一. Spark SQL提供的操作数据的方式

SQL

Dataframe API

Dataset API

一个用于处理结构化数据的Spark组件，强调的是“结构化数据”，而非“SQL”

二. Spark SQL不仅仅是SQL这么简单的事情，它还能做更多的事情

Hive: SQL

Spark SQL: SQL

三. Spark SQL架构以及处理流程

        Sql -->抽象语法树-->逻辑执行计划 -->优化过后的逻辑执行计划-->物理执行计划-->优化后的物理执行计划-->最终执行引擎上面执行

        Catalyst做了很多的优化工作

四. Dataframe/Dataset详解

Python无法使用Dataset，只可以使用Dataframe, Dataframe也是个分布式的集合，组织成带名字的列

Create an RDD of tuples or lists from the original RDD;

Create the schema represented by a StructType matching the structure of tuples or lists in the RDD created in the step 1.

Apply the schema to the RDD via createDataframe method provided by SparkSession.

Spark中RDD、Dataframe和DataSet的区别 - LestatZ - 博客园

五. 范例

1. 读取文件使用SQL

from pyspark.sql import SparkSession, Row
from pyspark.sql.types import *

def basic(spark):
    df = spark.read.json('file:///home/tarena/桌面/spark/spark_sql/1.json')
    df.show()
    # 显示数据类型
    df.printSchema()
    # 选择所要显示的列
    df.select('staic_total_route').show()
    df.select(df['staic_total_route'], df['staic_total_route'] + 1).show()
    df.filter(df['total_total_route'] > 21).show()
    df.groupBy("host_name").count().show()

    # 注册临时表
    df.createOrReplaceTempView('origin')
    sqlDF = spark.sql('select * from origin')
    sqlDF.show()

def schema_inference_example(spark):
    # Inferring the Schema Using Reflection
    sc = spark.sparkContext

    path = r'file:///home/tarena/桌面/spark/txt/age.txt'
    lines = sc.textFile(path)
    parts = lines.map(lambda l: l.split(" "))
    # print(parts.collect())
    people = parts.map(lambda p: Row(name=p[0], age=p[1]))

    # Infer the schema, and register the Dataframe as a table.
    schemaPeople = spark.createDataframe(people)
    schemaPeople.createOrReplaceTempView("people")

    sqlDF = spark.sql('select * from people')
    sqlDF.show()

    # sql之后转换成rdd进行操作
    sqlRdd = sqlDF.rdd.map(lambda name:'Name: ' + name.name)
    for name in sqlRdd.collect():
        print(name)

def programmatically_schema_example(spark):
    '''
    Create an RDD of tuples or lists from the original RDD;
    Create the schema represented by a StructType matching the structure of tuples or lists in the RDD created in the step 1.
    Apply the schema to the RDD via createDataframe method provided by SparkSession.
    '''

    sc = spark.sparkContext

    # Load a text file and convert each line to a Row.
    path = r'file:///home/tarena/桌面/spark/txt/age.txt'
    lines = sc.textFile(path)
    parts = lines.map(lambda l: l.split(" "))
    # Each line is converted to a tuple.
    people = parts.map(lambda p: (p[0], p[1].strip()))

    # The schema is encoded in a string.
    fields = [StructField(field_name, StringType(), True) for field_name in schemaString.split(' ')]
    schema = StructType(fields)

    # Apply the schema to the RDD.
    schemaPeople = spark.createDataframe(people, schema)
    # Creates a temporary view using the Dataframe
    schemaPeople.createOrReplaceTempView("people")

    # SQL can be run over Dataframes that have been registered as a table.
    results = spark.sql("SELECT * FROM people")
    results.show()

if __name__ == '__main__':
    spark = SparkSession.builder.appName('spark_sql').getOrCreate()
    # sql = SparkSession.builder.appName('spark_sql').enableHiveSupport().getOrCreate()

    # basic(spark)

    # schema_inference_example(spark)

    programmatically_schema_example(spark)
    spark.stop()

2. spark连接sql

from pyspark import SparkContext
from pyspark.sql import SQLContext


if __name__ == "__main__":
    sc = SparkContext(appName="mysqltest")
    sqlContext = SQLContext(sc)
    df = sqlContext.read.format("jdbc").options(
        url="jdbc:mysql://192.168.137.1:3306/employees?user=root&password=123456", dbtable="ztp").load()
    df.show()
    sc.stop()