Add Zipkin API V2 Reporter

[rmfitzpatrick]

Expands span reporter selection by breaking out Reporter into base
QueueReporter class and using reporting method-specific ThriftReporter
and ZipkinV2Reporter. Also breaks LocalAgentSender into two classes
including LocalAgentReader for ThriftReporter only usage of
TBufferedTransport.

These changes allow users to continue reporting spans oob to the local
jaeger-agent but with the option of reporting to Zipkin backend.

In the spirit of
jaegertracing/jaeger-client-java#399
and
jaegertracing/jaeger-client-go#310

Signed-off-by: Ryan Fitzpatrick rmfitzpatrick@signalfx.com

[rmfitzpatrick]

Fixing the issues found in integration.

edit: Issues resolved.

[codecov]

Codecov Report

Merging #197 into master will increase coverage by 0.31%.
The diff coverage is 92.3%.

@@            Coverage Diff             @@
##           master     #197      +/-   ##
==========================================
+ Coverage   94.75%   95.06%   +0.31%     
==========================================
  Files          25       26       +1     
  Lines        1773     1926     +153     
  Branches      224      256      +32     
==========================================
+ Hits         1680     1831     +151     
+ Misses         60       54       -6     
- Partials       33       41       +8

Impacted Files	Coverage Δ
jaeger_client/local_agent_net.py	95.65% <100%> (+0.3%)	⬆️
crossdock/server/endtoend.py	100% <100%> (ø)	⬆️
jaeger_client/config.py	93.15% <90%> (+2.46%)	⬆️
jaeger_client/zipkin_v2.py	91% <91%> (ø)
jaeger_client/reporter.py	95.31% <94.87%> (+3.43%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 00d3d4a...ce6fb0b. Read the comment docs.

[rmfitzpatrick]

I've increased test coverage for all proposed changes, including some potentially overreaching reporter tests. Apologies for the churn and any feedback would be greatly appreciated.

[rmfitzpatrick]

@black-adder and @yurishkuro, in light of jaegertracing/jaeger#925 understandably preventing these changes from being accepted, if a short term requirement is a thrift over http reporter in the client do you think it's reasonable for me to:

1. Attempt to add the append and flush methods to Preliminary refactor for supporting spans over HTTP #186 for that to land.
2. Propose an HTTPSender, while using the LocalAgentReader distinction proposed in this PR?

[black-adder]

@rmfitzpatrick I think that'd be great.