make av1 obu parsing optional

[jech]

• Make parsing of the OBU list in AV1Packet optional
• Avoid usage of OBUElements in av1_packet_test.
• Mark ReadFrames as being deprecated.

In Pion, the (*XXXPacket).Unmarshal operation is supposed to be
fast and perform no allocation, as it is sometimes used to check
just a single flag in the packet header. The AV1Packet
implementation breaks this property by parsing the whole list
of OBUs at Unmarshal time, even though it is likely to be unneeded.

Note that I strongly object to this solution, since it makes the
inefficient path (both in space and in time) the default, and thus
will make people write inefficient code. However, if that's the
cost to getting the brain-damange reverted, I won't scream too
loudly.

[codecov]

Codecov Report

Attention: Patch coverage is 90.00000% with 2 lines in your changes are missing coverage. Please review.

Project coverage is 86.19%. Comparing base (a663858) to head (bcbb9ea).

❗ Current head bcbb9ea differs from pull request most recent head ee561cf. Consider uploading reports for the commit ee561cf to get more accurate results

Files	Patch %	Lines
codecs/av1_packet.go	90.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #266      +/-   ##
==========================================
+ Coverage   86.05%   86.19%   +0.14%     
==========================================
  Files          22       22              
  Lines        1742     1746       +4     
==========================================
+ Hits         1499     1505       +6     
+ Misses        206      204       -2     
  Partials       37       37              

Flag	Coverage Δ
go	86.19% <90.00%> (+0.14%)	⬆️
wasm	85.68% <85.00%> (+0.09%)	⬆️

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.

[jech]

I've just made the accessor into a method, so its presence can be tested by callers (by checking whether the packet implements a given interface). I see that H264 and H265 have the same problem, so I've made the name codec-agnostic.