A Node.js module for parsing form data, especially file uploads.
[!CAUTION] As of April 2025, old versions like v1 and v2 are still the most
used, while they are deperecated for years -- they are also vulnerable to
attacks if you are not implementing it properly. Please upgrade! We are
here to help, and AI Editors & Agents could help a lot in such codemod-like
migrations.
[!TIP] If you are starting a fresh project, try formidable@latest (v3) or
you can check out the formidable-mini which is a super minimal version of
Formidable (not quite configurable yet, but when it does it could become the
basis for formidable@v4), using web standards like FormData API and File
API, and you can use it to stream uploads directly to S3 or other such
services.
Check
VERSION NOTES
for more information on v1, v2, and v3 plans, NPM dist-tags and branches.
This module was initially developed by
@felixge for
Transloadit, a service focused on uploading and
encoding images and videos. It has been battle-tested against hundreds of GBs of
file uploads from a large variety of clients and is considered production-ready
and is used in production for years.
Currently, we are few maintainers trying to deal with it. :) More contributors
are always welcome! :heart: Jump on
issue #412 which is
closed, but if you are interested we can discuss it and add you after strict
rules, like enabling Two-Factor Auth in your npm and GitHub accounts.
There are multiple variants to do this, but Formidable just need Node.js Request
stream, so something like the following example should work just fine, without
any third-party Express.js middleware.
Of course, with Koa v1, v2 or future v3 the things
are very similar. You can use formidable manually as shown below or through
the koa-better-body package which is
using formidable under the hood and support more features and different
request bodies, check its documentation for more info.
Note: this example is assuming Koa v2. Be aware that you should pass ctx.req
which is Node.js's Request, and NOT the ctx.request which is Koa's Request
object - there is a difference.
const Koa = require('koa');const formidable = require('formidable');const app = new Koa();app.on('error', (err) => { console.error('server error', err);});app.use(async (ctx, next) => { if (ctx.url === '/api/upload' && ctx.method.toLowerCase() === 'post') { const form = formidable({ multiples: true }); // not very elegant, but that's for now if you don't want to use `koa-better-body` // or other middlewares. await new Promise((resolve, reject) => { form.parse(ctx.req, (err, fields, files) => { if (err) { reject(err); return; } ctx.set('Content-Type', 'application/json'); ctx.status = 200; ctx.state = { fields, files }; ctx.body = JSON.stringify(ctx.state, null, 2); resolve(); }); }); await next(); return; } // show a file upload form ctx.set('Content-Type', 'text/html'); ctx.status = 200; ctx.body = ` <h2>With <code>"koa"</code> npm package</h2> <form action="/api/upload" enctype="multipart/form-data" method="post"> <div>Text field title: <input type="text" name="title" /></div> <div>File: <input type="file" name="koaFiles" multiple="multiple" /></div> <input type="submit" value="Upload" /> </form> `;});app.use((ctx) => { console.log('The next middleware is called'); console.log('Results:', ctx.state);});app.listen(3000, () => { console.log('Server listening on http://localhost:3000 ...');});
Benchmarks
The benchmark is quite old, from the old codebase. But maybe quite true though.
Previously the numbers was around ~500 mb/sec. Currently with moving to the new
Node.js Streams API it's faster. You can clearly see the differences between the
Node versions.
Note: a lot better benchmarking could and should be done in future.
options.minFileSize{number} - default 1 (1byte); the minium size of
uploaded file.
options.maxFileSize{number} - default 200 * 1024 * 1024 (200mb);
limit the size of uploaded file.
options.maxFields{number} - default 1000; limit the number of fields,
set 0 for unlimited
options.maxFieldsSize{number} - default 20 * 1024 * 1024 (20mb);
limit the amount of memory all fields together (except files) can allocate in
bytes.
options.hashAlgorithm{string | false} - default false; include
checksums calculated for incoming files, set this to some hash algorithm, see
crypto.createHash
for available algorithms
options.fileWriteStreamHandler{function} - default null, which by
default writes to host machine file system every file parsed; The function
should return an instance of a
Writable stream
that will receive the uploaded file data. With this option, you can have any
custom behavior regarding where the uploaded file data will be streamed for.
If you are looking to write the file uploaded in other types of cloud storages
(AWS S3, Azure blob storage, Google cloud storage) or private file storage,
this is the option you're looking for. When this option is defined the default
behavior of writing the file in the host machine file system is lost.
options.multiples{boolean} - default false; when you call the
.parse method, the files argument (of the callback) will contain arrays of
files for inputs which submit multiple files using the HTML5 multiple
attribute. Also, the fields argument will contain arrays of values for
fields that have names ending with '[]'.
options.filename{function} - default undefined Use it to control
newFilename. Must return a string. Will be joined with options.uploadDir.
options.filter{function} - default function that always returns true.
Use it to filter files before they are uploaded. Must return a boolean.
options.filename{function} function (name, ext, part, form) -> string
Note: If this size of combined fields, or size of some file is exceeded, an
'error' event is fired.
// The amount of bytes received for this form so far.form.bytesReceived;
// The expected number of bytes in this form.form.bytesExpected;
options.filter{function} function ({name, originalFilename, mimetype}) -> boolean
Note: use an outside variable to cancel all uploads upon the first error
You may overwrite this method if you are interested in directly accessing the
multipart stream. Doing so will disable any 'field' / 'file' events
processing which would occur otherwise, making you fully responsible for
handling the processing.
About uploadDir, given the following directory structure
__dirname would be the same directory as the source file itself (src)
`${__dirname}/../uploads`;
to put files in uploads.
Omitting __dirname would make the path relative to the current working
directory. This would be the same if server.js is launched from src but not
project-name.
null will use default which is os.tmpdir()
Note: If the directory does not exist, the uploaded files are silently
discarded. To make sure it exists:
import { createNecessaryDirectoriesSync } from 'filesac';const uploadPath = `${__dirname}/../uploads`;createNecessaryDirectoriesSync(`${uploadPath}/x`);
In the example below, we listen on couple of events and direct them to the
data listener, so you can do whatever you choose there, based on whether its
before the file been emitted, the header value, the header name, on field, on
file and etc.
Or the other way could be to just override the form.onPart as it's shown a bit
later.
A method that allows you to extend the Formidable library. By default we include
4 plugins, which esentially are adapters to plug the different built-in parsers.
The plugins added by this method are always enabled.
See src/plugins/ for more detailed look on default plugins.
The architecture is simple. The plugin is a function that is passed with the
Formidable instance (the form across the README examples) and the options.
Note: the plugin function's this context is also the same instance.
const formidable = require('formidable');const form = formidable({ keepExtensions: true });form.use((self, options) => { // self === this === form console.log('woohoo, custom plugin'); // do your stuff; check `src/plugins` for inspiration});form.parse(req, (error, fields, files) => { console.log('done!');});
Important to note, is that inside plugin this.options, self.options and
options MAY or MAY NOT be the same. General best practice is to always use the
this, so you can later test your plugin independently and more easily.
If you want to disable some parsing capabilities of Formidable, you can disable
the plugin which corresponds to the parser. For example, if you want to disable
multipart parsing (so the src/parsers/Multipart.js
which is used in src/plugins/multipart.js), then
you can remove it from the options.enabledPlugins, like so
const { Formidable } = require('formidable');const form = new Formidable({ hashAlgorithm: 'sha1', enabledPlugins: ['octetstream', 'querystring', 'json'],});
Be aware that the order MAY be important too. The names corresponds 1:1 to
files in src/plugins/ folder.
Pull requests for new built-in plugins MAY be accepted - for example, more
advanced querystring parser. Add your plugin as a new file in src/plugins/
folder (lowercased) and follow how the other plugins are made.
form.onPart
If you want to use Formidable to only handle certain parts for you, you can do
something similar. Or see
#387 for
inspiration, you can for example validate the mime-type.
const form = formidable();form.onPart = (part) => { part.on('data', (buffer) => { // do whatever you want here });};
For example, force Formidable to be used only on non-file "parts" (i.e., html
fields)
const form = formidable();form.onPart = function (part) { // let formidable handle only non-file parts if (part.originalFilename === '' || !part.mimetype) { // used internally, please do not override! form._handlePart(part); }};
File
export interface File { // The size of the uploaded file in bytes. // If the file is still being uploaded (see `'fileBegin'` event), // this property says how many bytes of the file have been written to disk yet. file.size: number; // The path this file is being written to. You can modify this in the `'fileBegin'` event in // case you are unhappy with the way formidable generates a temporary path for your files. file.filepath: string; // The name this file had according to the uploading client. file.originalFilename: string | null; // calculated based on options provided file.newFilename: string | null; // The mime type of this file, according to the uploading client. file.mimetype: string | null; // A Date object (or `null`) containing the time this file was last written to. // Mostly here for compatibility with the [W3C File API Draft](http://dev.w3.org/2006/webapi/FileAPI/). file.mtime: Date | null; file.hashAlgorithm: false | |'sha1' | 'md5' | 'sha256' // If `options.hashAlgorithm` calculation was set, you can read the hex digest out of this var (at the end it will be a string) file.hash: string | object | null;}
file.toJSON()
This method returns a JSON-representation of the file, allowing you to
JSON.stringify() the file which is useful for logging and responding to
requests.
Events
'progress'
Emitted after each incoming chunk of data that has been parsed. Can be used to
roll your own progress bar.
Emitted whenever a field / value pair has been received.
form.on('field', (name, value) => {});
'fileBegin'
Emitted whenever a new file is detected in the upload stream. Use this event if
you want to stream the file to somewhere else while buffering the upload on the
file system.
form.on('fileBegin', (formName, file) => { // accessible here // formName the name in the form (<input name="thisname" type="file">) or http filename for octetstream // file.originalFilename http filename or null if there was a parsing error // file.newFilename generated hexoid or what options.filename returned // file.filepath default pathnme as per options.uploadDir and options.filename // file.filepath = CUSTOM_PATH // to change the final path});
'file'
Emitted whenever a field / file pair has been received. file is an instance of
File.
form.on('file', (formname, file) => { // same as fileBegin, except // it is too late to change file.filepath // file.hash is available if options.hash was used});
'error'
Emitted when there is an error processing the incoming form. A request that
experiences an error is automatically paused, you will have to manually call
request.resume() if you want the request to continue firing 'data' events.
May have error.httpCode and error.code attached.
form.on('error', (err) => {});
'aborted'
Emitted when the request was aborted by the user. Right now this can be due to a
'timeout' or 'close' event on the socket. After this event is emitted, an
error event will follow. In the future there will be a separate 'timeout'
event (needs a change in the node core).
form.on('aborted', () => {});
'end'
Emitted when the entire request has been received, and all contained files have
finished flushing to disk. This is a great place for you to send your response.
Ryan Dahl for his work on
http-parser which heavily inspired the
initial multipart_parser.js.
Contributing
If the documentation is unclear or has a typo, please click on the page's Edit
button (pencil icon) and suggest a correction. If you would like to help us fix
a bug or add a new feature, please check our Contributing
Guide. Pull requests are welcome!
Thanks goes to these wonderful people
(emoji key):