项目中用到python操作hdfs的问题,一般都是使用python的hdfs包,然而这个包初始化起来太麻烦,需要:

from pyspark impport SparkConf, SparkContext
from
hdfs import * client = Client("http://127.0.0.1:50070")

  可以看到python需要指定master的地址,平时Scala使用的时候不用这样,如下:

import org.apache.hadoop.fs.{FileSystem, Path}
import org.apache.spark.{SparkConf, SparkContext}
hdfs = FileSystem.get(sc.hadoopConfiguration)

  如果我们要在本地测试和生产打包发布的时候,python这样需要每次修改master地址的方式很不方便,而且一般本地调试的时候一般hadoop需要的时候才开起来,Scala启动的时候是在项目目录的根目录直接启动hdfs,但是python调用hadoop的话需要本地开启hadoop服务,通过localhost:50070监听。于是想通过python调用Scala的Filesystem来实现这个操作。

  阅读spark的源码发现python是使用py4j这个py文件和java交互的,通过gateway启动jvm,这里的源码有很大用途,于是我做了修改:

#!/usr/bin/env python
# coding:utf-8

import re
import jieba
import atexit
import os
import select
import signal
import shlex
import socket
import platform
from subprocess import Popen, PIPE
from py4j.java_gateway import java_import, JavaGateway, GatewayClient
from common.Tools import loadData
from pyspark import SparkContext
from pyspark.serializers import read_int

if "PYSPARK_GATEWAY_PORT" in os.environ:
    gateway_port = int(os.environ["PYSPARK_GATEWAY_PORT"])
else:
    SPARK_HOME = os.environ["SPARK_HOME"]
    # Launch the Py4j gateway using Spark's run command so that we pick up the
    # proper classpath and settings from spark-env.sh
    on_windows = platform.system() == "Windows"
    script = "./bin/spark-submit.cmd" if on_windows else "./bin/spark-submit"
    submit_args = os.environ.get("PYSPARK_SUBMIT_ARGS", "pyspark-shell")
    if os.environ.get("SPARK_TESTING"):
        submit_args = ' '.join([
            "--conf spark.ui.enabled=false",
            submit_args
        ])
    command = [os.path.join(SPARK_HOME, script)] + shlex.split(submit_args)

    # Start a socket that will be used by PythonGatewayServer to communicate its port to us
    callback_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    callback_socket.bind(('127.0.0.1', 0))
    callback_socket.listen(1)
    callback_host, callback_port = callback_socket.getsockname()
    env = dict(os.environ)
    env['_PYSPARK_DRIVER_CALLBACK_HOST'] = callback_host
    env['_PYSPARK_DRIVER_CALLBACK_PORT'] = str(callback_port)

    # Launch the Java gateway.
    # We open a pipe to stdin so that the Java gateway can die when the pipe is broken
    if not on_windows:
        # Don't send ctrl-c / SIGINT to the Java gateway:
        def preexec_func():
            signal.signal(signal.SIGINT, signal.SIG_IGN)


        proc = Popen(command, stdin=PIPE, preexec_fn=preexec_func, env=env)
    else:
        # preexec_fn not supported on Windows
        proc = Popen(command, stdin=PIPE, env=env)

    gateway_port = None
    # We use select() here in order to avoid blocking indefinitely if the subprocess dies
    # before connecting
    while gateway_port is None and proc.poll() is None:
        timeout = 1  # (seconds)
        readable, _, _ = select.select([callback_socket], [], [], timeout)
        if callback_socket in readable:
            gateway_connection = callback_socket.accept()[0]
            # Determine which ephemeral port the server started on:
            gateway_port = read_int(gateway_connection.makefile(mode="rb"))
            gateway_connection.close()
            callback_socket.close()
    if gateway_port is None:
        raise Exception("Java gateway process exited before sending the driver its port number")

    # In Windows, ensure the Java child processes do not linger after Python has exited.
    # In UNIX-based systems, the child process can kill itself on broken pipe (i.e. when
    # the parent process' stdin sends an EOF). In Windows, however, this is not possible
    # because java.lang.Process reads directly from the parent process' stdin, contending
    # with any opportunity to read an EOF from the parent. Note that this is only best
    # effort and will not take effect if the python process is violently terminated.
    if on_windows:
        # In Windows, the child process here is "spark-submit.cmd", not the JVM itself
        # (because the UNIX "exec" command is not available). This means we cannot simply
        # call proc.kill(), which kills only the "spark-submit.cmd" process but not the
        # JVMs. Instead, we use "taskkill" with the tree-kill option "/t" to terminate all
        # child processes in the tree (http://technet.microsoft.com/en-us/library/bb491009.aspx)
        def killChild():
            Popen(["cmd", "/c", "taskkill", "/f", "/t", "/pid", str(proc.pid)])


        atexit.register(killChild)

    # Connect to the gateway
gateway = JavaGateway(GatewayClient(port=gateway_port), auto_convert=True)

# Import the classes used by PySpark
java_import(gateway.jvm, "org.apache.spark.SparkConf")
java_import(gateway.jvm, "org.apache.spark.api.java.*")
java_import(gateway.jvm, "org.apache.spark.api.python.*")
java_import(gateway.jvm, "org.apache.spark.ml.python.*")
java_import(gateway.jvm, "org.apache.spark.mllib.api.python.*")
# TODO(davies): move into sql
java_import(gateway.jvm, "org.apache.spark.sql.*")
java_import(gateway.jvm, "org.apache.spark.sql.hive.*")
java_import(gateway.jvm, "scala.Tuple2")
java_import(gateway.jvm, "org.apache.hadoop.fs.{FileSystem, Path}")
java_import(gateway.jvm, "org.apache.hadoop.conf.Configuration")
java_import(gateway.jvm, "org.apache.hadoop.*")
java_import(gateway.jvm, "org.apache.spark.{SparkConf, SparkContext}")
jvm = gateway.jvm
conf = jvm.org.apache.spark.SparkConf()
conf.setMaster("local").setAppName("test hdfs")
sc = jvm.org.apache.spark.SparkContext(conf)
print(sc.hadoopConfiguration())
FileSystem = jvm.org.apache.hadoop.fs.FileSystem
print(repr(FileSystem))
Path = jvm.org.apache.hadoop.fs.Path
hdfs = FileSystem.get(sc.hadoopConfiguration())
hdfs.delete(Path("/DATA/*/*/TMP/KAIVEN/*"))
print(‘目录删除成功’)

  可以参考py4j.java_gateway.launch_gateway,这个方法是python启动jvm的,本人做了一点小小的修改用java_import 调用了

org.apache.hadoop.fs.{FileSystem, Path},
org.apache.hadoop.conf.Configuration
这样的话sparkConf起来的时候就自动配置了Scala的配置。python调用scala或java包-LMLPHP
01-24 17:23