博客

读取文件的多种方式：

1、spark直接读取http数据，直接生成DataSet

2、InputStreamReader

3、Source.fromInputStream

package com.ku.test

import java.io.{BufferedReader, InputStreamReader}
import java.net.{URL, URLConnection}

import org.apache.http.client.methods.HttpGet
import org.apache.http.impl.client.DefaultHttpClient

import scala.io.Source

import org.apache.spark.SparkFiles
import org.apache.spark.sql.SparkSession


object TokenTest {

  def main(args: Array[String]): Unit = {

    //方法1
    val spark = SparkSession.builder()
      .appName("tokenTest")
      .master("local")
      .getOrCreate()
    spark.sparkContext.setLogLevel("WARN")

    val fileName = "test.txt"
    spark.sparkContext.addFile("http://127.0.0.1/test/" + fileName)
    spark.read
      .textFile("file:///" + SparkFiles.get(fileName))
      .filter(s => !s.isEmpty)
      .show(truncate = true, numRows = 20000)

    //方法2
    val url2: URL = new URL("http://127.0.0.1/test/test.txt")
    val urlConn: URLConnection = url2.openConnection()
    val in = new BufferedReader(new InputStreamReader(urlConn.getInputStream))
    var line: String = ""
    while (!"".equals(line = in.readLine())) {
      println(line)
    }
    


    //方法3
    val url3 = "http://127.0.0.1/test/test.txt"
    val httpclient = new DefaultHttpClient()
    try {
      val response = httpclient.execute(new HttpGet(url3))
      val entity = response.getEntity
      val file = Source.fromInputStream(entity.getContent) //.getLines().mkString //.foreach(row => println(row))
      for (line <- file.getLines) {
        println(line)
      }
    }
  }
}

先记录下，后面作详细更新