Serving files from the content API

[oeway]

As discussed with @Zsailer , the current get content api returns a json object which requires extra parsing in order to be used.

One thing we want to have in ImJoy is an endpoint that can serve the files directly with the correct MIME type in the Content-Type header so the browser can render these files directly, like the /view endpint from the notebook(but it doesn't always give the correct mime type).

I think I can build a server extension for it, but definitely nice to have it built-in with the content api.

[Zsailer]

Pinging @saulshanabrook and @telamonian. They've been thinking about the future of the contents API quite a bit, so I'd like to loop them in here.

[saulshanabrook]

Sounds generally sensible to me on another endpoint.

[krassowski]

How would that differ from existing /files/ endpoint?

jupyter_server/jupyter_server/files/handlers.py

Lines 14 to 79 in 606487a

	class FilesHandler(JupyterHandler):
	"""serve files via ContentsManager
	Normally used when ContentsManager is not a FileContentsManager.
	FileContentsManager subclasses use AuthenticatedFilesHandler by default,
	a subclass of StaticFileHandler.
	"""
	@property
	def content_security_policy(self):
	# In case we're serving HTML/SVG, confine any Javascript to a unique
	# origin so it can't interact with the notebook server.
	return super(FilesHandler, self).content_security_policy + "; sandbox allow-scripts"
	@web.authenticated
	def head(self, path):
	self.get(path, include_body=False)
	self.check_xsrf_cookie()
	return self.get(path, include_body=False)
	@web.authenticated
	async def get(self, path, include_body=True):
	# /files/ requests must originate from the same site
	self.check_xsrf_cookie()
	cm = self.contents_manager
	if await ensure_async(cm.is_hidden(path)) and not cm.allow_hidden:
	self.log.info("Refusing to serve hidden file, via 404 Error")
	raise web.HTTPError(404)
	path = path.strip("/")
	if "/" in path:
	_, name = path.rsplit("/", 1)
	else:
	name = path
	model = await ensure_async(cm.get(path, type="file", content=include_body))
	if self.get_argument("download", False):
	self.set_attachment_header(name)
	# get mimetype from filename
	if name.lower().endswith(".ipynb"):
	self.set_header("Content-Type", "application/x-ipynb+json")
	else:
	cur_mime = mimetypes.guess_type(name)[0]
	if cur_mime == "text/plain":
	self.set_header("Content-Type", "text/plain; charset=UTF-8")
	elif cur_mime is not None:
	self.set_header("Content-Type", cur_mime)
	else:
	if model["format"] == "base64":
	self.set_header("Content-Type", "application/octet-stream")
	else:
	self.set_header("Content-Type", "text/plain; charset=UTF-8")
	if include_body:
	if model["format"] == "base64":
	b64_bytes = model["content"].encode("ascii")
	self.write(decodebytes(b64_bytes))
	elif model["format"] == "json":
	self.write(json.dumps(model["content"]))
	else:
	self.write(model["content"])
	self.flush()

I now it is not a part of the contents API, but would it be a good idea to include it in the documentation?

[blink1073]

Closing as answered, the documentation issue is tracked in #298