fix: deserialize parquet error when stream's base table modify column type

[zhyass]

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

Summary

This PR fixes an issue where reading or deserializing Parquet data from a stream fails if the base table’s column type has been modified after the stream was created.

When a base table column’s data type changes, the system now drops the old column and re-adds it to regenerate a new column_id. This ensures consistency between the schema definition and column identifiers, preventing mismatched schemas during Parquet deserialization.

Modifying columns while Change Tracking is active should be avoided. Changing schema definitions during change tracking may break consistency between tracked changes and the current table schema, leading to incorrect or incomplete change records.

Fixes: #18827

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 

[github-actions]

🤖 Smart Auto-retry Analysis

Workflow: 18447157252

📊 Summary

• Total Jobs: 83
• Failed Jobs: 2
• Retryable: 0
• Code Issues: 2

❌ NO RETRY NEEDED

All failures appear to be code/test issues requiring manual fixes.

🔍 Job Details

• ❌ linux / sqllogic / cluster_with_minio_and_nginx (http_handler, ttc-go): Not retryable (Code/Test)
• ❌ linux / sqllogic / ee (parquet): Not retryable (Code/Test)

---

🤖 About

Automated analysis using job annotations to distinguish infrastructure issues (auto-retried) from code/test issues (manual fixes needed).