Refactoring Validation logic into separate function

[abdullah-workday]

Overview

• Refactoring packages/server/src/services/validation/index.ts to extract the validation logic into a separate function.

What is Changed

• packages/server/src/services/validation/index.ts -> Code Refactoring
• packages/server/src/services/validation/validateFlowData.test.ts -> Adding unit test

[gemini-code-assist]

Code Review

This pull request refactors the flow validation logic by extracting the core validation rules into a pure function, validateFlowData, and introduces a comprehensive unit test suite. The refactoring improves testability and separates data retrieval from validation logic. Feedback identifies a performance bottleneck in the hanging edges check that could be optimized using a Set for node lookups. Additionally, a logic issue was found where the regex for normalizing array index conditions is hardcoded to a specific parameter name, which may cause validation to fail for other array-type parameters.

[gemini-code-assist]

The hanging edges check has $O(E \times N)$ complexity because it performs nodes.some (an $O(N)$ operation) twice for every edge. For large flows, this can significantly impact performance. Using a Set for node ID lookups reduces this to $O(E)$.

    // Check for hanging edges
    const nodeIds = new Set(nodes.map((node) => node.id))
    for (const edge of edges) {
        const sourceExists = nodeIds.has(edge.source)
        const targetExists = nodeIds.has(edge.target)

[gemini-code-assist]

The regex for $index conditions is hardcoded to the string conditions. If an array parameter has a different name (e.g., actions), this validation logic will fail to correctly normalize the key. This issue also exists on line 135. Consider using the actual parameter name param.name to build the regex dynamically.