Can you use GridFS to store large zip files?

[Khalid]

I have stored zip files in GridFS, but I could not retrieve the whole ZIP file. How can pull it from GridFS chunks collection into the file system using the Python driver? Is there a way to serve zip files in GridFS instead of loading the chunks in the memory? 

I looked at the examples in the documentation but they are too simple and do not show the power of GridFS.

If this is a limitation in the Python driver, can you show me how to do it in other drivers (Java or Ruby)?

[Khalid]

It turned out it was straightforward to do in Java. Just get the input stream and copy it into a new file using Apache IOUtils.

GridFSDBFile zipFile = gridGfs.find(new ObjectId("539edfcc7836782a6bd2970b"));
InputStream inputStream = zipFile.getInputStream();
File file = new File("/Users/Khalid/file.zip");
if (!file.exists()) {file.createNewFile();}

FileOutputStream fileOutputStream = new FileOutputStream(file);

IOUtils.copy(inputStream, fileOutputStream);

[Alex at Ikanow]

Seems like you have a working solution, but in case it was useful I just wanted to mention a utility I wrote and open-sourced:

https://github.com/Alex-Ikanow/GridFSZipFile

This provides a zip file interface that backs directly onto the GridFS, ie you can extract individual files from within the zip archive without having to copy the whole thing to disk first:

    import com.ikanow.utility.GridFSRandomAccessFile;
    import net.sf.jazzlib.GridFSZipFile;
    //+MongoDB imports
    String dbName = "test";
    DBCollection db = new MongoClient().getDB(dbName);
    String fsName = "fs"; // (the default)
    ObjectId fileId = new ObjectId("FILEID");
    GridFSRandomAccessFile shareAsFile = new GridFSRandomAccessFile(db, fsName, fileId);
    net.sf.jazzlib.GridFSZipFile zipFile = new net.sf.jazzlib.GridFSZipFile("dummy", shareAsFile);
    //Then use the zipFile object however you would use a java.util.zip.ZipFile...