Add RTP Stats to stats report

[k0nserv]

This PR implements inbound and outbound RTP stats according to the
specification, with some
limitations.

Todo

• Inbound RTP
• Outbound RTP
• Tests

[codecov]

Codecov Report

Merging #225 (ec4ac28) into master (5614253) will increase coverage by 0.22%.
The diff coverage is 67.92%.

@@            Coverage Diff             @@
##           master     #225      +/-   ##
==========================================
+ Coverage   56.05%   56.27%   +0.22%     
==========================================
  Files         500      500              
  Lines       46523    47034     +511     
  Branches    12714    12779      +65     
==========================================
+ Hits        26077    26470     +393     
- Misses       9920     9935      +15     
- Partials    10526    10629     +103     

Impacted Files	Coverage Δ
interceptor/src/report/receiver/mod.rs	100.00% <ø> (ø)
webrtc/src/stats/mod.rs	43.47% <0.00%> (-11.80%)	⬇️
webrtc/src/peer_connection/peer_connection_test.rs	56.12% <57.14%> (+1.53%)	⬆️
interceptor/src/stats/interceptor.rs	74.38% <73.05%> (+3.74%)	⬆️
...tc/src/peer_connection/peer_connection_internal.rs	51.29% <79.16%> (+3.73%)	⬆️
interceptor/src/stats/mod.rs	81.18% <81.31%> (+11.42%)	⬆️
.../src/rtp_transceiver/rtp_sender/rtp_sender_test.rs	51.79% <83.33%> (+4.17%)	⬆️
...c/src/track/track_local/track_local_static_test.rs	54.42% <83.33%> (+1.03%)	⬆️
.../transport_layer_nack/transport_layer_nack_test.rs	82.00% <100.00%> (+0.75%)	⬆️
dtls/src/curve/named_curve.rs	72.50% <0.00%> (-15.38%)	⬇️
... and 15 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[k0nserv]

I'll explore extending this test to also have some RTCP components to it

[k0nserv]

Will tackle Remote{Inbound,Outbound}RTP in another PR

[lolgesten]

Some questions

[lolgesten]

This looks strange. If we have this:

packet_id: 123
bitfield: `00000000 00000101`

We expect the iterator to produce 123, 124, and 126.

But it looks like self.bitfield is "destroyed" on the first iteration which means the result would be 123, 124 – and missing the last value.

[k0nserv]

This specific line zeroes out the bit in question which ensure we eventually hit the loop condition.

Keep in mind that we are within a next method here, as such we only run until we can produce a value, subsequent calls to next should produce another value if we do indeed have more values to produce.

I'll write a test with a gap like you pointed out to verify

[k0nserv]

Thanks for the review!

[k0nserv]

This specific line zeroes out the bit in question which ensure we eventually hit the loop condition.

Keep in mind that we are within a next method here, as such we only run until we can produce a value, subsequent calls to next should produce another value if we do indeed have more values to produce.

I'll write a test with a gap like you pointed out to verify

[algesten]

I prefer the alternative impl for NackIter, but other than that. Good to go!