我怎样才能使我的文件 reading/writing 程序更快?
How can I make my file reading/writing program faster?
所以我有一个包含一堆天气数据的大文件。我必须将大文件中的每一行分配到其相应的状态文件中。因此总共将有 50 个具有自己数据的新状态文件。
大文件包含约 100 万行记录,如下所示:
COOP:166657,'NEW IBERIA AIRPORT ACADIANA REGIONAL LA US',200001,177,553
虽然电台的名称可能会有所不同并且有不同的字数。
目前我正在使用正则表达式查找模式并输出到文件,它必须按状态分组。如果我在不做任何修改的情况下读入整个文件,大约需要 46 秒。使用查找州缩写、创建文件并输出到该文件的代码,需要 10 多分钟。
这是我现在拥有的:
package climate;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.util.Arrays;
import java.util.Scanner;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* This program will read in a large file containing many stations and states,
* and output in order the stations to their corresponding state file.
*
* Note: This take a long time depending on processor. It also appends data to
* the files so you must remove all the state files in the current directory
* before running for accuracy.
*
* @author Marcus
*
*/
public class ClimateCleanStates {
public static void main(String[] args) throws IOException {
Scanner in = new Scanner(System.in);
System.out
.println("Note: This program can take a long time depending on processor.");
System.out
.println("It is also not necessary to run as state files are in this directory.");
System.out
.println("But if you would like to see how it works, you may continue.");
System.out.println("Please remove state files before running.");
System.out.println("\nIs the States directory empty?");
String answer = in.nextLine();
if (answer.equals("N")) {
System.exit(0);
in.close();
}
System.out.println("Would you like to run the program?");
String answer2 = in.nextLine();
if (answer2.equals("N")) {
System.exit(0);
in.close();
}
String[] statesSpaced = new String[51];
File statefile, dir, infile;
// Create files for each states
dir = new File("States");
dir.mkdir();
infile = new File("climatedata.csv");
FileReader fr = new FileReader(infile);
BufferedReader br = new BufferedReader(fr);
String line;
line = br.readLine();
System.out.println();
// Read in climatedata.csv
final long start = System.currentTimeMillis();
while ((line = br.readLine()) != null) {
// Remove instances of -9999
if (!line.contains("-9999")) {
String stateFileName = null;
Pattern p = Pattern.compile(".* ([A-Z][A-Z]) US");
Matcher m = p.matcher(line);
if (m.find()){
stateFileName = m.group(1);
stateFileName = "States/" + stateFileName + ".csv";
statefile = new File(stateFileName);
FileWriter stateWriter = new FileWriter(statefile, true);
stateWriter.write(line + "\n");
// Progress reporting
//System.out.printf("Writing [%s] to file [%s]\n", line,
// statefile);
stateWriter.flush();
stateWriter.close();
}
}
}
System.out.println("Elapsed " + (System.currentTimeMillis() - start) + " ms");
br.close();
fr.close();
in.close();
}
}
不要在循环中做任何你不需要做的事情。制作一张以文件名为关键字的文件地图,并让它们保持打开状态。根据需要给他们写信,在完成之前不要关闭任何一个。您也永远不需要冲洗它们。你正在击败所有好的缓冲。完成后,您可以考虑增大缓冲区。
您可以使用 Map
来跟踪状态文件而不是每次都关闭它们:
Map<String, FileWriter> fileMap = new HashMap<String, FileWriter>();
while ((line = br.readLine()) != null) {
if (!line.contains("-9999")) {
if (m.find()) {
stateFileName = m.group(1);
stateFileName = "States/" + stateFileName + ".csv";
FileWriter stateFileWriter = fileMap.get(stateFileName);
if (stateFileWriter == null) {
stateFileWriter = new FileWriter(stateFileName, true);
fileMap.put(stateFileName, stateFileWriter);
}
stateFileWriter.write(line + "\n");
}
}
}
// flush the writers and close once you have parsed the entire file
for(Map.Entry<String, FileWriter> entry : fileMap.entrySet()) {
FileWriter writer = entry.getValue();
writer.flush();
writer.close();
}
所以我有一个包含一堆天气数据的大文件。我必须将大文件中的每一行分配到其相应的状态文件中。因此总共将有 50 个具有自己数据的新状态文件。
大文件包含约 100 万行记录,如下所示:
COOP:166657,'NEW IBERIA AIRPORT ACADIANA REGIONAL LA US',200001,177,553
虽然电台的名称可能会有所不同并且有不同的字数。
目前我正在使用正则表达式查找模式并输出到文件,它必须按状态分组。如果我在不做任何修改的情况下读入整个文件,大约需要 46 秒。使用查找州缩写、创建文件并输出到该文件的代码,需要 10 多分钟。
这是我现在拥有的:
package climate;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.util.Arrays;
import java.util.Scanner;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* This program will read in a large file containing many stations and states,
* and output in order the stations to their corresponding state file.
*
* Note: This take a long time depending on processor. It also appends data to
* the files so you must remove all the state files in the current directory
* before running for accuracy.
*
* @author Marcus
*
*/
public class ClimateCleanStates {
public static void main(String[] args) throws IOException {
Scanner in = new Scanner(System.in);
System.out
.println("Note: This program can take a long time depending on processor.");
System.out
.println("It is also not necessary to run as state files are in this directory.");
System.out
.println("But if you would like to see how it works, you may continue.");
System.out.println("Please remove state files before running.");
System.out.println("\nIs the States directory empty?");
String answer = in.nextLine();
if (answer.equals("N")) {
System.exit(0);
in.close();
}
System.out.println("Would you like to run the program?");
String answer2 = in.nextLine();
if (answer2.equals("N")) {
System.exit(0);
in.close();
}
String[] statesSpaced = new String[51];
File statefile, dir, infile;
// Create files for each states
dir = new File("States");
dir.mkdir();
infile = new File("climatedata.csv");
FileReader fr = new FileReader(infile);
BufferedReader br = new BufferedReader(fr);
String line;
line = br.readLine();
System.out.println();
// Read in climatedata.csv
final long start = System.currentTimeMillis();
while ((line = br.readLine()) != null) {
// Remove instances of -9999
if (!line.contains("-9999")) {
String stateFileName = null;
Pattern p = Pattern.compile(".* ([A-Z][A-Z]) US");
Matcher m = p.matcher(line);
if (m.find()){
stateFileName = m.group(1);
stateFileName = "States/" + stateFileName + ".csv";
statefile = new File(stateFileName);
FileWriter stateWriter = new FileWriter(statefile, true);
stateWriter.write(line + "\n");
// Progress reporting
//System.out.printf("Writing [%s] to file [%s]\n", line,
// statefile);
stateWriter.flush();
stateWriter.close();
}
}
}
System.out.println("Elapsed " + (System.currentTimeMillis() - start) + " ms");
br.close();
fr.close();
in.close();
}
}
不要在循环中做任何你不需要做的事情。制作一张以文件名为关键字的文件地图,并让它们保持打开状态。根据需要给他们写信,在完成之前不要关闭任何一个。您也永远不需要冲洗它们。你正在击败所有好的缓冲。完成后,您可以考虑增大缓冲区。
您可以使用 Map
来跟踪状态文件而不是每次都关闭它们:
Map<String, FileWriter> fileMap = new HashMap<String, FileWriter>();
while ((line = br.readLine()) != null) {
if (!line.contains("-9999")) {
if (m.find()) {
stateFileName = m.group(1);
stateFileName = "States/" + stateFileName + ".csv";
FileWriter stateFileWriter = fileMap.get(stateFileName);
if (stateFileWriter == null) {
stateFileWriter = new FileWriter(stateFileName, true);
fileMap.put(stateFileName, stateFileWriter);
}
stateFileWriter.write(line + "\n");
}
}
}
// flush the writers and close once you have parsed the entire file
for(Map.Entry<String, FileWriter> entry : fileMap.entrySet()) {
FileWriter writer = entry.getValue();
writer.flush();
writer.close();
}