[Bug] The better implementation of StatementParser

[smm321]

Search before asking

• I searched in the issues and found nothing similar.

Paimon UI version

main

Compute Engine

Flink

Minimal reproduce step

Currently,org.apache.paimon.web.engine.flink.common.parser.StatementParser processes input strings in a 'split line Separator' way. It can't handle the statements which contain a line separator itself.

What doesn't meet your expectations?

In my option, a more graceful way to solve this issue is that we create a Calcite Sql Parser with a FlinkSqlParser Factory.

/**
 * CustomSqlParser to parse Sql list.
 */
public class CustomSqlParser {

    private static final SqlParser.Config config;
    private SqlParser parser;
    static {
        config = SqlParser.configBuilder()
                .setParserFactory(FlinkSqlParserImpl.FACTORY)
                .setConformance(FlinkSqlConformance.DEFAULT)
                .setLex(Lex.JAVA)
                .setIdentifierMaxLength(256)
                .build();
    }

    public CustomSqlParser(String sql){
        parser = SqlParser.create(sql, config);
    }

    public SqlParser getParser(){
        return parser;
    }
}

Then sqlParser can parse kinds of statements and return a sqlNodeList. The last thing we need to do is get the sqlKind of every sqlNode such as

CustomSqlParser customSqlParser = new CustomSqlParser(statementSetString);
SqlNodeList sqlNodeList = customSqlParser.getParser().parseStmtList();
for (SqlNode sqlNode : sqlNodeList) {
     System.out.println(sqlNode.getKind());
}

We just use AST level of Apache Calcite that do not need to validate if the tables or columns are legal or not

Anything else?

no

Are you willing to submit a PR?

• I'm willing to submit a PR!