quicksql使用mybatis时的注意事项

一、配置双数据源

因为qsql-server会根据runnerType来判断，如果不配置双数据源的话，到时使用到spark时会出现强制转换错误。

            switch (runnerType) {
                case DEFAULT:
                    resultSet = getJDBCResultSet(h, runner, sql);
                    break;
                case JDBC:
                    resultSet = getJDBCResultSet(h, runner, sql);
                    break;
                case SPARK:
                    resultSet = getSparkResultSet(h, runner, sql);
                    break;
                case FLINK:
                    resultSet = getFlinkResultSet(h, runner, sql);
                    break;
            }

jdbc

ResultSet resultSet = (ResultSet) runner.sql(sql).collect();;

spark

 Map.Entry<List<Attribute>, List<GenericRowWithSchema>> sparkData = (Entry<List<Attribute>, List<GenericRowWithSchema>>) runner
                .sql(sql).collect();

二、使用到spark时还需修改qsql-server的源码

修改ColumnMetaData的构造函数中label的赋值。

com.qihoo.qsql.server.QuicksqlServerMeta

 public QueryResult executeQuery(Object object) throws Exception {
        if (object == null) {
            return new QueryResult(new ArrayList<>(), new ArrayList<>());
        }
        Map.Entry<List<Attribute>, List<GenericRowWithSchema>> sparkData = (Entry<List<Attribute>, List<GenericRowWithSchema>>) object;
        if (sparkData == null || CollectionUtils.isEmpty(sparkData.getKey())) {
            throw new SparkException("collect data error");
        }
        List<Attribute> attributes = sparkData.getKey();
        List<GenericRowWithSchema> value = sparkData.getValue();

        List<Object> data = new ArrayList<>();
        List<ColumnMetaData> meta = new ArrayList<>();
        value.stream().forEach(column -> {
            data.add(column.values());
        });
        
        //修改ColumnMetaData，由于之前默认别名为空，我把他设置为列名，quicksql在spark情况中没有处理有别名的情况，直接设置为空。
        attributes.stream().forEach(attribute -> {
            ScalarType columnType = getColumnType(attribute.dataType());
            meta.add(new ColumnMetaData(0, false, true, false, false,
                    1, true, -1, (String) attribute.name(), attribute.name(), (String) null, -1, -1, (String) null, (String) null,
                    columnType, true, false, false, columnType.columnClassName()));
        });
        return new QueryResult(meta, data);
    }

原因：因为mybatis使用到映射时优先选用别名

org.apache.ibatis.executor.resultset.ResultSetWrapper

public ResultSetWrapper(ResultSet rs, Configuration configuration) throws SQLException {
    super();
    this.typeHandlerRegistry = configuration.getTypeHandlerRegistry();
    this.resultSet = rs;
    final ResultSetMetaData metaData = rs.getMetaData();
    final int columnCount = metaData.getColumnCount();
    for (int i = 1; i <= columnCount; i++) {
    
      //server端不修改的话，这里到时会存储null
      columnNames.add(configuration.isUseColumnLabel() ? metaData.getColumnLabel(i) : metaData.getColumnName(i));
      jdbcTypes.add(JdbcType.forCode(metaData.getColumnType(i)));
      classNames.add(metaData.getColumnClassName(i));
    }
  }

private void loadMappedAndUnmappedColumnNames(ResultMap resultMap, String columnPrefix) throws SQLException {
    List<String> mappedColumnNames = new ArrayList<>();
    List<String> unmappedColumnNames = new ArrayList<>();
    final String upperColumnPrefix = columnPrefix == null ? null : columnPrefix.toUpperCase(Locale.ENGLISH);
    final Set<String> mappedColumns = prependPrefixes(resultMap.getMappedColumns(), upperColumnPrefix);
    for (String columnName : columnNames) {
    
     //这里会报空指针
      final String upperColumnName = columnName.toUpperCase(Locale.ENGLISH);
      if (mappedColumns.contains(upperColumnName)) {
        mappedColumnNames.add(upperColumnName);
      } else {
        unmappedColumnNames.add(columnName);
      }
    }