[client] Add support of column aliasing for PojoConverter

[VladBanar]

Purpose

Linked issue: closes #3439

Currently, POJO field names must exactly match Fluss table column names. This creates rigid naming constraints that conflict with Java naming conventions (camelCase) and database naming conventions (snake_case).

Brief change log

1. Created @ColumnName Annotation
   Example: @ColumnName("user_id") on a field named userId
2. Modified PojoType.java
   Updated discoverAllInstanceFields() to check for @ColumnName annotations
   When @ColumnName is present, uses the annotation value as the property key instead of the field name
   Maintains backward compatibility: fields without annotation use original field names
   Properties map is now keyed by column names (mapped or original)
3. Converter Integration
   PojoToRowConverter and RowToPojoConverter automatically work with the updated property mapping
   No changes needed in converters—they query properties by column name, which now respects@ColumnName mappings

Tests

Added new test PojoTypeTest:testColumnNameAnnotation

API and Format

No

Documentation

Updated website/docs/apis/java-client.md

• Added note about @ColumnName annotation in Defining POJOs section