from_gregorian but in GPST

[gwbres]

Hello @ChristopherRabotin,

I would be interested in having an Epoch construction method similar to Epoch::from_gregorian_utc but in TimeScale::GPST.

from_gregorian_utc is the best thing to use when all data fields are already known.

Such feature would solve the requirement I face in two of my parsers:

• in RINEX when parsing NAV ephemeris - the SV broadcasts Epoch expressed in their timescale (GPST for example)
• in SP3, all Epochs are expressed in a unique GNSS time scale

[ChristopherRabotin]

Yes, that's a good idea. Is this something we can implement in version 4? It sounds like it would be quite a bit easier in there. What do you think?

…

On Wed, Aug 30, 2023, 02:28 gwbres ***@***.***> wrote: Hello @ChristopherRabotin <https://github.com/ChristopherRabotin>, I would be interested in having an Epoch construction method similar to Epoch::from_gregorian_utc but in TimeScale::GPST. from_gregorian_utc is the best thing to use when all data fields are already known. Such feature would solve the requirement I face in two of my parsers: - in RINEX when parsing NAV ephemeris - the SV broadcasts Epoch expressed in their timescale (GPST for example) - in SP3, all Epochs are expressed in a unique GNSS time scale — Reply to this email directly, view it on GitHub <#257>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABEZV2FKQUUO3W2KHEEVI3LXX32RVANCNFSM6AAAAAA4EEI7RA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[gwbres]

It sounds like it would be quite a bit easier in there. What do you think?

Yes, definitely easier if the new approach gets finalized.

I don't have a need for this immediatly either

[ChristopherRabotin]

Would this make more sense in #229 to avoid proliferation of more initializers?

[gwbres]

I agree that if Epochs can be built from durations directly, it's okay to get rid of all the from_xxx_seconds,nanoseconds .
This will reduce code quantity and improve maintainability. (you did great in cleaning up the Epoch module BTW).

Now we have Epoch::from_gregorian_str(x:x:x:x:x $TS) which is very convenient to parse formatted epochs in specific timescales. Ideally you want a mirror operation to go with that

[ChristopherRabotin]

Hi Guillaume,
This feature is already available it seems, and tested in the regression_test_gh_288 specifically with the GPST from str. I will add another test on the branch for 331 now.