feat(checks): add contains_any and contains_all checks

[TheManishCode]

Summary

• Adds ContainsAny / ContainsAll built-in checks for validating that a response contains at least one, or all, of a list of expected strings (e.g. topic coverage), so users no longer need custom FnCheck lambdas for this.
• Follows the existing StringMatching/RegexMatching pattern in text_matching.py: JSONPath extraction (text_key/values_key), Unicode normalization, and configurable case sensitivity.
• Refactors the shared string-formatting helper (_format_str) out of StringMatching into a module-level function so it can be reused without changing RegexMatching's API.

Test plan

• New unit tests in tests/builtin/test_contains_matching.py covering pass/fail cases, case sensitivity, JSONPath extraction, Unicode normalization, empty/missing/invalid values, and constructor validation errors.
• ruff format --check / ruff check clean.
• basedpyright --level error libs/giskard-checks — 0 errors.
• pytest libs/giskard-checks -m "not functional" — 751 passed (pre-existing RegoPolicy/celine-regorus failures unrelated, no Windows wheel for that dependency).

Closes #2361

[gemini-code-assist]

Code Review

This pull request introduces two new text-matching checks, ContainsAny and ContainsAll, along with comprehensive unit tests and a refactored _format_str helper function. The review feedback highlights an optimization opportunity in both checks, pointing out that the current implementation for computing matched and missing values has $O(N^2)$ complexity and performs redundant string formatting. It is recommended to apply the suggested single-pass $O(N)$ loop to partition the values more efficiently.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

The current implementation of ContainsAny has an $O(N^2)$ complexity for computing missing because it performs a linear lookup (v not in matched) for each element in values. Additionally, it calls _format_str twice per matching element. We can optimize this to a single $O(N)$ pass by iterating through values once and partitioning them into matched and missing lists directly.

Suggested change

	formatted_text = _format_str(text, self.normalization_form, self.case_sensitive)
	matched = [
	v
	for v in values
	if _format_str(v, self.normalization_form, self.case_sensitive)
	in formatted_text
	]
	missing = [v for v in values if v not in matched]
	formatted_text = _format_str(text, self.normalization_form, self.case_sensitive)
	matched = []
	missing = []
	for v in values:
	if _format_str(v, self.normalization_form, self.case_sensitive) in formatted_text:
	matched.append(v)
	else:
	missing.append(v)

[gemini-code-assist]

The current implementation of ContainsAll has an $O(N^2)$ complexity for computing missing because it performs a linear lookup (v not in matched) for each element in values. Additionally, it calls _format_str twice per matching element. We can optimize this to a single $O(N)$ pass by iterating through values once and partitioning them into matched and missing lists directly.

Suggested change

	formatted_text = _format_str(text, self.normalization_form, self.case_sensitive)
	matched = [
	v
	for v in values
	if _format_str(v, self.normalization_form, self.case_sensitive)
	in formatted_text
	]
	missing = [v for v in values if v not in matched]
	formatted_text = _format_str(text, self.normalization_form, self.case_sensitive)
	matched = []
	missing = []
	for v in values:
	if _format_str(v, self.normalization_form, self.case_sensitive) in formatted_text:
	matched.append(v)
	else:
	missing.append(v)

[TheManishCode]

Hi maintainers 👋 — this PR is ready for review but blocked on two org-level gates that only a maintainer can clear:

1. authorize check (Integration Tests workflow) fails because I'm an external contributor — it needs the safe for build label.
2. Main CI and GitHub Actions Security Analysis workflows are stuck in action_required — GitHub's "approve workflow runs from first-time contributors" gate.

In the meantime I've addressed the Gemini Code Assist review feedback (single-pass O(N) partitioning in ContainsAny/ContainsAll instead of O(N²)) and verified locally: ruff format/ruff check clean, basedpyright --level error 0 errors, vermin 3.12-compat clean, pip-audit no known vulnerabilities, and pytest libs/giskard-checks -m "not functional" 751 passed (same pre-existing unrelated celine-regorus/Windows-wheel failures noted in the PR description).

Could someone approve the workflow run and/or add the safe for build label so CI can actually execute? Thanks!