java判断文件编码

最新推荐文章于 2024-07-03 02:34:55 发布

_小海_

最新推荐文章于 2024-07-03 02:34:55 发布

阅读量1.1k

点赞数

分类专栏： java 文章标签： linux java

java 专栏收录该内容

20 篇文章 0 订阅

订阅专栏

/**
* 获得远程URL文件的编码格式
*/
public static String getReomoteURLFileEncode(URL url) {
   CodepageDetectorProxy detector = CodepageDetectorProxy.getInstance();
   detector.add(new ParsingDetector(false));
   detector.add(JChardetFacade.getInstance());
   detector.add(ASCIIDetector.getInstance());
   detector.add(UnicodeDetector.getInstance());
   java.nio.charset.Charset charset = null;
   try {
    System.out.println(url);
    charset = detector.detectCodepage(url);
   } catch (Exception ex) {
    ex.printStackTrace();
   }
   if (charset != null) {
    return charset.name();
   } else {
    return "utf-8";
   }
}


/**
* 获得文件流的编码格式
*/
public static String getInputStreamEncode(InputStream is) {
   CodepageDetectorProxy detector = CodepageDetectorProxy.getInstance();
   detector.add(new ParsingDetector(false));
   detector.add(JChardetFacade.getInstance());
   detector.add(ASCIIDetector.getInstance());
   detector.add(UnicodeDetector.getInstance());
   java.nio.charset.Charset charset = null;
   try {
    charset = detector.detectCodepage(is, 0);
   } catch (Exception ex) {
    ex.printStackTrace();
   }
   if (charset != null) {
    return charset.name();
   } else {
    return "utf-8";
   }
}

/**
* 获得本地文件的编码格式
*/
public static String getLocalteFileEncode(String filePath) {
   CodepageDetectorProxy detector = CodepageDetectorProxy.getInstance();
   detector.add(new ParsingDetector(false));
   detector.add(JChardetFacade.getInstance());
   detector.add(ASCIIDetector.getInstance());
   detector.add(UnicodeDetector.getInstance());
   java.nio.charset.Charset charset = null;
   File file = new File(filePath);
   try {
    charset = detector.detectCodepage(file.toURI().toURL());
   } catch (Exception ex) {
    ex.printStackTrace();
   }
   if (charset != null) {
    return charset.name();
   } else {
    return "utf-8";
   }
}
/**
* 获得字符串的编码格式
*/
public static String getStringEncode(String str) {
   CodepageDetectorProxy detector = CodepageDetectorProxy.getInstance();
   detector.add(new ParsingDetector(false));
   detector.add(JChardetFacade.getInstance());
   detector.add(ASCIIDetector.getInstance());
   detector.add(UnicodeDetector.getInstance());
   java.nio.charset.Charset charset = null;
   InputStream myIn=new ByteArrayInputStream(str.getBytes());
   try {
    charset = detector.detectCodepage(myIn,3);
   } catch (Exception ex) {
    ex.printStackTrace();
   }
   if (charset != null) {
    return charset.name();
   } else {
    return "utf-8";
   }
}