If c.S3ContentsManager.s3_base_uri = 's3://s3nb/', no notebooks will be found.
If c.S3ContentsManager.s3_base_uri = 's3://s3nb/notebooks/', if functions as expected.

Seems that NamedTemporaryFiles under Windows are opened automatically. This causes issues when codec.open is then used to open the file in the IPython V3 code. I changed the code to use the following method as opposed to codecs_open in the save and get methods.

Am happy to integrate it in if need be.

Thanks for this work.

Regards
Paul

     def _codecs_open(self, filename, file_handle, mode='rb', encoding=None, errors='strict', buffering=1):
        # Copy of codecs.open
        if encoding is not None:
            if 'U' in mode:
                # No automatic conversion of '\n' is done on reading and writing
                mode = mode.strip().replace('U', '')
                if mode[:1] not in set('rwa'):
                    mode = 'r' + mode
            if 'b' not in mode:
                # Force opening of the file in binary mode
                mode = mode + 'b'
        try:
            file = open(filename, mode, buffering)
        except:
            file = file_handle
        if encoding is None:
            return file
        info = codecs.lookup(encoding)
        srw = codecs.StreamReaderWriter(file, info.streamreader, info.streamwriter, errors)
        # Add attributes to simplify introspection
        srw.encoding = encoding
        return srw

Save was changed as follows (get the same)

 def _save_notebook(self, path, nb):
        self.log.debug('_save_notebook: %s', locals())

        k = boto.s3.key.Key(self.bucket)
        k.key = self._path_to_s3_key(path)

        try:
            with tempfile.NamedTemporaryFile() as t:
                with self._codecs_open(t.name, t, mode='w', encoding='utf-8') as f:
                    # write tempfile with utf-8 encoding
                    nbformat.write(nb, f, version=nbformat.NO_CONVERT)
                    # upload as bytes (t's fp didn't advance)
                    t.seek(0)
                    k.set_contents_from_file(t)
        except Exception as e:
            raise web.HTTPError(400, u"Unexpected Error Writing Notebook: %s %s" % (path, e))

Sometimes (depending on many factors) saving to S3 can produce a noticeable delay.

Saving in a background thread would prevent this, I think? @raphaellig?

When duplicating a notebook, it just opens the same notebook in a new tab. No copy is created in S3.

@krismolendyke

404: Not Found

This error comes out:

{"reason": null, "traceback": "Traceback (most recent call last):\n File \"/usr/lib/python2.7/site-packages/IPython/html/base/handlers.py\", line 441, in wrapper\n result = yield gen.maybe_future(method(self, *args, **kwargs))\n File \"/usr/lib64/python2.7/site-packages/tornado/gen.py\", line 1015, in run\n value = future.result()\n File \"/usr/lib64/python2.7/site-packages/tornado/concurrent.py\", line 237, in result\n raise_exc_info(self._exc_info)\n File \"/usr/lib64/python2.7/site-packages/tornado/gen.py\", line 285, in wrapper\n yielded = next(result)\n File \"/usr/lib/python2.7/site-packages/IPython/html/services/contents/handlers.py\", line 126, in get\n path=path, type=type, format=format, content=content,\n File \"/usr/lib/python2.7/site-packages/s3nb-0.0.4.37-py2.7.egg/s3nb/ipy3.py\", line 190, in get\n model = self._s3_key_file_to_model(k, timeformat=S3_TIMEFORMAT_GET_KEY)\n File \"/usr/lib/python2.7/site-packages/s3nb-0.0.4.37-py2.7.egg/s3nb/ipy3.py\", line 63, in _s3_key_file_to_model\n self.log.debug(\"_s3_key_file_to_model: %s: %s\", key, key.name)\nAttributeError: 'NoneType' object has no attribute 'name'\n", "message": "Unknown server error"}

I seem to be getting a "method not allowed" error when I attempt to rename folders. I can do other things, such as create folders and notebooks, so I don't believe permissions are an issue. Thanks for building this!