[dhairyakhandelwal880-dotcom]

Issue number: closes #8796

Summary

Changes

• Add math.isfinite(float(value)) checks in MetricManager.add_metric (aws_lambda_powertools/metrics/base.py), AmazonCloudWatchEMFProvider.add_metric (aws_lambda_powertools/metrics/provider/cloudwatch_emf/cloudwatch.py), and DatadogProvider.add_metric (aws_lambda_powertools/metrics/provider/datadog/datadog.py).
• Warn with PowertoolsUserWarning and skip non-finite values (NaN, inf, -inf) per ADR 0006, preserving all other valid metrics in the flush.
• Add functional tests verifying warning emission and skipping of non-finite values across Metrics, AmazonCloudWatchEMFProvider, SingleMetric, and DatadogMetrics.

User experience

Before this change, passing a NaN or infinite value to add_metric (e.g. from division by zero in an average calculation) passed the isinstance(value, numbers.Number) check. At flush time, json.dumps emitted bare NaN / Infinity, corrupting the entire EMF line for all metrics in that flush into invalid JSON.

After this change, non-finite values are caught with a clear PowertoolsUserWarning, skipped, and all other valid metrics in the flush are serialized and published successfully without breaking the EMF line.

Verification

• Ran pytest on functional test suites:
  • tests/functional/metrics/required_dependencies/test_metrics_cloudwatch_emf.py: 102 passed
  • tests/functional/metrics/datadog/test_metrics_datadog.py: 35 passed
• Validated formatting and linting:
  • ruff check --cache-dir /tmp/ruff_cache aws_lambda_powertools/metrics/ tests/functional/metrics/: 0 errors
  • ruff format --check --cache-dir /tmp/ruff_cache aws_lambda_powertools/metrics/ tests/functional/metrics/: 0 modifications
• Verified with AI assistant and personally validated all test runs.

---

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[leandrodamascena]

Thanks for the PR @dhairyakhandelwal880-dotcom. This one is already fixed in #8797, which was linked on the issue and approved before yours came in, so I'm closing this as a duplicate. Your #8814 is reviewed and good, see there. Next time, a quick look at the issue for an existing PR saves you the work.