unzipping ePub file doesn't work

Question

I am developping an app to decompress an .epub file into SDCARD in Android. I already read the Can't Unzip EPub File TOPIC. IT WORKED FOR .zip files but not for .epub files. Can somone tell me where is the problem ? here is the exception log:

03-21 13:35:44.281: W/System.err(1255): java.io.FileNotFoundException: /mnt/sdcard/unzipped11/META-INF/container.xml: open failed: ENOENT (No such file or directory)

I am using this code:

private void decom() throws IOException {
    ZipFile zipFile = new ZipFile(Environment.getExternalStorageDirectory()+"/dir.zip");
    String path = Environment.getExternalStorageDirectory() + "/unzipped10/";

    Enumeration<?> files = zipFile.entries();
    _dirChecker("");
    while (files.hasMoreElements()) {
        ZipEntry entry = (ZipEntry) files.nextElement();
        Log.v("ZipEntry", ""+entry);
        Log.v("isDirectory", ""+entry.isDirectory());

        if (entry.isDirectory()) {
            File file = new File(path + entry.getName());
            file.mkdir();
            System.out.println("Create dir " + entry.getName());
        } else {
            File f = new File(path + entry.getName());
            FileOutputStream fos = new FileOutputStream(f);
            InputStream is = zipFile.getInputStream(entry);
            byte[] buffer = new byte[1024];
            int bytesRead = 0;
            while ((bytesRead = is.read(buffer)) != -1) {
                fos.write(buffer, 0, bytesRead);
            }
            fos.close();
            System.out.println("Create File " + entry.getName());
        }
    }
}

Dunkirk

Do you get the log output that it creates the META-INF directory before you get the log output that it's attempting to create the file META-INF/container.xml? – Flavio

no and i don't know why – Dunkirk

it doesn't even detect it's a directory – Dunkirk

Answer 1

Based on your response to my comment, it sounds like the parent directory to the file is not created before the entry for the file in the archive is attempted to be written.

It sounds like you may need to alter the code that deals with a file entry in the zip file to create parent directories if they do not yet exist. You may also need to alter the code which creates the directories to check if the directory already exists before creating it.

Try something like this:

    while (files.hasMoreElements()) {
        ZipEntry entry = (ZipEntry) files.nextElement();
        Log.d(TAG, "ZipEntry: "+entry);
        Log.d(TAG, "isDirectory: " + entry.isDirectory());

        if (entry.isDirectory()) {
            File file = new File(path + entry.getName());
            file.mkdir();
            Log.d(TAG, "Create dir " + entry.getName());
        } else {
            File f = new File(path + entry.getName());
            f.getParentFile().mkdirs();
            FileOutputStream fos = new FileOutputStream(f);
            InputStream is = zipFile.getInputStream(entry);
            byte[] buffer = new byte[1024];
            int bytesRead = 0;
            while ((bytesRead = is.read(buffer)) != -1) {
                fos.write(buffer, 0, bytesRead);
            }
            fos.close();
            Log.d(TAG, "Create File " + entry.getName());
        }
    }
    Log.d(TAG, "Done extracting epub file");

For me this produces the following output using a test epub (moby dick from Google's epub samples: https://code.google.com/p/epub-samples/downloads/list)

ZipEntry: mimetype
isDirectory: false
Create File mimetype
ZipEntry: META-INF/
isDirectory: true
Create dir META-INF/
ZipEntry: META-INF/container.xml
isDirectory: false
Create File META-INF/container.xml
ZipEntry: OPS/
isDirectory: true
Create dir OPS/
ZipEntry: OPS/chapter_001.xhtml
isDirectory: false
Create File OPS/chapter_001.xhtml
ZipEntry: OPS/chapter_002.xhtml
isDirectory: false
Create File OPS/chapter_002.xhtml
ZipEntry: OPS/chapter_003.xhtml
isDirectory: false
...
Create File OPS/toc-short.xhtml
ZipEntry: OPS/toc.xhtml
isDirectory: false
Create File OPS/toc.xhtml
Done extracting epub file

Flavio

it doesn't work, (by the way, i'am working with the Virtual Device) – Dunkirk

Can you give me more information about that? Is it the same error as before? – Flavio

I just updated the code sample to fix a problem (moved the mkdirs() call up one line) and I changed the logging over to use Android's logging. I added the output I got from a sample Android app which shows that I was able to extract an epub with the provided code. – Flavio

If it still doesn't work with this code, you might try running epubcheck against your epub. github.com/IDPF/epubcheck or validator.idpf.org – Flavio