MassUpload

Lets the user upload lots of huge files.

Why?

There are lots of libraries that do this already, but they're missing a key question: how do we upload a file to the server? This is important: as you go beyond tiny files, multipart/form-data just won't cut it.

MassUpload lets the developer decide. So you, being a clever developer, can upload several gigabytes' worth of data and resume when interrupted.

Usage

First, include the library. (It depends on Backbone.) You may use RequireJs:

require([ 'path/to/mass-upload' ], function(MassUpload) { ... });

Or you can use Browserify or plain Node require:

var MassUpload = require('js-mass-upload');

Now you'll need to implement a few asynchronous functions. The return values are ignored; instead, these functions are expected to call the appropriate callbacks that are passed to them.

var app = new MassUpload({
  doListFiles: function(progress, done) { ... },
  doUploadFile: function(upload, progress, done) { ... },
  doDeleteFile: function(upload, done) { ... },
});

app.on('change:status', function() { ... });
app.on('change:listFilesProgress', function() { ... });
app.on('change:uploadProgress', function() { ... });

$('form#mass-upload').prepend(app.el);

app.fetchFileInfosFromServer() // kicks everything off

Finally, write your UI. It may iterate over MassUpload.uploads, a list of files; onStateChanged will be called whenever the list changes (because the total size of the upload will change).

Broad ideas

If you're going to support resume (or even if you aren't but you want to keep yourself open to the possibility some day), then the user interface could get tricky: how do you get users to re-select the files they selected last time they were on this page?

The answer is to be transparent:

• Keep a list on the server of what the user has uploaded.
• Show that list to the user.
• Let the user edit that list, by deleting files or adding new files.

MassUpload does this.

Client-side API

You'll need to implement a few functions and pass them in the MassUpload app's constructor.

But first let's define some datatypes:

• File: See the W3 File API. This represents a file on the client side. The web browser creates this object.
• ProgressEvent: See the W3 Progress Events API.
• MassUpload: an object containing:
  • state (a State)
  • uploads (an Array of Upload objects)
  • deleteUpload(upload): deletes an upload
• MassUpload.FileInfo: depicts a file on the server. An object containing:
  • name (a String)
  • lastModified (an integer Number of milliseconds since Unix epoch)
  • total (an integer Number depicting total size)
  • loaded (an integer Number less than or equal to size)
• MassUpload.Upload: depicts a single object which either is on the server or should be on the server. An object containing:
  • file (a File, if the user has selected this file on the client)
  • fileInfo (a FileInfo, if the file is on the server, either partially uploaded or completely uploaded)
  • error (an Error or null)
  • isComplete(): true iff fileInfo.loaded === fileInfo.total
• MassUpload.State: an immutable object containing:
  • loaded and total (integer Numbers, like in a ProgressEvent)
  • status ("listing", "uploading" or "waiting")
  • errors (an Array of Error objects)
  • isComplete(): true iff state === "waiting" && loaded === total && total > 0 && errors.length === 0
• MassUpload.Error: an error that prevents this upload from going smoothly. Errors can be removed by retrying operations or removing the files that they apply to. They are immutable and contain:
  • failedCall: "listFiles", "uploadFile" or "deleteFile"
  • failedCallArgument: undefined, a File or a FileInfo, as appropriate
  • detail: a raw error object (you define what that should be; perhaps an XMLHttpRequest)

Got that? Great. Here's what you need to implement:

---

function doListFiles(progress, done) { ... }

Lists files already uploaded to the server. Called when initializing.

• Calls progress(progressEvent), where progressEvent is a ProgressEvent, during loading.
• Calls done(null, data), where data is an Array of FileInfo objects, when loading is complete.
• Calls done(error), where error is an Error (perhaps with a .xhr property), if listing fails.

Parameters:

• progress: To be called during loading.
• done: To be called on error or after completion.

---

function doUploadFile(upload, progress, done) { ... }

Uploads a file to the server. Called asynchronously.

• Calls progress(progressEvent), where progressEvent is a ProgressEvent, during loading.
• Calls done(null) when upload is complete.
• Calls done(error), where error is an Error (perhaps with a .xhr property), if listing fails.

Parameters:

• upload: An Upload. Call upload.get('file') to get its File. The File may already be partially uploaded to the server.
• progress: To be called during upload.
• done: To be called on error or after completion.

Returns: a function that, when called, aborts the operation as soon as possible. (The abort function should asynchronously call done(new Error('aborted')).)

---

function doDeleteFile(upload, success, error) { ... }

Deletes a file from the server. Called asynchronously.

• Calls done(null) when the delete is complete.
• Calls done(error), where error is an Error (perhaps with a .xhr property), if deleting fails.

Parameters:

• upload: An Upload. Call upload.get('fileInfo') to get its FileInfo object.
• done: To be called on error or after completion.

---

function onStateChanged(state, massUpload) { ... }

Does something when the state changes.

In practical terms, callbacks should listen and:

• Display global progress.
• Display error messages (both global and alongside File and FileInfo representations).
• Prompt the user to resolve errors.
• Submit the form when upload is complete.

Parameters:

• state, a State.
• massUpload, a MassUpload.

Talking with a server

Let's portray an example web service in which the user owns a bunch of folders. For each folder, the user owns some files.

Let's give each file a GUID based on its filename, size and lastModified.

Let's also assume we want to be able to resume uploads. We can't do that with multipart/form-data; instead, we'll upload raw binary blobs.

Our server API might look something like this:

Method	URL	Parameters	Description
GET	/folders		Lists folders
POST	/folders	name, permalink	Creates a folder
DELETE	/folders/:permalink		Deletes a folder
PUT	/folders/:permalink	name	Renames a folder
GET	/folders/:permalink/files		Lists a folder's contents
DELETE	/folders/:permalink/files		Deletes all files from the folder
PUT	/folders/:permalink/files/:guid	name, size, lastModified	Creates an empty file with the given GUID
HEAD	/folders/:permalink/files/:guid		Describes how much of the file is uploaded, using a Content-Range header.
PATCH	/folders/:permalink/files/:guid	(raw bytes)	Uploads the file's raw bytes using the passed Content-Range header. Fails if the Content-Range would create a gap in the file or go past the end of the file; succeeds once the file is completely uploaded.
DELETE	/folders/:permalink/files/:guid		Deletes the file from the server

Then our client-side API methods should do something like this:

Function	Behavior
doListFiles(progress, done)	Mirrors GET /folders/:permalink/files
doUploadFile(file, progress, done)	Determines guid; calls HEAD /folders/:permalink/files/:guid, optionally calls PUT /folders/:permalink/files/:guid if the file does not exist; calculates the missing Content-Range and calls file.slice() to create a Blob starting at that range; calls PATCH /folders/:permalink/files/:guid sending the Blob. (In jQuery.ajax(), you can send a blob using processData: false.)
doDeleteFile(fileInfo, done)	Determines guid; calls DELETE /folders/:permalink/files/:guid; calls success if the server responds with success or 404, otherwise calls error.

Contributing

To build and contribute:

1. Install NodeJS
2. Clone this repository: git clone https://github.com/overview/js-mass-upload.git
3. npm install in this directory
4. gulp to build
5. gulp test to unit-test
6. Add a test; return to step 4; make test pass; return to step 4
7. git commit and create a pull request

License

Copyright 2013 The Overview Project

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.