fix(exporter-zipkin): correct status tags names

[t2t2]

Which problem is this PR solving?

As per OpenTelemetry to Zipkin Transformation spec:

Span Status MUST be reported as a key-value pair in tags to Zipkin, unless it is UNSET.
In the latter case it MUST NOT be reported.

The following table defines the OpenTelemetry Status to Zipkin tags mapping.

Status	Tag Key	Tag Value
Code	otel.status_code	Name of the code, either OK or ERROR. MUST NOT be set if the code is UNSET.
Description	error	Description of the Status. MUST be set if the code is ERROR, use an empty string if Description has no value. MUST NOT be set for OK and UNSET codes.

Short description of the changes

Corrected the tag names (ot. -> otel.), check for unset status

For reference: zipkin exporter in otel-java

[codecov]

Codecov Report

Merging #2519 (06d880b) into main (0c7f1c3) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #2519   +/-   ##
=======================================
  Coverage   93.23%   93.23%           
=======================================
  Files         137      137           
  Lines        5042     5043    +1     
  Branches     1066     1067    +1     
=======================================
+ Hits         4701     4702    +1     
  Misses        341      341           

Impacted Files	Coverage Δ
...ges/opentelemetry-exporter-zipkin/src/transform.ts	100.00% <100.00%> (ø)
...ckages/opentelemetry-exporter-zipkin/src/zipkin.ts	100.00% <100.00%> (ø)

[johnbley]

https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/trace/sdk_exporters/zipkin.md suggests that for a status of ERROR, the error messages should be in an error tag/attribute. It specifically calls out that it overrides the generic status mapping rule. Would like to see a fix for that and a test covering the error case. See https://github.com/open-telemetry/opentelemetry-swift/pull/232/files for a recent swift change I did that was similar.

[t2t2]

@johnbley Updated to be based on the zipkin specific doc. Should be already covered by should map OpenTelemetry SpanStatus.message to a Zipkin tag test case.

---

Seems like unrelated rounding fail in test?

[dyladan]

Most likely unrelated. I'm running again to see if it is consistent

[johnbley]

LGTM now, thanks!