代码-第8章 Structured Streaming-林子雨编著《Spark编程基础（Python版，第2版）》

厦门大学林子雨编著《Spark编程基础（Python版，第2版）》教材中的命令行和代码（教材官网）
提供了教材中的所有章节的命令行和代码，可以直接复制粘贴去执行。
查看《Spark编程基础（Python版，第2版）》教材中的所有命令行和代码

代码文件StructuredNetworkWordCount.py

#!/usr/bin/env python3
 
from pyspark.sql import SparkSession
from pyspark.sql.functions import split
from pyspark.sql.functions import explode
 
if __name__ == "__main__":
    spark = SparkSession \
        .builder \
        .appName("StructuredNetworkWordCount") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("socket") \
        .option("host", "localhost") \
        .option("port", 9999) \
        .load()
 
    words = lines.select(
        explode(
            split(lines.value, " ")
        ).alias("word")
    )
 
    wordCounts = words.groupBy("word").count()
 
    query = wordCounts \
        .writeStream \
        .outputMode("complete") \
        .format("console") \
        .trigger(processingTime="8 seconds") \
        .start()
 
    query.awaitTermination()

cd /usr/local/hadoop
sbin/start-dfs.sh

nc -lk 9999

cd /usr/local/spark/mycode/structuredstreaming/
/usr/local/spark/bin/spark-submit StructuredNetworkWordCount.py

csvDF = spark \
    .readStream \ 
    .format("csv") \
    .option("seq", ";") \
    .load("SOME_DIR")

代码文件spark_ss_filesource_generate.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
 
# 导入需要用到的模块
import os
import shutil
import random
import time
 
TEST_DATA_TEMP_DIR = '/tmp/'
TEST_DATA_DIR = '/tmp/testdata/'
 
ACTION_DEF = ['login', 'logout', 'purchase']
DISTRICT_DEF = ['fujian', 'beijing', 'shanghai', 'guangzhou']
JSON_LINE_PATTERN = '{{"eventTime": {}, "action": "{}", "district": "{}"}}\n'
 
# 测试的环境搭建，判断文件夹是否存在，如果存在则删除旧数据，并建立文件夹
def test_setUp():
    if os.path.exists(TEST_DATA_DIR):
        shutil.rmtree(TEST_DATA_DIR, ignore_errors=True)
    os.mkdir(TEST_DATA_DIR)
 
# 测试环境的恢复，对文件夹进行清理
def test_tearDown():
    if os.path.exists(TEST_DATA_DIR):
        shutil.rmtree(TEST_DATA_DIR, ignore_errors=True)
 
# 生成测试文件
def write_and_move(filename, data):
    with open(TEST_DATA_TEMP_DIR + filename,
              "wt", encoding="utf-8") as f:
        f.write(data)
 
    shutil.move(TEST_DATA_TEMP_DIR + filename,
                TEST_DATA_DIR + filename)
 
if __name__ == "__main__":
    test_setUp()
 
    for i in range(1000):
        filename = 'e-mall-{}.json'.format(i)
 
        content = ''
        rndcount = list(range(100))
        random.shuffle(rndcount)
        for _ in rndcount:
            content += JSON_LINE_PATTERN.format(
                str(int(time.time())),
                random.choice(ACTION_DEF),
                random.choice(DISTRICT_DEF))
        write_and_move(filename, content)
 
        time.sleep(1)
 
    test_tearDown()

代码文件spark_ss_filesource.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
 
# 导入需要用到的模块
import os
import shutil
from pprint import pprint
 
from pyspark.sql import SparkSession
from pyspark.sql.functions import window, asc
from pyspark.sql.types import StructType, StructField
from pyspark.sql.types import TimestampType, StringType
 
# 定义JSON文件的路径常量
TEST_DATA_DIR_SPARK = 'file:///tmp/testdata/'
 
if __name__ == "__main__":
    # 定义模式，为时间戳类型的eventTime、字符串类型的操作和省份组成
    schema = StructType([
        StructField("eventTime", TimestampType(), True),
        StructField("action", StringType(), True),
        StructField("district", StringType(), True)])
 
    spark = SparkSession \
        .builder \
        .appName("StructuredEMallPurchaseCount") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("json") \
        .schema(schema) \
        .option("maxFilesPerTrigger", 100) \
        .load(TEST_DATA_DIR_SPARK)
 
    # 定义窗口
    windowDuration = '1 minutes'
 
    windowedCounts = lines \
        .filter("action = 'purchase'") \
        .groupBy('district', window('eventTime', windowDuration)) \
        .count() \
        .sort(asc('window'))
 
    query = windowedCounts \
        .writeStream \
        .outputMode("complete") \
        .format("console") \
        .option('truncate', 'false') \
        .trigger(processingTime="10 seconds") \
        .start()
 
    query.awaitTermination()

cd /usr/local/hadoop
sbin/start-dfs.sh

cd /usr/local/spark/mycode/structuredstreaming/file
python3 spark_ss_filesource_generate.py

cd /usr/local/spark/mycode/structuredstreaming/file
/usr/local/spark/bin/spark-submit spark_ss_filesource.py

cd /usr/local/kafka
bin/zookeeper-server-start.sh config/zookeeper.properties

cd /usr/local/kafka
bin/kafka-server-start.sh config/server.properties

cd /usr/local/kafka
bin/kafka-console-consumer.sh \
> --bootstrap-server localhost:9092 --topic wordcount-topic

cd /usr/local/kafka
bin/kafka-console-consumer.sh \
> --bootstrap-server localhost:9092 --topic wordcount-result-topic

代码文件/spark_ss_kafka_producer.py

#!/usr/bin/env python3
 
import string
import random
import time
 
from kafka import KafkaProducer
 
if __name__ == "__main__":
    producer = KafkaProducer(bootstrap_servers=['localhost:9092'])
 
    while True:
        s2 = (random.choice(string.ascii_lowercase) for _ in range(2))
        word = ''.join(s2)
        value = bytearray(word, 'utf-8')
 
        producer.send('wordcount-topic', value=value) \
            .get(timeout=10)
 
        time.sleep(0.1)

sudo apt-get install pip3

sudo pip3 install kafka-python

cd /usr/local/spark/mycode/structuredstreaming/kafka/
python3 spark_ss_kafka_producer.py

代码文件spark_ss_kafka_consumer.py

#!/usr/bin/env python3
 
from pyspark.sql import SparkSession
 
if __name__ == "__main__":
    spark = SparkSession \
        .builder \
        .appName("StructuredKafkaWordCount") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("kafka") \
        .option("kafka.bootstrap.servers", "localhost:9092") \
        .option("subscribe", 'wordcount-topic') \
        .load() \
        .selectExpr("CAST(value AS STRING)")
 
    wordCounts = lines.groupBy("value").count()
 
    query = wordCounts \
        .selectExpr("CAST(value AS STRING) as key", "CONCAT(CAST(value AS STRING), ':', CAST(count AS STRING)) as value") \
        .writeStream \
        .outputMode("complete") \
        .format("kafka") \
        .option("kafka.bootstrap.servers", "localhost:9092") \
        .option("topic", "wordcount-result-topic") \
        .option("checkpointLocation", "file:///tmp/kafka-sink-cp") \
        .trigger(processingTime="8 seconds") \
        .start()
 
    query.awaitTermination()

cd /usr/local/spark/mycode/structuredstreaming/kafka/
/usr/local/spark/bin/spark-submit \
> --packages org.apache.spark:spark-sql-kafka-0-10_2.12:3.4.0 \
> spark_ss_kafka_consumer.py

代码文件spark_ss_rate.py

#!/usr/bin/env python3
 
from pyspark.sql import SparkSession
 
if __name__ == "__main__":
    spark = SparkSession \
        .builder \
        .appName("TestRateStreamSource") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("rate") \
        .option('rowsPerSecond', 5) \
        .load()
 
    print(lines.schema)
 
    query = lines \
        .writeStream \
        .outputMode("update") \
        .format("console") \
        .option('truncate', 'false') \
        .start()
 
    query.awaitTermination()

代码文件StructuredNetworkWordCountFileSink.py

#!/usr/bin/env python3
 
from pyspark.sql import SparkSession
from pyspark.sql.functions import split
from pyspark.sql.functions import explode
from pyspark.sql.functions import length
 
if __name__ == "__main__":
    spark = SparkSession \
        .builder \
        .appName("StructuredNetworkWordCountFileSink") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("socket") \
        .option("host", "localhost") \
        .option("port", 9999) \
        .load()
 
    words = lines.select(
        explode(
            split(lines.value, " ")
        ).alias("word")
    )
 
    all_length_5_words = words.filter(length("word") == 5)
 
    query = all_length_5_words \
        .writeStream \
        .outputMode("append") \
        .format("parquet") \
        .option("path", "file:///tmp/filesink") \
        .option("checkpointLocation", "file:///tmp/file-sink-cp") \
        .trigger(processingTime="8 seconds") \
        .start()
 
    query.awaitTermination()

nc -lk 9999

cd /usr/local/spark/mycode/structuredstreaming
/usr/local/spark/bin/spark-submit StructuredNetworkWordCountFileSink.py

cd /tmp/filesink
ls

代码文件spark_ss_test_delay.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
 
# 导入需要用到的模块
import os
import shutil
from functools import partial
 
from pyspark.sql import SparkSession
from pyspark.sql.functions import window
from pyspark.sql.types import StructType, StructField
from pyspark.sql.types import TimestampType, StringType
 
# 定义CSV文件的路径常量
TEST_DATA_DIR = '/tmp/testdata/'
TEST_DATA_DIR_SPARK = 'file:///tmp/testdata/'
 
# 搭建测试环境，判断CSV文件夹是否存在，如果存在则删除旧数据，并建立文件夹
def test_setUp():
    if os.path.exists(TEST_DATA_DIR):
        shutil.rmtree(TEST_DATA_DIR, ignore_errors=True)
    os.mkdir(TEST_DATA_DIR)
 
# 恢复测试环境，对CSV文件夹进行清理
def test_tearDown():
    if os.path.exists(TEST_DATA_DIR):
        shutil.rmtree(TEST_DATA_DIR, ignore_errors=True)
 
# 编写模拟输入的函数，传入CSV文件名和数据。注意写入应当是原子性的
# 如果写入时间较长，应当先写入临时文件再移动到CSV目录内
# 这里采取直接写入的方式
def write_to_csv(filename, data):
    with open(TEST_DATA_DIR + filename, "wt", encoding="utf-8") as f:
        f.write(data)
 
if __name__ == "__main__":
    test_setUp()
 
    # 定义模式，该模式为字符串类型的word和时间戳类型的eventTime两个列组成
    schema = StructType([
        StructField("word", StringType(), True),
        StructField("eventTime", TimestampType(), True)])
 
    spark = SparkSession \
        .builder \
        .appName("StructuredNetworkWordCountWindowedDelay") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format('csv') \
        .schema(schema) \
        .option("sep", ";") \
        .option("header", "false") \
        .load(TEST_DATA_DIR_SPARK)
 
    # 定义窗口
    windowDuration = '1 hour'
 
    windowedCounts = lines \
        .withWatermark("eventTime", "1 hour") \
        .groupBy('word', window('eventTime', windowDuration)) \
        .count()
 
    query = windowedCounts \
        .writeStream \
        .outputMode("update") \
        .format("console") \
        .option('truncate', 'false') \
        .trigger(processingTime="8 seconds") \
        .start()
 
    # 写入测试文件file1.csv
    write_to_csv('file1.csv', """
正常;2023-10-01 08:00:00
正常;2023-10-01 08:10:00
正常;2023-10-01 08:20:00
""")
 
    # 处理当前数据
    query.processAllAvailable()
 
    # 此时事件时间更新到上次看到的最大的2023-10-01 08:20:00
 
    write_to_csv('file2.csv', """
正常;2023-10-01 20:00:00
一小时以内延迟到达;2023-10-01 10:00:00
一小时以内延迟到达;2023-10-01 10:50:00
""")
 
    # 处理当前数据
    query.processAllAvailable()
 
    # 此时事件时间更新到上次看到的最大的2023-10-01 20:00:00
 
    write_to_csv('file3.csv', """
正常;2023-10-01 20:00:00
一小时以外延迟到达;2023-10-01 10:00:00
一小时以外延迟到达;2023-10-01 10:50:00
一小时以内延迟到达;2023-10-01 19:00:00
""")
 
    # 处理当前数据
    query.processAllAvailable()
 
    query.stop()
 
    test_tearDown()

cd /usr/local/spark/mycode/structuredstreaming/watermark/
/usr/local/spark/bin/spark-submit spark_ss_test_delay.py

代码文件StructuredNetworkWordCountWithMonitor.py

#!/usr/bin/env python3
 
from pprint import pprint
import time
 
from pyspark.sql import SparkSession
from pyspark.sql.functions import split
from pyspark.sql.functions import explode
 
if __name__ == "__main__":
    spark = SparkSession \
        .builder \
        .appName("StructuredNetworkWordCount") \
        .getOrCreate()
 
    spark.sparkContext.setLogLevel('WARN')
 
    lines = spark \
        .readStream \
        .format("socket") \
        .option("host", "localhost") \
        .option("port", 9999) \
        .load()
 
    words = lines.select(
        explode(
            split(lines.value, " ")
        ).alias("word")
    )
 
    wordCounts = words.groupBy("word").count()
 
    query = wordCounts \
        .writeStream \
        .outputMode("complete") \
        .format("console") \
        .queryName('write_to_console') \
        .trigger(processingTime="8 seconds") \
        .start()
 
    while True:
        if query.lastProgress:
            if query.lastProgress['numInputRows'] > 0:
                pprint(query.lastProgress)
 
        pprint(query.status)
 
        time.sleep(5)

cd /usr/local/hadoop
sbin/start-dfs.sh

nc -lk 9999

cd /usr/local/spark/mycode/structuredstreaming/monitor
/usr/local/spark/bin/spark-submit StructuredNetworkWordCountWithMonitor.py

厦大数据库实验室博客