API_doc.md

API Document

Overview

QSQL supports called API directly by other applications to execute queries. If you want to do this, import QSQL dependency into your project first.

<dependency>
		<groupId>com.qihoo.qsql</groupId>
        <artifactId>qsql-core</artifactId>
        <version>${project.version}</version>
</dependency>

Examples

QSQL provides a simple API for users to query by SQL, users don't need to care about the storage which is defined in SQL from clause, only focus on data itself.

Example 1: A sample without any configuration, only query and show results .

//Create a dynamic SQL runner to execute SQL query.
SqlRunner runner = SqlRunner.builder().ok();
//Execute and run.
runner.sql("SELECT age, stu_name, dep_name "
     + "FROM student stu INNER JOIN department dep "
     + "on stu.dep_id = dep.id")
    .show()
    .run();

Example 2: A little more custom configurations.

SqlRunner runner = SqlRunner.builder()
    .setTransformRunner(RunnerType.SPARK) //Choose Spark to execute query. 
    .setAppName("A Simple Example") //Set application name. default is current ts.
    .setAcceptedResultsNum(1000) //Set maxinum result set size, default is 1000.
    .setMaster("local[*]")	//Set execute mode, default is local[*].
    .setSchemaPath()  //Set path of schema in the form of file.
    .setProperties()  //Set properties of concrete engine.
    .ok();

runner.sql("SELECT * FROM student GROUP BY age")
    .asTextFile("/tmp/part") //Save results on hdfs or local filesystem.
    .run();

QSQL only supports these API currently, more API will be developed gradually.

Submit Job

After writing the code in your project, you should package this class, then put it into your LINUX server. You need to submit the task using the submit command corresponding to the runner used in your program. If you use spark runner, you can submit just like this:

spark-submit 

Note: If you have not set the runner, just use java -jar set up your package is also well.