refactor(functions): Optimize jq function performance

[b41sh]

I hereby agree to the terms of the CLA available at: https://docs.databend.com/dev/policies/cla/

Summary

This PR introduces a significant performance optimization for the jq function when working with jsonb data. Previously, invoking the jq function required an additional conversion step where jsonb data had to be transformed into serde_json::Value. After the execution of the jq function, another conversion was necessary to convert the result back into jsonb. This overhead not only added complexity but also resulted in suboptimal performance.

Changes

• Eliminated the need for converting jsonb to serde_json::Value executing the jq function.
• Enabled direct execution of the jq function on jsonb data, thus avoiding the extra conversion steps.

fixes: #[Link the issue here]

Tests

• Unit Test
• Logic Test
• Benchmark Test
• No Test - Explain why

Type of change

• Bug Fix (non-breaking change which fixes an issue)
• New Feature (non-breaking change which adds functionality)
• Breaking Change (fix or feature that could cause existing functionality not to work as expected)
• Documentation Update
• Refactoring
• Performance Improvement
• Other (please describe):

---

This change is 

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 78557f8ab7

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".