NACK responder: bypass auxiliary SSRCs

[aalekseevx]

Description

The current implementation of RTX in the NACK interceptor implies that there may be packets with different SSRCs in a single RTP stream (stream in terms of the interceptor interface), e.g., with SSRCRetransmission or (more importantly for this PR) SSRCForwardErrorCorrection. These packets should not be stored in the buffer or used for responding to NACKs.

I suggest skipping such packets in the responder interceptor. This might be a breaking change for some exotic use cases, but I can't think of any sensible scenario where the NACK interceptor is used to resend packets with a non-main SSRC. Therefore, I believe this is a reasonable default.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.34%. Comparing base (c06f448) to head (4cda00c).
Report is 1 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #310      +/-   ##
==========================================
- Coverage   71.38%   71.34%   -0.05%     
==========================================
  Files          79       79              
  Lines        4526     4530       +4     
==========================================
+ Hits         3231     3232       +1     
- Misses       1158     1160       +2     
- Partials      137      138       +1     

Flag	Coverage Δ
go	71.34% <100.00%> (+0.09%)	⬆️
wasm	69.38% <100.00%> (-0.11%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[aalekseevx]

Thank you @nils-ohlmeier :)