簡介
工作的時候,需要將一個大的文本文件按行分割成幾個小文件。本來懶得寫,想網上copy一下得了,但是 google 了一遍,找了幾個代碼寫的有點亂,嘗試了之後發現效率太慢了,一個 1000000 行 200M 的文件,按每個文件 2000 行分割,要6分多鐘才能跑完。沒辦法自己寫了個,試了幾次,基本都是 4 秒內跑完,貼出來記錄下,下次用就直接 copy 出來用。
代碼
public static List<File> splitDataToSaveFile(int rows, File sourceFile, String targetDirectoryPath) {
long startTime = System.currentTimeMillis();
List<File> fileList = new ArrayList<>();
log.info("開始分割文件");
File targetFile = new File(targetDirectoryPath);
if (!sourceFile.exists() || rows <= 0 || sourceFile.isDirectory()) {
return null;
}
if (targetFile.exists()) {
if (!targetFile.isDirectory()) {
return null;
}
} else {
targetFile.mkdirs();
}
try (FileInputStream fileInputStream = new FileInputStream(sourceFile);
InputStreamReader inputStreamReader = new InputStreamReader(fileInputStream, StandardCharsets.UTF_8);
BufferedReader bufferedReader = new BufferedReader(inputStreamReader)) {
StringBuilder stringBuilder = new StringBuilder();
String lineStr;
int lineNo = 1, fileNum = 1;
while ((lineStr = bufferedReader.readLine()) != null) {
stringBuilder.append(lineStr).append("\r\n");
if (lineNo % rows == 0) {
File file = new File(targetDirectoryPath + File.separator + fileNum + sourceFile.getName());
writeFile(stringBuilder.toString(), file);
//清空文本
stringBuilder.delete(0, stringBuilder.length());
fileNum++;
fileList.add(file);
}
lineNo++;
}
if ((lineNo - 1) % rows != 0) {
File file = new File(targetDirectoryPath + File.separator + fileNum + sourceFile.getName());
writeFile(stringBuilder.toString(), file);
fileList.add(file);
}
long endTime = System.currentTimeMillis();
log.info("分割文件結束,耗時:{}秒", (endTime - startTime) / 1000);
} catch (Exception e) {
log.error("分割文件異常", e);
}
return fileList;
}
private static void writeFile(String text, File file) {
try (
FileOutputStream fileOutputStream = new FileOutputStream(file);
OutputStreamWriter outputStreamWriter = new OutputStreamWriter(fileOutputStream, StandardCharsets.UTF_8);
BufferedWriter bufferedWriter = new BufferedWriter(outputStreamWriter, 1024)
) {
bufferedWriter.write(text);
} catch (IOException e) {
e.printStackTrace();
}
}