CODEWITHSUNDEEP
javaencodingiofile-handling
Michael Hillman
Michael Hillman

Reputation: 1277

Reading and writing file in ISO-8859-1 encoding?

I have file encoded in ISO-8859-1. I'm trying to read it in as a single String, do some regex substitutions on it, and write it back out in the same encoding.

However, the resulting file I get always seems to be UTF-8 (according to Notepad++ at least), mangling some characters.

Can anyone see what I'm doing wrong here?

private static void editFile(File source, File target) {

    // Source and target encoding
    Charset iso88591charset = Charset.forName("ISO-8859-1");

    // Read the file as a single string
    String fileContent = null;

    try (Scanner scanner = new Scanner(source, iso88591charset)) {
    
        fileContent = scanner.useDelimiter("\\Z").next();
                
    } catch (IOException exception) {
        LOGGER.error("Could not read input file as a single String.", exception);
        return;
    }

    // Do some regex substitutions on the fileContent string
    String newContent = regex(fileContent);

    // Write the file back out in target encoding
    try (BufferedWriter writer = new BufferedWriter(new OutputStreamWriter(new FileOutputStream(target), iso88591charset))) {
    
        writer.write(newContent);
        
    } catch (Exception exception) {
        LOGGER.error("Could not write out edited file!", exception);
    }
}

Upvotes: 0

Views: 3167

Answers (1)

Jonasz
Jonasz

Reputation: 1802

There is nothing actually wrong with your code. Notepad++ sees the file encoded in UTF-8 because on a basic level there is no difference between UTF-8 and the encoding you're trying to use. Only specific characters differ and some (a lot) are missing from ISO compared to UTF. You can read more here or by simply searching ISO-8859-1 vs UTF-8 in Google.

I've created a simple project with your code and tested it with characters that are different for the ISO encoding - the result is a file that IntelliJ (and probably Notepad++ as well - cannot easily check, I'm on Linux) recognizes as ISO-8859-1. Apart from that, I've added another class that makes use of new (JDK11) features from Files class. The new Scanner(source, charset) that you've used was added in JDK10, so I think that you may be using 11 already. Here's the simplified code:

private static void editFile(File source, File target) {
    Charset charset = StandardCharsets.ISO_8859_1;
    String fileContent;
    try {
        fileContent = Files.readString(source.toPath(), charset);
    } catch (IOException exception) {
        System.err.println("Could not read input file as a single String.");
        exception.printStackTrace();
        return;
    }
    String newContent = regex(fileContent);
    try {
        Files.writeString(target.toPath(), newContent, charset);
    } catch (IOException exception) {
        System.err.println("Could not write out edited file!");
        exception.printStackTrace();
    }
}

Feel free to clone the repository or check it on GitHub and use whichever code version you prefer.

Upvotes: 2

Related Questions