feat: add show_sample_rows tag to override PII-based sample hiding

[joostboon]

Summary

• Adds a new show_sample_rows tag that is the inverse of the existing PII tag behavior
• When enable_samples_on_show_sample_rows_tags: true is set (default: false), models or columns tagged with show_sample_rows will have their samples shown even when PII tags would otherwise suppress them
• disable_test_samples at the test level still takes full precedence over this override

Changes

• get_config_var.sql: Added enable_samples_on_show_sample_rows_tags (default: false) and show_sample_rows_tags (default: ["show_sample_rows"]) config vars
• is_show_sample_rows_table.sql (new): Macro that checks if a model's tags match any show_sample_rows_tags; mirrors is_pii_table.sql
• get_pii_columns_from_parent_model.sql: Skips columns tagged with show_sample_rows when building the PII columns list
• test.sql: Wraps the is_pii_table / should_disable_sampling_for_pii checks so they are bypassed when the model has a show_sample_rows tag

Test plan

• Tag a model with both pii and show_sample_rows, set both feature flags to true — confirm samples are returned
• Tag a column with both pii and show_sample_rows, both flags enabled — confirm column appears in samples
• Set enable_samples_on_show_sample_rows_tags: false with a show_sample_rows-tagged model — confirm PII hiding still applies
• Only disable_samples_on_pii_tags: true, no show_sample_rows tag — confirm existing behavior unchanged
• Both flags enabled, model tagged show_sample_rows, test has meta.disable_test_samples: true — confirm samples still hidden
• Run existing PII sampling tests: test_sampling_pii.py, test_column_pii_sampling.py, test_disable_samples_config.py

🤖 Generated with Claude Code

Summary by CodeRabbit

• New Features

  • Configurable tags and a visibility check to allow overriding PII-based sampling restrictions.

• Bug Fixes

  • Anomaly descriptions now list anomalous dimension values and summarize large groups.
  • Decimal serialization handles special numeric values (Infinity/NaN) correctly.

• Tests

  • Added integration tests validating anomaly description behavior for few vs many failures.

• Chores

  • Added .claude/ to ignore patterns.

[github-actions]

👋 @joostboon
Thank you for raising your pull request.
Please make sure to add tests and document all user-facing changes.
You can do this by editing the docs files in the elementary repository.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: bdd75e18-3a14-42dc-9c1b-7d2783b63688

📥 Commits

Reviewing files that changed from the base of the PR and between 8ad4467 and 37c9701.

📒 Files selected for processing (2)

• macros/edr/system/system_utils/get_pii_columns_from_parent_model.sql
• macros/edr/system/system_utils/is_show_sample_rows_table.sql

🚧 Files skipped from review as they are similar to previous changes (2)

• macros/edr/system/system_utils/get_pii_columns_from_parent_model.sql
• macros/edr/system/system_utils/is_show_sample_rows_table.sql

---

📝 Walkthrough

Walkthrough

Adds a feature-flagged flow to preserve sample rows for models with configured "show sample rows" tags, refactors anomaly test result description to summarize anomalous dimension values, adds integration tests for dimension-anomaly descriptions, tightens Decimal serialization for special values, and updates .gitignore.

Changes

Cohort / File(s)	Summary
Configuration macros/edr/system/system_utils/get_config_var.sql	Adds enable_samples_on_show_sample_rows_tags (default false) and show_sample_rows_tags (default ["show_sample_rows"]).
Show-sample-rows feature macros/edr/system/system_utils/is_show_sample_rows_table.sql, macros/edr/system/system_utils/get_pii_columns_from_parent_model.sql, macros/edr/materializations/test/test.sql	New macro is_show_sample_rows_table(flattened_test); PII column detection respects the show-sample-tags feature (skips columns when matched); test materialization gate preserves sample_limit when is_show_sample_rows_table is true, avoiding PII-based sample disabling.
Anomaly description logic macros/edr/data_monitoring/anomaly_detection/store_anomaly_test_results.sql	Rebuilds test_results_description from collected anomalous_rows, summarizing up to 5 anomalous dimension_values (with remaining count) or concatenating per-row descriptions when ≤5; falls back to last scored row or default message when appropriate.
Tests & serialization integration_tests/tests/test_dimension_anomalies.py, integration_tests/tests/adapter_query_runner.py	Adds two integration tests checking alert descriptions for dimension anomalies (≤5 vs >5); updates _serialize_value() to convert Decimal special normalized exponents (Infinity/NaN) to float and preserve previous int/float mapping otherwise.
Misc .gitignore	Adds .claude/ to ignore patterns.

Sequence Diagram(s)

sequenceDiagram
  rect rgba(200,200,255,0.5)
    participant TestRunner as DbtProject.test
  end
  rect rgba(200,255,200,0.5)
    participant Materialization as test.materialization
    participant Macro as is_show_sample_rows_table
    participant PII as get_pii_columns_from_parent_model
  end
  rect rgba(255,200,200,0.5)
    participant AnomalyMacro as store_anomaly_test_results
    participant Results as TestResults
  end

  TestRunner->>Materialization: invoke test materialization (flattened_test)
  Materialization->>Macro: call is_show_sample_rows_table(flattened_test)
  Macro-->>Materialization: true/false
  alt show_sample_rows true
    Materialization->>PII: preserve sample_limit (skip PII-based disabling)
  else show_sample_rows false
    Materialization->>PII: evaluate PII tags -> may set sample_limit=0
  end
  TestRunner->>AnomalyMacro: persist anomaly results
  AnomalyMacro->>Results: build test_results_description (anomalous_rows summary)
  AnomalyMacro-->>TestRunner: emit results

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 I hopped through tags both bright and small,
I kept sample rows for those who call,
Counted anomalies, five then more,
Wrote tests and tidy decimal lore,
A little rabbit cheers these changes — cottontail encore! 🥕

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the primary feature being added: a new 'show_sample_rows' tag mechanism that can override PII-based sample hiding, which is the central change across all modified files.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/add-show-sample-rows-tag

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@macros/edr/system/system_utils/get_pii_columns_from_parent_model.sql`:
- Around line 41-54: This file was auto-modified by the SQL formatter; run the
project's pre-commit hooks (or directly run sqlfmt) locally and commit the
formatted changes so CI passes. Reformat the macro that defines
enable_show_tags, raw_show_tags, show_tags and the loop over
column_node/all_column_tags_lower to match the repo's sqlfmt/pre-commit rules,
then stage and push the resulting changes.
- Around line 41-46: The current normalization of show_sample_rows_tags
mistakenly treats strings as iterable (so "show" becomes ["s","h","o","w"]);
update the block that sets enable_show_tags / raw_show_tags / show_tags to
follow the same pattern used for pii_tags and get_column_tags: first check if
raw_show_tags is string and wrap it into a single-item list, else if it's
iterable use it, then map to lower and cast to list; adjust the show_tags
assignment (the variable name show_tags and the source raw_show_tags)
accordingly so the downstream intersection check uses a proper list of tag
strings.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 180b1499-7df2-47b3-b07a-5ef781014c1f

📥 Commits

Reviewing files that changed from the base of the PR and between 7a2b542 and 5b68070.

📒 Files selected for processing (8)

• .gitignore
• integration_tests/tests/adapter_query_runner.py
• integration_tests/tests/test_dimension_anomalies.py
• macros/edr/data_monitoring/anomaly_detection/store_anomaly_test_results.sql
• macros/edr/materializations/test/test.sql
• macros/edr/system/system_utils/get_config_var.sql
• macros/edr/system/system_utils/get_pii_columns_from_parent_model.sql
• macros/edr/system/system_utils/is_show_sample_rows_table.sql

[joostboon]

Thanks for the review! Both issues have been addressed:

• sqlfmt formatting: applied and committed
• String iterable bug: fixed in both is_show_sample_rows_table.sql and get_pii_columns_from_parent_model.sql — now using the same is string guard pattern as the existing pii_tags normalization to avoid strings being split into individual characters