[PoC] Change Processor.OnEmit to return record passed to next registered processor

[pellared]

Towards #5219

Related to open-telemetry/opentelemetry-specification#4065

Alternative proposal: #5469

Changes Processor.OnEmit to:

OnEmit(ctx context.Context, record Record) (Record, error)

Pros:

• MAY be seen as acceptable by spec (however, probably the spec should still be refined)
• Almost no performance overhead, zero-allocation is possible. There are more operations on the stack.

Cons:

• In my opinion, ugly design; more: Logs: Allow the registered processor to be an independent pipeline opentelemetry-specification#4010 (comment)
• Inconsistent API design with span processor

Personally, I dislike the design of this proposal as I outlined here open-telemetry/opentelemetry-specification#4010 (comment)

Processor benchmark results:

$ benchstat old.txt new.txt
goos: linux
goarch: amd64
pkg: go.opentelemetry.io/otel/sdk/log
cpu: Intel(R) Core(TM) i9-10885H CPU @ 2.40GHz
                                    │   old.txt    │               new.txt                │
                                    │    sec/op    │    sec/op     vs base                │
Processor/Simple-16                   527.3n ± 19%   574.0n ± 15%   +8.86% (p=0.022 n=10)
Processor/Batch-16                    1.013µ ±  6%   1.089µ ±  4%   +7.45% (p=0.003 n=10)
Processor/ModifyTimestampSimple-16    573.2n ± 12%   646.1n ±  5%  +12.72% (p=0.005 n=10)
Processor/ModifyTimestampBatch-16     1.036µ ±  4%   1.106µ ±  8%   +6.71% (p=0.050 n=10)
Processor/ModifyAttributesSimple-16   616.2n ± 12%   688.8n ±  8%  +11.77% (p=0.035 n=10)
Processor/ModifyAttributesBatch-16    1.085µ ±  9%   1.158µ ±  3%   +6.78% (p=0.024 n=10)
geomean                               772.2n         841.8n         +9.02%

                                    │  old.txt   │              new.txt              │
                                    │    B/op    │    B/op     vs base               │
Processor/Simple-16                   416.0 ± 0%   416.0 ± 0%       ~ (p=1.000 n=10)
Processor/Batch-16                    622.5 ± 2%   616.0 ± 1%       ~ (p=0.137 n=10)
Processor/ModifyTimestampSimple-16    417.0 ± 0%   416.0 ± 0%       ~ (p=0.656 n=10)
Processor/ModifyTimestampBatch-16     618.5 ± 3%   609.0 ± 2%       ~ (p=0.137 n=10)
Processor/ModifyAttributesSimple-16   465.0 ± 0%   465.0 ± 0%       ~ (p=1.000 n=10)
Processor/ModifyAttributesBatch-16    652.5 ± 2%   651.0 ± 2%       ~ (p=0.781 n=10)
geomean                               522.1        519.5       -0.51%

                                    │   old.txt    │               new.txt               │
                                    │  allocs/op   │ allocs/op   vs base                 │
Processor/Simple-16                   1.000 ± 0%     1.000 ± 0%       ~ (p=1.000 n=10) ¹
Processor/Batch-16                    0.000 ± 0%     0.000 ± 0%       ~ (p=1.000 n=10) ¹
Processor/ModifyTimestampSimple-16    1.000 ± 0%     1.000 ± 0%       ~ (p=1.000 n=10) ¹
Processor/ModifyTimestampBatch-16     0.000 ± 0%     0.000 ± 0%       ~ (p=1.000 n=10) ¹
Processor/ModifyAttributesSimple-16   2.000 ± 0%     2.000 ± 0%       ~ (p=1.000 n=10) ¹
Processor/ModifyAttributesBatch-16    1.000 ± 0%     1.000 ± 0%       ~ (p=1.000 n=10) ¹
geomean                                          ²               +0.00%                ²
¹ all samples are equal
² summaries must be >0 to compute geomean


                        │   old.txt   │            new.txt            │
                        │   sec/op    │   sec/op     vs base          │
BatchProcessorOnEmit-16   385.9n ± 8%   385.5n ± 7%  ~ (p=0.493 n=10)

                        │    old.txt     │               new.txt                │
                        │      B/s       │     B/s       vs base                │
BatchProcessorOnEmit-16   1047.85Mi ± 9%   79.17Mi ± 6%  -92.44% (p=0.000 n=10)

                        │  old.txt   │            new.txt             │
                        │    B/op    │    B/op     vs base            │
BatchProcessorOnEmit-16   0.000 ± 0%   0.000 ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal

                        │  old.txt   │            new.txt             │
                        │ allocs/op  │ allocs/op   vs base            │
BatchProcessorOnEmit-16   0.000 ± 0%   0.000 ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 84.6%. Comparing base (047df28) to head (66cf41b).
Report is 263 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##            main   #5470     +/-   ##
=======================================
- Coverage   84.6%   84.6%   -0.1%     
=======================================
  Files        267     267             
  Lines      17825   17826      +1     
=======================================
- Hits       15086   15085      -1     
- Misses      2426    2428      +2     
  Partials     313     313             

Files with missing lines	Coverage Δ
sdk/log/batch.go	100.0% <100.0%> (ø)
sdk/log/logger.go	100.0% <100.0%> (ø)
sdk/log/simple.go	100.0% <100.0%> (ø)

... and 1 file with indirect coverage changes