Skip to content

Latest commit

 

History

History
275 lines (224 loc) · 9.66 KB

ContentTypeParser.md

File metadata and controls

275 lines (224 loc) · 9.66 KB

Fastify

Content-Type Parser

Natively, Fastify only supports 'application/json' and 'text/plain' content types. If the content type is not one of these, an FST_ERR_CTP_INVALID_MEDIA_TYPE error will be thrown. Other common content types are supported through the use of plugins.

The default charset is utf-8. If you need to support different content types, you can use the addContentTypeParser API. The default JSON and/or plain text parser can be changed or removed.

Note: If you decide to specify your own content type with the Content-Type header, UTF-8 will not be the default. Be sure to include UTF-8 like this text/html; charset=utf-8.

As with the other APIs, addContentTypeParser is encapsulated in the scope in which it is declared. This means that if you declare it in the root scope it will be available everywhere, while if you declare it inside a plugin it will be available only in that scope and its children.

Fastify automatically adds the parsed request payload to the Fastify request object which you can access with request.body.

Note that for GET and HEAD requests the payload is never parsed. For OPTIONS and DELETE requests the payload is only parsed if the content type is given in the content-type header. If it is not given, the catch-all parser is not executed as with POST, PUT and PATCH, but the payload is simply not parsed.

⚠ Security Notice

When using with RegExp to detect Content-Type, you should beware of how to properly detect the Content-Type. For example, if you need application/*, you should use /^application\/([\w-]+);?/ to match the essence MIME type only.

Usage

fastify.addContentTypeParser('application/jsoff', function (request, payload, done) {
  jsoffParser(payload, function (err, body) {
    done(err, body)
  })
})

// Handle multiple content types with the same function
fastify.addContentTypeParser(['text/xml', 'application/xml'], function (request, payload, done) {
  xmlParser(payload, function (err, body) {
    done(err, body)
  })
})

// Async is also supported in Node versions >= 8.0.0
fastify.addContentTypeParser('application/jsoff', async function (request, payload) {
  const res = await jsoffParserAsync(payload)

  return res
})

// Handle all content types that matches RegExp
fastify.addContentTypeParser(/^image\/([\w-]+);?/, function (request, payload, done) {
  imageParser(payload, function (err, body) {
    done(err, body)
  })
})

// Can use default JSON/Text parser for different content Types
fastify.addContentTypeParser('text/json', { parseAs: 'string' }, fastify.getDefaultJsonParser('ignore', 'ignore'))

Fastify first tries to match a content-type parser with a string value before trying to find a matching RegExp. If you provide overlapping content types, Fastify tries to find a matching content type by starting with the last one passed and ending with the first one. So if you want to specify a general content type more precisely, first specify the general content type and then the more specific one, like in the example below.

// Here only the second content type parser is called because its value also matches the first one
fastify.addContentTypeParser('application/vnd.custom+xml', (request, body, done) => {} )
fastify.addContentTypeParser('application/vnd.custom', (request, body, done) => {} )

// Here the desired behavior is achieved because fastify first tries to match the
// `application/vnd.custom+xml` content type parser
fastify.addContentTypeParser('application/vnd.custom', (request, body, done) => {} )
fastify.addContentTypeParser('application/vnd.custom+xml', (request, body, done) => {} )

Using addContentTypeParser with fastify.register

When using addContentTypeParser in combination with fastify.register, await should not be used when registering routes. Using await causes the route registration to be asynchronous and can lead to routes being registered before the addContentTypeParser has been set.

Correct Usage

const fastify = require('fastify')();


fastify.register((fastify, opts) => {
  fastify.addContentTypeParser('application/json', function (request, payload, done) {
    jsonParser(payload, function (err, body) {
      done(err, body)
    })
  })

  fastify.get('/hello', async (req, res) => {});
});

Besides the addContentTypeParser API there are further APIs that can be used. These are hasContentTypeParser, removeContentTypeParser and removeAllContentTypeParsers.

hasContentTypeParser

You can use the hasContentTypeParser API to find if a specific content type parser already exists.

if (!fastify.hasContentTypeParser('application/jsoff')){
  fastify.addContentTypeParser('application/jsoff', function (request, payload, done) {
    jsoffParser(payload, function (err, body) {
      done(err, body)
    })
  })
}

removeContentTypeParser

With removeContentTypeParser a single or an array of content types can be removed. The method supports string and RegExp content types.

fastify.addContentTypeParser('text/xml', function (request, payload, done) {
  xmlParser(payload, function (err, body) {
    done(err, body)
  })
})

// Removes the both built-in content type parsers so that only the content type parser for text/html is available
fastify.removeContentTypeParser(['application/json', 'text/plain'])

removeAllContentTypeParsers

In the example from just above, it is noticeable that we need to specify each content type that we want to remove. To solve this problem Fastify provides the removeAllContentTypeParsers API. This can be used to remove all currently existing content type parsers. In the example below we achieve the same as in the example above except that we do not need to specify each content type to delete. Just like removeContentTypeParser, this API supports encapsulation. The API is especially useful if you want to register a catch-all content type parser that should be executed for every content type and the built-in parsers should be ignored as well.

fastify.removeAllContentTypeParsers()

fastify.addContentTypeParser('text/xml', function (request, payload, done) {
  xmlParser(payload, function (err, body) {
    done(err, body)
  })
})

Notice: The old syntaxes function(req, done) and async function(req) for the parser are still supported but they are deprecated.

Body Parser

You can parse the body of a request in two ways. The first one is shown above: you add a custom content type parser and handle the request stream. In the second one, you should pass a parseAs option to the addContentTypeParser API, where you declare how you want to get the body. It could be of type 'string' or 'buffer'. If you use the parseAs option, Fastify will internally handle the stream and perform some checks, such as the maximum size of the body and the content length. If the limit is exceeded the custom parser will not be invoked.

fastify.addContentTypeParser('application/json', { parseAs: 'string' }, function (req, body, done) {
  try {
    const json = JSON.parse(body)
    done(null, json)
  } catch (err) {
    err.statusCode = 400
    done(err, undefined)
  }
})

See example/parser.js for an example.

Custom Parser Options
  • parseAs (string): Either 'string' or 'buffer' to designate how the incoming data should be collected. Default: 'buffer'.
  • bodyLimit (number): The maximum payload size, in bytes, that the custom parser will accept. Defaults to the global body limit passed to the Fastify factory function.

Catch-All

There are some cases where you need to catch all requests regardless of their content type. With Fastify, you can just use the '*' content type.

fastify.addContentTypeParser('*', function (request, payload, done) {
  let data = ''
  payload.on('data', chunk => { data += chunk })
  payload.on('end', () => {
    done(null, data)
  })
})

Using this, all requests that do not have a corresponding content type parser will be handled by the specified function.

This is also useful for piping the request stream. You can define a content parser like:

fastify.addContentTypeParser('*', function (request, payload, done) {
  done()
})

and then access the core HTTP request directly for piping it where you want:

app.post('/hello', (request, reply) => {
  reply.send(request.raw)
})

Here is a complete example that logs incoming json line objects:

const split2 = require('split2')
const pump = require('pump')

fastify.addContentTypeParser('*', (request, payload, done) => {
  done(null, pump(payload, split2(JSON.parse)))
})

fastify.route({
  method: 'POST',
  url: '/api/log/jsons',
  handler: (req, res) => {
    req.body.on('data', d => console.log(d)) // log every incoming object
  }
})

For piping file uploads you may want to check out this plugin.

If you want the content type parser to be executed on all content types and not only on those that don't have a specific one, you should call the removeAllContentTypeParsers method first.

// Without this call, the request body with the content type application/json would be processed by the built-in JSON parser
fastify.removeAllContentTypeParsers()

fastify.addContentTypeParser('*', function (request, payload, done) {
  const data = ''
  payload.on('data', chunk => { data += chunk })
  payload.on('end', () => {
    done(null, data)
  })
})