com.yahoo.bard.webservice.sql

Class SqlResultSetProcessor

java.lang.Object

com.yahoo.bard.webservice.sql.SqlResultSetProcessor

public class SqlResultSetProcessor
extends Object

Process the results from a DruidQuery to a sql backend.

Constructor Summary

Constructors

Constructor and Description
SqlResultSetProcessor(DruidAggregationQuery<?> druidQuery, ApiToFieldMapper apiToFieldMapper, com.fasterxml.jackson.databind.ObjectMapper objectMapper, SqlTimeConverter sqlTimeConverter) Builds something to process a set of sql results and return them as the same format as a GroupBy query to Druid.

Method Summary

Modifier and Type	Method and Description
com.fasterxml.jackson.databind.JsonNode	buildDruidResponse() Processes the results from the sql ResultSet and writes them out as the json format returned for a GroupByQuery.
protected static Map<String,Function<String,Number>>	getAggregationTypeMapper(DruidAggregationQuery<?> druidQuery) Creates a map from each aggregation name, i.e.
void	process(ResultSet sqlResultSet) Reads the result set and converts it into a result that druid would produce.
protected void	processRow(Map<String,Function<String,Number>> resultTypeMapper, com.fasterxml.jackson.core.JsonGenerator jsonWriter, String[] row) Processes a single row of results from the result set.
protected static void	writeNumberField(com.fasterxml.jackson.core.JsonGenerator jsonWriter, String name, Number number) Writes a Number as either a Double or Long in json.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SqlResultSetProcessor

public SqlResultSetProcessor(DruidAggregationQuery<?> druidQuery,
                             ApiToFieldMapper apiToFieldMapper,
                             com.fasterxml.jackson.databind.ObjectMapper objectMapper,
                             SqlTimeConverter sqlTimeConverter)

Builds something to process a set of sql results and return them as the same format as a GroupBy query to Druid.

Parameters:

druidQuery - The original query that was converted to a sql query.

apiToFieldMapper - The mapping from api to physical name.

objectMapper - The mapper for all JSON processing.

sqlTimeConverter - The time converter used with making the query.

Method Detail

buildDruidResponse

public com.fasterxml.jackson.databind.JsonNode buildDruidResponse()

Processes the results from the sql ResultSet and writes them out as the json format returned for a GroupByQuery.

Returns:

the equivalent json.

processRow

protected void processRow(Map<String,Function<String,Number>> resultTypeMapper,
                          com.fasterxml.jackson.core.JsonGenerator jsonWriter,
                          String[] row)
                   throws IOException

Processes a single row of results from the result set.

Parameters:

resultTypeMapper - The mapping from an aggregation to a function which corrects it's type.

jsonWriter - The generator for writing the json results.

row - The result row.

Throws:

IOException - if failed while writing json.

process

public void process(ResultSet sqlResultSet)
             throws SQLException

Reads the result set and converts it into a result that druid would produce.

Parameters:

sqlResultSet - The result set of the druid query.

Throws:

SQLException - if results can't be read.

writeNumberField

protected static void writeNumberField(com.fasterxml.jackson.core.JsonGenerator jsonWriter,
                                       String name,
                                       Number number)
                                throws IOException

Writes a Number as either a Double or Long in json.

Parameters:

jsonWriter - The writer used to build json.

name - The name of the field to write with json.

number - The Number value of the field to write with json.

Throws:

IOException - if results can't be written.

getAggregationTypeMapper

protected static Map<String,Function<String,Number>> getAggregationTypeMapper(DruidAggregationQuery<?> druidQuery)

Creates a map from each aggregation name, i.e. ("longSum", "doubleSum"), to a function which will parse to the correct type, i.e. (long, double). If no type is found it will do nothing.

Parameters:

druidQuery - The query to make a mapper for.

Returns:

the map from aggregation name to Double.parseDouble(java.lang.String) Long.parseLong(java.lang.String, int).