feat: make metrics build specific

[thugrock7]

Description

CPU and memory metrics which are added, making them linux specific. For other platforms using no-op

Testing

Tested on Mac, linux

[codecov]

Codecov Report

Attention: Patch coverage is 22.38806% with 52 lines in your changes missing coverage. Please review.

Project coverage is 58.61%. Comparing base (e8225ec) to head (0572059).

Files	Patch %	Lines
...entation/opentelemetry/internal/metrics/metrics.go	0.00%	27 Missing ⚠️
...on/opentelemetry/internal/metrics/linux_metrics.go	37.50%	23 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #237      +/-   ##
==========================================
- Coverage   59.89%   58.61%   -1.29%     
==========================================
  Files          70       71       +1     
  Lines        2364     2375      +11     
==========================================
- Hits         1416     1392      -24     
- Misses        863      903      +40     
+ Partials       85       80       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[varkey98]

done reviewing, missed this earlier

[varkey98]

@shashank11p @puneet-traceable pls take a look as well

[varkey98]

Adding some context here: The reason is that /proc/pid/stat file is only available on linux and trying to read from that file on other platforms results in an error

[varkey98]

@thugrock7 Pls test it on windows as well

[puneet-traceable]

no tests for this function? It's fine for this but in general for testing such scenarios

1. either file should become a parameter to the function and then you can create custom files to test.
2. can pass an io.Reader to the function to read file. (*os.File) implements an io.Reader. During test you can use a mock version.

[thugrock7]

Tested running this by http-server/client scenario on linux, Mac, windows. Works fine
Also confirmed the exported metrics on TPA in linux, mac os.