Clarification on frequency code of SBAS

[bobzwik]

Hello,
I've been digging into the code to figure out which frequencies of each constellation is used depending on the pos1-frequency parameter.

I found code2idx, and which each subfunction (for each constellation), I can understand which observation code is associated with the L1, L2, L5, L6, L7 options of pos1-frequency.

I've however found a mismatch. The comment for code2idx says that the L5 frequency of SBAS should return the code 2, but code2freq_SBS actually returns 1 for L5. This means that the SBAS L5 band is used when selecting L1+L2 for pos1-frequency, right? Is this intentional? Is the code wrong, or the comment?

Thanks in advance for the clarification!

[ourairquality]

Also relevant is the definition of seliflc() which uses index 2 for SBAS. This area of the code was changed recently, probably to pivot to L1/L5 low cost receivers, and BDS now has good 1P 5P signals, and perhaps code2freq_SBS() was missed.

Rather than patching code2freq_SBS(), an alternatively would be to map all the L5 signals to index 1 and then seliflc() becomes a NOP and might even be removed. But then that leaves the naming of the frequency indices confusing.

Fwiw I my own wip branch these are just frequency indices, labelled F1 to F6, and the rinex signal bands can be arbitrarily mapped to frequency indices per systems with a runtime config option. seliflc() then just uses the first two indices across all systems.