feat(parser): add column data type fragment parser

[minleejae]

What

• Add CCJSqlParserUtil.parseColDataType(String).
• Add a configurable overload accepting Consumer<CCJSqlParser> for dialect and feature settings.
• Require the complete input to be consumed, rejecting trailing column constraints or modifiers instead of returning a partial type.
• Cover PostgreSQL array dimensions, MySQL ENUM arguments, parser configuration, null/empty input, and partial-input rejection.

Why

ColDataType() is available on the generated parser, but there is no public CCJSqlParserUtil entry point for applications that receive a type fragment rather than a complete DDL statement. Such applications currently need to instantiate the generated parser directly or wrap the fragment in synthetic DDL.

The strict EOF check is important for type fragments: returning varchar(255) from varchar(255) NOT NULL would silently discard meaningful input.

API example

ColDataType type = CCJSqlParserUtil.parseColDataType(
        "enum('small','medium')",
        parser -> parser.withDialect(AbstractJSqlParser.Dialect.MYSQL));

Testing

• ./gradlew check
• Added focused tests in CCJSqlParserUtilTest.

AI assistance

OpenAI Codex was used to review existing parser utility conventions, prepare the API and tests, and run validation. The generated change was reviewed for strict full-input behavior and compatibility with parser configuration.

[manticore-projects]

Thank you!