from pyspark.sql import SQLContext, SparkSession

if __name__ == '__main__':
  # spark 初始化
  spark = SparkSession. \
    Builder(). \
    appName('sql'). \
    master('local'). \
    getOrCreate()
  # mysql 配置(需要修改)
  prop = {'user': 'xxx', 
      'password': 'xxx', 
      'driver': 'com.mysql.cj.jdbc.Driver'}
  # database 地址(需要修改)
  url = 'jdbc:mysql://host:port/database'
  # 讀取表
  data = spark.read.jdbc(url=url, table='tb_newCity', properties=prop)
  # 打印data數(shù)據(jù)類型
  print(type(data))
  # 展示數(shù)據(jù)
  data.show()
  # 關(guān)閉spark會話
  spark.stop()

import pandas as pd
from pyspark import SparkContext
from pyspark.sql import SQLContext, Row

if __name__ == '__main__':
  # spark 初始化
  sc = SparkContext(master='local', appName='sql')
  spark = SQLContext(sc)
  # mysql 配置(需要修改)
  prop = {'user': 'xxx',
      'password': 'xxx',
      'driver': 'com.mysql.cj.jdbc.Driver'}
  # database 地址(需要修改)
  url = 'jdbc:mysql://host:port/database'

  # 創(chuàng)建spark DataFrame
  # 方式1：list轉(zhuǎn)spark DataFrame
  l = [(1, 12), (2, 22)]
  # 創(chuàng)建并指定列名
  list_df = spark.createDataFrame(l, schema=['id', 'value']) 
  
  # 方式2：rdd轉(zhuǎn)spark DataFrame
  rdd = sc.parallelize(l) # rdd
  col_names = Row('id', 'value') # 列名
  tmp = rdd.map(lambda x: col_names(*x)) # 設(shè)置列名
  rdd_df = spark.createDataFrame(tmp) 
  
  # 方式3：pandas dataFrame 轉(zhuǎn)spark DataFrame
  df = pd.DataFrame({'id': [1, 2], 'value': [12, 22]})
  pd_df = spark.createDataFrame(df)

  # 寫入數(shù)據(jù)庫
  pd_df.write.jdbc(url=url, table='new', mode='append', properties=prop)
  # 關(guān)閉spark會話
  sc.stop()