Avoid usage of getSpanWithoutDefault when possible.

[bogdandrutu]

We check for isValid everywhere so this will not be a performance hit, explicitly because the DefaultSpan.getInvalid().getContext().isValid() will be memoized.

[codecov]

Codecov Report

Merging #1690 into master will increase coverage by 0.10%.
The diff coverage is 73.33%.

@@             Coverage Diff              @@
##             master    #1690      +/-   ##
============================================
+ Coverage     85.36%   85.46%   +0.10%     
+ Complexity     1385     1383       -2     
============================================
  Files           164      164              
  Lines          5445     5435      -10     
  Branches        562      560       -2     
============================================
- Hits           4648     4645       -3     
+ Misses          590      587       -3     
+ Partials        207      203       -4     

Impacted Files	Coverage Δ	Complexity Δ
...xt/src/main/java/io/opentelemetry/Application.java	0.00% <0.00%> (ø)	0.00 <0.00> (ø)
...ntelemetry/trace/propagation/HttpTraceContext.java	96.03% <100.00%> (-0.08%)	27.00 <0.00> (-2.00)
...opagation/B3PropagatorInjectorMultipleHeaders.java	100.00% <100.00%> (+8.33%)	4.00 <0.00> (ø)
.../propagation/B3PropagatorInjectorSingleHeader.java	100.00% <100.00%> (+4.34%)	5.00 <0.00> (ø)
...extensions/trace/propagation/JaegerPropagator.java	87.50% <100.00%> (+1.08%)	26.00 <0.00> (ø)
...tensions/trace/propagation/OtTracerPropagator.java	82.75% <100.00%> (+4.63%)	10.00 <0.00> (ø)
...ava/io/opentelemetry/sdk/trace/SpanBuilderSdk.java	95.79% <100.00%> (ø)	39.00 <0.00> (ø)

---

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 d9bd619...57b1ce9. Read the comment docs.

[jkwatson]

It looks like there's only one usage left after this. Can we get rid of that one, as well, and get rid of the method altogether?

[bogdandrutu]

@jkwatson you read my mind, but that one usage left is a bit different. I will followup on that soon.

[jkwatson]

Thanks!

[anuraaga]

Was looking forward to this, thanks!