[Feature Request]: Add Flexible Validation Rules (Assertions) to Unit Test Golden Data Sets

### What would you like to happen?

Description: Issue Type: Improvement / Feature Request Priority: Major

Summary: Currently, the Golden Data Set validation in Pipeline Unit Tests performs a strict comparison: the output rows must exactly match the values in the dataset. If a single character differs (or if the sort order changes), the test fails.

This rigidity makes it difficult to test pipelines that generate dynamic data, such as:

Timestamps/Dates: (e.g., sysdate or execution_date).

Generated IDs: (e.g., UUIDs or sequence numbers).

Variable outputs: (e.g., Live API responses).

Proposed Solution: Enhance the Golden Data Set configuration to allow Validation Rules (Assertions) per field, rather than just static values.

Instead of only checking Value == Expected, allow the user to define rules such as:

Not Null: Pass if the field has any value.

Data Type Check: Pass if the value is a specific type (e.g., String, Integer).

Numeric Ranges: Pass if Value > X or Value < Y.

String Length: Pass if Length > X.

Regex Matching: Pass if the string matches a pattern (e.g., ^[0-9]{3}-[0-9]{2}$).

Contains: Pass if the string contains a specific substring.

Benefit: This would drastically improve the usability of Unit Tests for real-world scenarios where data is dynamic. Developers could validate the structure and integrity of the data without needing to know the exact runtime values beforehand.

### Issue Priority

Priority: 3

### Issue Component

Component: Pipelines

Start by locating the Pipeline Unit Tests and Golden Data Set validation and configuration entry points; no specific files or tests are named in the issue. Define the supported per-field assertions and how they are represented, then verify that dynamic values can be checked for nullability, type, ranges, length, regex, and substring matches without requiring exact values.

Written by the indexing model from the issue text.