiding this comment.
code looks fine, but what about 3.39 -> 3.41? there is a bit of a gap in the conversion logic, right? 3.41 -> 4?
| destination IDS will remain empty. | ||
| .. [#ps3to4] In Data Dictionary 3.39.0 and older, all dynamic quantities in the | ||
| ``pulse_schedule`` IDS had their own time array. In DD 4.0.0 this was |
There was a problem hiding this comment.
what about 3.40 etc? was it not restructured at that point?
There was a problem hiding this comment.
Correct, but IO requested to make the conversion logic available for DD3.x to 4.x.
what about 3.39 -> 3.41? there is a bit of a gap in the conversion logic, right?
3.39 -> 3.41 follows regular conversion logic and will not resample dynamic quantities. It would be trivial to update the logic to execute when converting from DD <=3.39.0 to DD >= 3.40.0, though. @olivhoenen any thoughts?
3.41 -> 4?
Nothing needs to be done w.r.t. the time bases between these DD versions.
There was a problem hiding this comment.
That's true, on the one hand we should not expect imas-python to have to fix every single non-backward compatible changes to the DD via ad-hoc conversion functions (but it's nice to have that option). But in this case, you've implemented the function already (with DD4 in mind but the same transformation applies for DD > 3.39.0), so I think there is no harm extending the range.
There was a problem hiding this comment.
Clear! The conversion now also applies when going to a DD version >= 3.40.0.
| provenance.node[0].reference[0].name = source.value | ||
|
|
||
|
|
||
| def _pulse_schedule_3to4( |
There was a problem hiding this comment.
should these go in a separate file perhaps? or do we do that when we add another large migration?
There was a problem hiding this comment.
We could, but these _pulse_schedule_3to4* functions also depend on some of the other functions and globals from this module (e.g. _get_target_item, and logger).
Separating them into separate files is slightly messy with circular imports etc. At this point that is not really worth it IMO :)
olivhoenen
requested review from
SimonPinches
and removed request for
SimonPinches
3 participants
DD3 to DD4 conversion mechanism for the pulse schedule IDS.
Fixes #21.