Parse-request NPM

parse-request

Parse requests in the Browser and Node (with added support for multer and passport). Made for Cabin.

Install

npm install parse-request

How does it work

This package is used internally by Cabin's middleware, and we highly recommend you to simply use Cabin instead of this package in particular.

This package exports a function that accepts an Object options argument:

options (Object) - a configuration object
- req (Object) - an Express/Connect request object (defaults to false) (you must either pass this option or ctx option, but not both)
- ctx (Object) - a Koa context object (defaults to false) (you must either pass this option or req option, but not both)
- responseHeaders (String or Object) - we highly recommend that you pass res._headers (see Cabin's middleware logic if you need an example of this)
- userFields (Array) - defaults to [ 'id', 'email', 'full_name', 'ip_address' ], list of fields to cherry-pick from the user object parsed out of req.user
- sanitizeFields (Array) - defaults to the list of Strings provided under Sensitive Field Names Automatically Masked below
- sanitizeHeaders (Array) - defaults to the list of Strings provided under Sensitive Header Names Automatically Masked below (case insensitive)
- maskCreditCards (Boolean) - defaults to true, and specifies whether or not credit card numbers are masked
- maskBuffers (Boolean) - defaults to true, and will rewrite Buffer's, ArrayBuffer's, and SharedArrayBuffer's recursively as an object of { type: <String>, byteLength: <Number> }. Note that this will save you on disk log storage size as logs will not output verbose stringified buffers – e.g. imagine a 10MB file image upload sent across the request body as a Buffer!)
- maskStreams (Boolean) - defauls to true, and will rewrite Stream's to { type: 'Stream' } (this is useful for those using multer v2.x (streams version), or those that have streams in req.body, req.file, or req.files)
- checkId (Boolean) - defaults to true, and prevents Strings that closely resemble primary key ID's from being masked (e.g. properties named _id, id, ID, product_id, product-id, productId, productID, and product[id] won't get masked or show as a false-positive for a credit card check)
- checkCuid (Boolean) - defaults to true, and prevents cuid values from being masked
- checkObjectId (Boolean) - defaults to true, and prevents MongoDB BSON ObjectId from being masked
- checkUUID (Boolean) - defaults to true, and prevents uuid values from being masked
- rfdc (Object) - defaults to { proto: false, circles: false } (you should not need to customize this, but if necessary refer to rfdc documentation)
- parseBody (Boolean) - defaults to true, if you set to false we will not parse nor clone the request body property (this overrides all other parsing settings related)
- parseQuery (Boolean) - defaults to true, if you set to false we will not parse nor clone the request query property (this overrides all other parsing settings related)
- parseFiles (Boolean) - defaults to true, if you set to false we will not parse nor clone the request file nor files properties (this overrides all other parsing settings related)

It automatically detects whether the request is from the Browser, Koa, or Express, and returns a parsed object with populated properties.

Here's an example object parsed:

{
  "id": "5d126d86160cea56950f80a9",
  "timestamp": "2019-06-25T18:52:54.000Z",
  "is_http": true,
  "request": {
    "method": "POST",
    "query": {
      "foo": "bar",
      "beep": "boop"
    },
    "headers": {
      "host": "127.0.0.1:59746",
      "accept-encoding": "gzip, deflate",
      "user-agent": "node-superagent/3.8.3",
      "authorization": "Basic ********************",
      "accept": "application/json",
      "cookie": "foo=bar;beep=boop",
      "content-type": "multipart/form-data; boundary=--------------------------104476455118209968089794",
      "content-length": "1599",
      "connection": "close"
    },
    "cookies": {
      "foo": "bar",
      "beep": "boop"
    },
    "url": "/?foo=bar&beep=boop",
    "body": "{\"product_id\":\"5d0350ef2ca74d11ee6e4f00\",\"name\":\"nifty\",\"surname\":\"lettuce\",\"bank_account_number\":\"1234567890\",\"card\":{\"number\":\"****-****-****-****\"},\"stripe_token\":\"***************\",\"favorite_color\":\"green\"}",
    "timestamp": "2019-06-25T18:52:54.589Z",
    "id": "fbbce5d4-02d9-4a81-9a70-909631317e7d",
    "http_version": "1.1",
    "files": "{\"avatar\":[{\"fieldname\":\"avatar\",\"originalname\":\"avatar.png\",\"encoding\":\"7bit\",\"mimetype\":\"image/png\",\"buffer\":{\"type\":\"Buffer\",\"byteLength\":216},\"size\":216}],\"boop\":[{\"fieldname\":\"boop\",\"originalname\":\"boop-1.txt\",\"encoding\":\"7bit\",\"mimetype\":\"text/plain\",\"buffer\":{\"type\":\"Buffer\",\"byteLength\":7},\"size\":7},{\"fieldname\":\"boop\",\"originalname\":\"boop-2.txt\",\"encoding\":\"7bit\",\"mimetype\":\"text/plain\",\"buffer\":{\"type\":\"Buffer\",\"byteLength\":7},\"size\":7}]}"
  },
  "user": {
    "ip_address": "::ffff:127.0.0.1"
  },
  "response": {
    "headers": {
      "x-powered-by": "Express",
      "x-request-id": "fbbce5d4-02d9-4a81-9a70-909631317e7d",
      "content-security-policy": "default-src 'none'",
      "x-content-type-options": "nosniff",
      "content-type": "text/html; charset=utf-8",
      "content-length": "1213",
      "x-response-time": "48.658ms",
      "date": "Tue, 25 Jun 2019 18:52:54 GMT",
      "connection": "close"
    },
    "http_version": "1.1",
    "status_code": 200,
    "reason_phrase": "OK",
    "timestamp": "2019-06-25T18:52:54.000Z",
    "duration": 48.658
  },
  "duration": 1.350323,
  "app": {
    "name": "parse-request",
    "version": "1.0.11",
    "node": "v10.15.3",
    "hash": "f99bb8f28be5c6dc76bed76f6dd8984accc5c5fa",
    "environment": "test",
    "hostname": "users-MacBook-Pro.local",
    "pid": 22165
  }
}

A few extra details about the above parsed properties:

id (String) - is a newly created BSON ObjectId used to uniquely identify this log
timestamp (String) - is the ISO-8601 date time string parsed from the id (thanks to MongoDB BSON ObjectID.getTimestamp method)
is_http (Boolean) - defaults to true to indicate that this was a parsed HTTP request
duration (Number) - is the number of milliseconds that parseRequest took to parse the request object (note that this uses process.hrtime which this package polyfills thanks to browser-process-hrtime)
user (Object) - is parsed from the user object on req.user automatically (e.g. you are using passport):
- ip_address (String) - IP address parsed
- ... - additional fields are optionally parsed from req.user
request (Object) - request object information parsed from options.req:
- id (String) - is conditionally added if req.id is a String (we highly recommend that you use express-request-id in your project, which will automatically add this property if X-Request-Id if it is set, otherwise it will generate it as a new UUID)
- file (Object) - is conditionally added if you have a req.file property (e.g. if you're using multer)
- files (Array) - is conditionally added if you have a req.files property (e.g. if you're using multer)
- http_version (String) - is parsed from req.httpVersion or req.httpVersionMajor and req.httpVersionMinor
- timesamp (String) - is the ISO-8601 date time string parsed from when the request was received (we highly recommend that you use request-received for this to be parsed as accurately as possible, although we do support a few widely-used fallback approaches)
- headers (Object) - the raw request headers (lowercased)
response (Object) - response object information parsed from options.responseHeaders (we use http-headers to parse this information):
- http_version (String) - is parsed from the response HTTP headers major and minor HTTP version
- timestamp (String) - is the ISO-8601 date time string parsed from the response's Date header
- duration (Number) - is the number of milliseconds parsed from the X-Response-Time HTTP header (we highly recommend that you use response-time and request-received)
- headers (Object) - the raw response headers (lowercased)
- status_code (Number) - the response's status code (see RFC spec on Status Code and Reason Phrase)
- reason_phrase (String) - the response's reason phrase (see RFC spec on Status Code and Reason Phrase)

Please see Credit Card Masking and Sensitive Field Names Automatically Masked below for more information about how request.body, request.file, and request.files are parsed and conditionally masked for security.

Credit Card Masking

We also have built-in credit-card number detection and masking using the credit-card-type library.

This means that credit card numbers (or fields that are very similar to a credit card) will be automatically masked. If you'd like to turn this off, pass false to maskCreditCards**

Sensitive Field Names Automatically Masked

See sensitive-fields for the complete list.

Sensitive Header Names Automatically Masked

The Authorization HTTP header has its <credentials> portion automatically masked.

This means that if you are using BasicAuth or JSON Web Tokens ("JWT"), then your tokens will be hidden.

Usage

We highly recommend to simply use Cabin as this package is built-in!

VanillaJS

The example below uses xhook which is used to intercept HTTP requests made in the browser.

<script src="https://polyfill.io/v3/polyfill.min.js?features=performance,WeakRef"></script>
<script src="https://unpkg.com/xhook"></script>
<script src="https://unpkg.com/parse-request"></script>
<script type="text/javascript">
  (function() {
    xhook.after(function(req, res) {
      var req = parseRequest({ req });
      console.log('req', req);
      // ...
    });
  })();
</script>

Required Browser Features

We recommend using https://polyfill.io (specifically with the bundle mentioned in VanillaJS above):

<script src="https://polyfill.io/v3/polyfill.min.js?features=performance,WeakRef"></script>

performance.now() is not supported in op_mini all
WeakRef is not supported in Opera 91

Koa

const parseRequest = require('parse-request');

// ...

app.get('/', (ctx, next) => {
  const req = parseRequest({ req: ctx });
  console.log('req', req);
  // ...
});

Express

const parseRequest = require('parse-request');

// ...

app.get('/', (req, res, next) => {
  const req = parseRequest({ req });
  console.log('req', req);
  // ...
});

If you override req.body and need to preserve original in logs

Sometimes developers overwrite req.body or req.body properties – therefore if you want to preserve the original request, you can add req._originalBody = req.body (or ctx.request._originalBody = ctx.request.body if you're using Koa) at the top of your route middleware (or as a global route middleware).

If you want to disable body parsing just for a specific route (e.g. prevent log output from showing the body)

If you're using Express:

const disableBodyParsing = Symbol.for('parse-request.disableBodyParsing');

// ...

app.get('/', (req, res, next) => {
  req[disableBodyParsing] = true;
  next();
});

If you're using Koa:

const disableBodyParsing = Symbol.for('parse-request.disableBodyParsing');

// ...

app.get('/', (ctx, next) => {
  ctx.req[disableBodyParsing] = true;
  next();
});

If you want to disable file parsing just for a specific route (e.g. prevent log output from showing the file(s))

If you're using Express:

const disableFileParsing = Symbol.for('parse-request.disableFileParsing');

// ...

app.get('/', (req, res, next) => {
  req[disableFileParsing] = true;
  next();
});

If you're using Koa:

const disableFileParsing = Symbol.for('parse-request.disableFileParsing');

// ...

app.get('/', (ctx, next) => {
  ctx.req[disableFileParsing] = true;
  next();
});