cansecurity is your all-in-one security library for user authentication management and authorization in node (and specifically expressjs and restify) apps.
You can use use authentication, per-route authorization, and even declarative authorization. You can authenticate against any authentication system you like, just provide the interface function.
It's this simple:
var express = require('express'), app = express(), cs = require('cansecurity'), cansec = cs.init(/* init params */);
app.use(cansec.validate);
app.user(app.router);
// send200 is a shortcut route to send a 200 response
// open route
app.get("/public",send200);
// only authorized if logged in, or as certain roles, or some combination
app.get("/secure/loggedin",cansec.restrictToLoggedIn,send200);
app.get("/secure/customloggedin",cansec.setUnauthenticatedCode({code:302,location:"/login"}),cansec.restrictToLoggedIn,send200);
app.get("/secure/user/:user",cansec.restrictToSelf,send200);
app.get("/secure/roles/admin",cansec.restrictToRoles("admin"),send200);
app.get("/secure/roles/adminOrSuper",cansec.restrictToRoles(["admin","super"]),send200);
app.get("/secure/selfOrRoles/:user/admin",cansec.restrictToSelfOrRoles("admin"),send200);
app.get("/secure/selfOrRoles/:user/adminOrSuper",cansec.restrictToSelfOrRoles(["admin","super"]),send200);
// only authorized if "searchParam" is set to the same value as the user ID field set in cs.init();
app.get("/secure/param",cansec.restrictToParam("searchParam"),send200);
app.get("/secure/paramOrRole",cansec.restrictToParamOrRoles("searchParam","admin"),send200);
app.get("/secure/paramOrMultipleRoles",cansec.restrictToParamOrRoles("searchParam",["admin","super"]),send200);
// only authorized if getCheckObject() returns an object, with field owner, that has a value matching the user ID field
app.get("/secure/field",cansec.restrictToField("owner",getCheckObject),send200);
app.get("/secure/fields",cansec.restrictToField(["owner","recipient"],getCheckObject),send200);
app.get("/secure/fieldOrRole",cansec.restrictToFieldOrRoles("owner","admin",getCheckObject),send200);
app.get("/secure/fieldOrRoles",cansec.restrictToFieldOrRoles("owner",["admin","super"],getCheckObject),send200);
app.get("/secure/fieldsOrRole",cansec.restrictToFieldOrRoles(["owner","recipient"],"admin",getCheckObject),send200);
app.get("/secure/fieldsOrRoles",cansec.restrictToFieldOrRoles(["owner","recipient"],["admin","super"],getCheckObject),send200);
// only authorized if the request parameter "private" has the value "true", and then restrict to logged in
app.get("/secure/conditionalDirect",cansec.ifParam("private","true").restrictToLoggedIn,send200);
app.get("/secure/conditionalIndirect",cansec.ifParam("private","true").restrictToRoles(["admin","super"]),send200);
And if you prefer declarative authorization, even easier:
// inside app.js:
// instantiate the user validator
app.use(cansec.validate);
// instantiate the declarative authorizer
app.use(cansec.authorizer(pathToAuthConfigFile));
// inside "pathToAuthConfigFile"
{
"routes": [
// [verb,path,default,[test params,] test condition]
["GET","/api/user","user.roles.admin === true"],
["GET","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
["GET","/api/user/:user",{"private":"true"},"user.roles.admin === true || user.id === req.param('user')"],
["PUT","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
["GET","/api/user/:user/roles","user.roles.admin === true || user.id === req.param('user')"],
["PUT","/api/user/:user/roles","user.roles.admin === true"]
]
}
For any breaking changes, please see the end of this README.
cansecurity will manage your user authentication, including cross-server stateless sessions. It can use either native express sessions and or its own stateless sessions. cansecurity stateless sessions can keep a user logged in automatically across multiple nodejs instances, essentially creating free single-sign-on.
The flow looks like this:
cansecurity will handle authorization in your requests, determining if a user should be allowed to perform a certain request, based on your rules.
You can tell cansecurity to manage your authorization imperatively (via middleware code) or declaratively (using a config file). Whatever works better for you is just fine for cansecurity!
Installation is simple, just install the npm module:
npm install cansecurity
First initialize your cansecurity instance, then use its validation to validate users.
To initialize cansecurity, you must first require() it, and then init() it, which will return the middleware you can use:
var cs = require('cansecurity');
var cansec = cs.init(initConfig);
The initConfig
has the following properties:
sessionExpiry
: OPTIONAL. Integer in minutes how long sessions should last, default is15
. Used both for expressjs sessions and CS sessions. SettingsessionExpiry
will only affect how long a session is valid for cansecurity. It will not affect the underlying expressjs session itself.sessionKey
: OPTIONAL. String. Secret key shared between nodejs servers to provide single-sign-on.privateKey
: OPTIONAL. String. PEM-encoded RSA private key to use to encrypt the token.publicKey
: OPTIONAL. String. PEM-encoded RSA public key to use to validate the token.validate
: REQUIRED. Function that will get a user by username, and possibly validate their password, asynchronously. For more details, see below.encryptHeader
: OPTIONAL. With a value of true, the entire JWT is encrypted usingrc4-hmac-md5
algorithm.authHeader
: OPTIONAL. Replaces the headerX-CS-Auth
in which the server sends its token and user information back to the requestor/browser with the specified header name.X-CS-Auth
is used only for sending the request from the server to the requestor (browser). The request to the server always isAuthentication
.invalidTokenMessage
: OPTIONAL. Custom message to send back if an authentication is attempted with an invalid token.debug
: OPTIONAL. Print debug messages about each authentication attempt to the console. It will not include the actual password.
cansecurity
supports both shared key token validation and public/private key token validation. Which it uses depends on what parameters you pass:
sessionKey
: shared session key using"HS256"
algorithm.privateKey
/publicKey
: public/private key using"RS256"
algorithm.- none of the above: will automatically generate a shared session key, a random 64-character string.
Because the default, if no option is provided, is to auto-generate a shared session key, you must provide either a public/private key pair or a session if you want to share tokens between servers.
The behaviour of cansecurity
changes based on which parameters you provide:
- No
sessionKey
,publicKey
orprivateKey
: auto-generate a random shared session key, as described above. - Only
sessionKey
: sign and validate tokens using provided session key with"HS256"
algorithm. - Only
privateKey
: sign tokens using provided private key with"RS256"
algorithm. Any attempt to validate tokens will fail. - Only
publicKey
: validate tokens using provided public key with"RS256"
algorithm. Any attempt to sign tokens will fail. - Both
privateKey
andpublicKey
: sign and validate tokens using provided private and public keys with"RS256"
algorithm. sessionKey
and eitherpublicKey
orprivateKey
: launch-time error.
Note that encrypted headers do not yet work with private/public keys, as they rely on the same session key to encrypt.
Validation is straightforward. Once you have set up cansecurity properly, it functions as standard expressjs middleware:
app.use(cansec.validate);
app.use(app.router);
This should be done before your router.
If the user is successfully authenticated, then the user object will be placed in two locations:
req["X-CS-Auth"];
req.session["X-CS-AUTH"].user; // only if expressjs sessions have been enabled
However, for safety, you should retrieve it using the convenience method:
require('cansecurity').getUser(req);
// or simply
cs.getUser(req);
You can also determine how the current user was authorized, credentials (e.g. password) and token, by calling
require('cansecurity').getAuthMethod(req);
// returns "credentials" or "token"
// or simply
cs.getAuthMethod(req);
This is very useful in cases when you need the existing password for an action. A common use case is changing a password or setting security parameters. You might be fine with the user logging in via token for most actions (happens all the time when you go back to Facebook or Twitter from the same browser as last time), but if they want to change their password, they need to send the password again (try changing your Facebook or Gmail password, or Gmail two-factor authentication).
The only required option to cs.init()
is the function to allow cansecurity to get a user from your system: validate()
.
Must have the following signature
validate(username,password,callback);
The validate()
function is expected to retrieve user information from your preferred user store. It may validate a password for the user as well, and indicate to the callback if it succeeded or failed. The signature and expected parameters to the callback are as follows:
callback(success,user,message);
Where:
success
: boolean, if we succeeded in retrieving the user and, if requested, validating password credentials
user
: the actual user object. This can be a function, a JavaScript object, anything you want. It will be placed inside the session and the request for you to use later. If retrieval/validation was successful, this must not be null/undefined.
message
= the error message in case of retrieval/validation failure. This can be anything you want, and will be passed along with the 401 unauthenticated response.
If the user was already authenticated via session, token or some other method, then validateUser()
will be called with password
parameter set to undefined
. If password
is set to anything other than undefined
(including a blank string), then validateUser()
is expected to validate the password along with retrieving the user.
cansec.init({
validate: function(username,password,callback) {
if (password !== undefined) {
// validate the username and password, then retrieve the user credentials
} else {
// just retrieve the user credentials
}
}
});
When authentication fails, cansecurity will directly return 401 with the message "unauthenticated".
- If authentication is required and succeeds, it will set request["X-CS-Auth"], and request.session["X-CS-Auth"] if sessions are enabled, and then call next() to jump to the next middleware.
- If authentication is required and fails, it will return
401
with the text messageunauthenticated
. You can customize the return code and optional headers on a per-path basis. - If authentication is not required, it will jump to the next middleware
If the user has provided HTTP Basic Authentication credentials in the form of username/password and the authentication via validate()
fails. In that case, cansecurity will return a 401
.
If you wish to customize the response code or headers for unauthenticated, add a middleware setting a unique "unauthenticated" object before the restriction that requires the user to be authenticated. For example:
app.get("/secure/loggedin",cansec.restrictToLoggedIn,send200);
app.get("/secure/customloggedin",cansec.setUnauthenticatedCode({code:302,location:"/login"}),cansec.restrictToLoggedIn,send200);
In the above examples, the restrictions are almost identical, except that a user not logged in (or with invalid credentials) attempting to access /secure/loggedin
will receive a 401
http response code, whereas a user attempting to access /secure/customloggedin
will receive a 302
http response code and the http header Location: /login
will be set on the response.
With each request, the following algorithm is followed:
- Was there an HTTP Basic
Authentication
header? If so, validate using the credentials. If they succeed, the user is authenticated, else send back a401
unauthenticated and include a response X-CS-Auth header of"error invalidpass"
. If not, go to the next step. - Was there an HTTP Bearer
Authentication
header? If so, validate the JSON Web Token using the auth header. If they succeed, the user is authenticated, else they are not and return a401
witherror=invalidtoken
. If not, go to the next step. - Is there a valid and non-expired expressjs session? If so, the user is authenticated. If not, go to the next step.
- The user is not authenticated.
Note that failing to get an authentication for all of the above steps does not necessarily indicate that a 401 should be sent back. It is entirely possible that the user is accessing a resource that does not require authentication! This part of the cansecurity library is entirely about authentication; authorization is a different topic.
To summarize:
- Passing credentials in an
Authentication
header (Bearer
with JWT orBasic
with username/password) that fail, always will return a401
. - Passing no credentials but with a valid expressjs session will pass the request with the user as authenticated to the app.
- Passing no credentials with no valid expressjs sessions will pass the request with the user as unauthenticated to the app.
cansecurity passes details about success or failure of authentication in the custom X-CS-Auth
HTTP response header. Of course, a failed authentication will return a 401
, but the reason for failure will be in the appropriate header listed in this section. Similarly, a successful authentication - by any means - will allow the request to go through returning a 200
, 201
, 403
, 404
, etc., depending on the app. cansecurity will, however, return the session token and logged in user via the X-CS-Auth
HTTP response header.
The X-CS-Auth
response header contains error responses or success tokens.
A successful authentication provides the following format:
success token username expiry
Where:
success
: result of the authentication, either success
or error
token
: a JSON Web token if success
, or an error message if error
username
: the username if success
expiry
: when this token will expire, as a JavaScript (Unix) second timestamp, provided by Math.floor(new Date().getTime()/1000)
Because a new token is created anew with each request, the expiry window is rolling - x minutes from the last valid request.
Both the username
and expiry
are convenience methods. They are contained within the JSON Web Token provided in the response.
The JWT payload contains the following fields:
sub
: the subject of the JWT. This is identical to theusername
field in the header response.exp
: the expiry of the token. This is identical to theexpiry
field in the header response.cs-user
: the actual logged in user when authentication by any means was successful.
The cs-user
is a "JST private claim". Normally, it is a JSON-encoded string, but it really depends on what your validate()
function returns in the user
parameter of the callback
.
Note: You need to be really careful with what validate()
returns. Everything in there goes into the cs-user
field in the JWT. While it only goes in the header to the authenticated user, it still is sending out everything you send. You might not want the password - even hashed - in the header fields.
A failed authentication provides the following format:
error message
Where:
result
: result of the authentication, either success
or error
message
: the error message, if any
Note for usage in CORS situations. cansecurity automatically adds the following header to every response:
Access-Control-Expose-Headers: X-CS-Auth
Of course, it does so intelligently, so it adds it to an existing list of headers (does not trounce them) or creates it.
Extensive performance testing has not been done. However, all of the algorithms are symmetric, which are very high-performance. The expensive part is validate(), which may require your app to look in a data source or database. However, since the majority of requests will simply hit the local session, the user will be stored locally, and it is not an issue. The hit will only be for the first authentication for each user, as well as when a user switches between nodejs servers using cansecurity stateless sessions.
See the directory ./example/
Authorization is the process of checking if a user is allowed to perform a certain action (in our case, execute a certain route), assuming they have already been authenticated (or not).
cansecurity can do authorization as route-specific middleware (imperative) or via a config file (declarative).
cansecurity authorization also includes conditionals, allowing the authorization to be applied only if certain parameters are met.
This is the traditional express usage: declare a route, chain up some middleware functions. The easiest way to demonstrate this is with an example, following which we will describe all of the options and APIs.
express = require('express'),
cansec = require('cansecurity').init({});
server = express.createServer();
// do lots of server initialization
app.get("/some/route/:user",cansec.restrictToLoggedIn,routeHandler);
app.get("/some/route",cansec.ifParam("private","true").restrictToRoles("admin"),routeHandler);
Usage of cansecurity authorization is only possible if you are using cansecurity authentication as well. To use authorization, you do two steps: initialization and middleware.
The authorization component of cansecurity is initialized at the same time as the authentication component:
var cs = require('cansecurity'), cansec;
cansec = cs.init({});
or more simply:
var cansec = require('cansecurity').init({});
In initialization, you set two key authorization parameters as properties of the config object passed to cs.init(). Both are objects and both are optional.
- fields: OPTIONAL. Fields within the User object that was retrieved/set during authentication. These fields are expected to be part of a POJSO (Plain Old JavaScript Object) stored as part of validate(), and are used in some of the restrictTo* authorization middleware. There are currently two fields:
-
- fields.id: Property of the User object that contains the user ID. OPTIONAL. Defaults to "id"
-
- fields.roles: Property of the User object that contains the user roles, as an array of strings. OPTIONAL. Defaults to "roles"
- params: OPTIONAL. Names of params passed as part of the expressjs route, and retrievable as this.params[param]. These params are used as part in some of the restrictTo* authorization middleware. There is currently one field:
-
- params.id: Param in which the user ID is normally stored, if none is provided, then "user" is used. For example, if params.id === "foo", then the route should have /user/:foo.
Initialization returns the object that has the restrictTo* middleware.
As in the example above, once you have authentication and authorization set up and initialized, you may use authorization middleware:
// execute routeHandler() if user is logged in, else send 401
app.get("/some/route/:user",cansec.restrictToLoggedIn,routeHandler);
// execute routeHandler if req.param("user") === user[fields.id], where 'user' is as returned by validate(), else send 401
app.get("/my/data/:user",cansec.restrictToSelf,routeHandler);
// execute routeHandler if user[fields.roles] contains "admin", where 'user' is as returned by validate(), else send 401
app.get("/admin/:user",cansec.restrictToRoles("admin"),routeHandler);
// execute routeHandler if one of:
// req.param("secret") === "false"
// req.param("secret") === "true" && user[fields.roles] contains "admin"
app.get("/user",cansec.ifParam("secret","true").restrictToRoles("admin"),routeHandler);
cansecurity authorization will directly return a 403
and message unauthorized
if authorization is required, i.e. a restrictTo* middleware is called, and fails.
Obviously, authentication comes before authorization, and if the user fails to authenticate, you may get a 401 from the authentication section without ever trying authorization.
The following authorization middleware methods are available. Each one is followed by an example. There are two sections
- Regular API: Regular restrictTo* that are always applied.
- Conditional API: Conditions under which to apply the regular restrictTo* interfaces.
Regular API interfaces are used to restrict access, each example is given below.
- restrictToLoggedIn - user must have logged in
app.get("/some/route/:user",cansec.restrictToLoggedIn,routeHandler);
- restrictToSelf - user must have logged in and the user ID in the user object from authentication (fields.id above) must equal some parameter in the URL or body (params.id)
var cs = require('cansecurity');
cansec = cs.init({
fields: {id: "userid"},
params: {id: "user"}
});
// only allow a person to see their own stuff
app.get("/some/route/:user",cansec.restrictToSelf,routeHandler);
// note that the param in the route is ":user", which matches the params.id:"user" in cansec.init()
- restrictToRoles - user must have logged in and the user must have in his/her "roles" property (fields.roles) in the user object from authentication one of the named roles (one role as string or multiple in array). Roles argument to the function may be a string or an array of strings.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.get("/api/admin",cansec.restrictToRoles("admin"),routeHandler);
app.get("/api/siteadmin",cansec.restrictToRoles(["admin","superadmin"]),routeHandler);
// will require the User object from authentication to have a property "roles", which is an array of strings. If one of the strings is "admin", then /api/admin will be allowed. If one of the roles is "admin" or "superadmin", then /api/siteadmin will be allowed.
- restrictToSelfOrRoles - combination of the previous two. Roles argument to the function may be a string or an array of strings.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.put("/api/user/:user",cansec.restrictToSelfOrRoles("admin"),routeHandler);
/*
* Will work if one of the following is true:
* 1) The logged in user has a property "userid" and it matches exactly the value of the param ":user"; OR
* 2) The logged in user has a property "roles" which is an array of strings, one of which is "admin"
*/
- restrictToParam - user must have logged in and some field in the user object (fields.id) from authentication must equal some parameter in the URL or body (params.id). Param argument to the function may be a string or an array of strings.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.put("/api/user/search",cansec.restrictToParam("searchParam"),routeHandler);
/*
* Will work if the logged in user has a property "userid" (since init() set fields.id to "userid"), and the value of that property matches req.param("searchParam").
* Useful for using parameters in searches.
*/
- restrictToParamOrRoles - user must have logged in and some field in the user object (fields.id) from authentication must equal some parameter in the URL or body (params.id) or user must have a specific role. Param argument and roles argument to the function may each be a string or an array of strings.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.put("/api/user/search",cansec.restrictToParamOrRoles("searchParam",["admin","superadmin"]),routeHandler);
app.put("/api/address/search",cansec.restrictToParamOrRoles(["searchParam","addParam"],"admin"),routeHandler);
/*
* Will work if one of the following is true:
* 1) the logged in user has a property "userid" (since init() set fields.id to "userid"), and the value of that property matches req.param("searchParam"), or, in the second example, one of "searchParam" or "addParam".
* 2) The logged in user has the role, as one of the array of strings of the property "roles", set to "admin" or "superadmin" (for the first example), or "admin" (for the second example).
*/
- restrictToField - user must have logged in and some field in the user object (fields.id) from authentication must equal the response to a given callback with a given field or fields parameter.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.get("/api/user/search",cansec.restrictToField("owner",getObjectFn),routeHandler);
/*
* Will call getObjectFn(req,res) to get a regular JavaScript object, and then try to match the requested fields, in the above example "owner", to the ID of the User from authentication. The ID from authentication is user.userid, as given in the init() for fields.id.
*/
- restrictToFieldOrRoles - user must have logged in and some field in the user object (fields.id) from authentication must equal the response to a given callback with a given field or fields parameter, or the user must have a role or roles.
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.get("/api/user/search",cansec.restrictToFieldOrRoles("owner","admin",getObjectFn),routeHandler);
/*
* Will call getObjectFn(req,res) to get a regular JavaScript object, and then try to match the requested fields, in the above example "owner", to the ID of the User from authentication. The ID from authentication is user.userid, as given in the init() for fields.id.
*/
app.get("/api/user/search",cansec.restrictToFieldOrRoles(["owner","recipient"],["admin","superadmin"],getOwnerFn),routeHandler);
A typical use case for restrictToField and its variant restrictToFieldOrRoles is that you may load an object, and want to restrict its access to the owner of an object. For example, let us assume that you are loading an employee record. For that, restrictToSelf would be fine, since the User ID from authentication is likely to match the ID for requesting the employee record. The following example shows this use case:
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.get("/api/employee/:user",cansec.restrictToSelfOrRoles("admin"),sendDataFn);
However, what if you are retrieving a record whose authorization requirements are not known until it is loaded. For example, you are loading a paystub, whose URL is /api/paystubs/34567. Until you load the paystub, you don't actually know who the owner is. Of course, you might make it accessible only via a more detailed API as /api/employee/12345/paystubs/34567, but let us assume that you need to do it directly, with the first simplified API. Until you load the actual paystubs object, and see that the employee is, indeed, 12345, the one who logged in, you don't know whether or not to show it. The following example describes how to simply implement this use case:
var cansec = require('cansecurity').init({
fields: {id: "userid", roles:"roles"},
params: {id: "user"}
});
app.get("/api/employee/:user",cansec.restrictToSelfOrRoles("admin"),sendDataFn);
app.get("/api/paystub/:payid",payStubLoad,cansec.restrictToField("employee",getObjectFn),sendDataFn);
In this example, we load the paystub, but do not send it. The paystub object retrieved by payStubLoad looks like this:
{
id: "34567",
employee: "12345",
date: "2011-01-31",
amount: "$100"
}
This is then stored in the request object. Now getObjectFn can return the same object, which has employee as "12345". This is then matched to User.userid, which will allow it to proceed.
The conditional API simply creates conditions under which the regular restrictions are applied. If the conditions are not met, then restrictions are not applied.
- ifParam - apply the restriction only if a certain parameter matches a certain value.
What if you have a resource that is normally accessible by all, but if certain parameters are applied, e.g. ?secret=true, then it should be restricted to an admin?
app.get("/api/employee",cansec.ifParameter("secret","true").restrictToRoles("admin"),sendDataFn);
In the above example, anyone can do a GET on /api/employee, but if they pass the parameter ?secret=true, then they will have to be logged in and have the role "admin" defined.
In our example, sendDataFn also checks for that parameter. If it is not set, then it sends public data about the employee list; if it is set, it sends public and private data, trusting that cansecurity prevented someone from getting in with ?secret=true unless they are authorized.
A common pattern, as shown in the last example above, is to retrieve an object, check it against the user, and then determine whether or not to allow the request to proceed. cansecurity would love to be able to just do the check directly. It even knows which field/property to check: "employee". It has two problems, however:
- It doesn't know where the object is stored. Sure, most use cases store it in the request object somewhere, or possibly the response object, but cansecurity does not want to impose on your application where that is. Thus, it just delegates to you, saying, "give me a function to retrieve the object, if I pass you the request."
- It doesn't know what the object looks like. It may be a POJSO (Plain Old JavaScript Object), like above, or one that supports that style, like Spine Models. But what if it is something more complex, a function, with special parameters? What if it is a Backbone Model, which requires using getters? By asking your application to provide a function, it completely abstracts out the issue, and says, "whatever, as long as you pass me back a POJSO, I am happy."
The function to clear is exposed once Cansecurity is initialized.
app.get("/logout",function(req, res){
cansec.clear(req, res);
res.send(200);
});
If you want to deny access to everything - always send a 403 - unless it is explicitly approved, just add a "deny all" line at the end of your routes.
app.get("/api/employee/:user",cansec.restrictToSelfOrRoles("admin"),sendDataFn);
app.get("/public/page",send200);
// lots more
app.get("*",function(req,res,next){res.send(403);});
Declarative authorization is given to drastically clean up your authorizations. Normal cansecurity authorization lets you inject authorization into every route, like so.
app.get("/secure/loggedin",cansec.restrictToLoggedIn,send200);
app.get("/secure/user/:user",cansec.restrictToSelf,send200);
app.get("/secure/roles/admin",cansec.restrictToRoles("admin"),send200);
app.get("/secure/roles/adminOrSuper",cansec.restrictToRoles(["admin","super"]),send200);
app.get("/secure/selfOrRoles/:user/admin",cansec.restrictToSelfOrRoles("admin"),send200);
app.get("/secure/selfOrRoles/:user/adminOrSuper",cansec.restrictToSelfOrRoles(["admin","super"]),send200);
// only authorized if "searchParam" is set to the same value as the user ID field set in cs.init();
app.get("/secure/param",cansec.restrictToParam("searchParam"),send200);
app.get("/secure/paramOrRole",cansec.restrictToParamOrRoles("searchParam","admin"),send200);
app.get("/secure/paramOrMultipleRoles",cansec.restrictToParamOrRoles("searchParam",["admin","super"]),send200);
// only authorized if getCheckObject() returns an object, with field owner, that has a value matching the user ID field
app.get("/secure/field",cansec.restrictToField("owner",getCheckObject),send200);
app.get("/secure/fields",cansec.restrictToField(["owner","recipient"],getCheckObject),send200);
app.get("/secure/fieldOrRole",cansec.restrictToFieldOrRoles("owner","admin",getCheckObject),send200);
app.get("/secure/fieldOrRoles",cansec.restrictToFieldOrRoles("owner",["admin","super"],getCheckObject),send200);
app.get("/secure/fieldsOrRole",cansec.restrictToFieldOrRoles(["owner","recipient"],"admin",getCheckObject),send200);
app.get("/secure/fieldsOrRoles",cansec.restrictToFieldOrRoles(["owner","recipient"],["admin","super"],getCheckObject),send200);
// only authorized if the request parameter "private" has the value "true", and then restrict to logged in
app.get("/secure/conditionalDirect",cansec.ifParam("private","true").restrictToLoggedIn,send200);
app.get("/secure/conditionalIndirect",cansec.ifParam("private","true").restrictToRoles(["admin","super"]),send200);
This is worlds better than before, when authorization was one of:
- didn't exist
- stuck inside the routes (really bad for separation of concerns, leading to unmaintainable code)
- binary, logged in or out but nothing more fine-tuned than that
But it still requires lots of code in the routes. What if you could just declare in a config file what authorization rules you want?
{
"routes": [
// [verb,path,[test params,][require logged in],[loader,]test condition]
["GET","/api/user",true,"user.roles.admin === true"],
["GET","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
["GET","/api/user/:user",{"private":"true"},true,"user.roles.admin === true || user.id === req.param('user')"],
["PUT","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
["GET","/api/user/:user/roles","user.roles.admin === true || user.id === req.param('user')"],
["PUT","/api/user/:user/roles","user.roles.admin === true"]
]
}
cansecurity provides you precisely this ability!
To use declarative authorization, you take two steps:
- Set up the config file
app.use(cansec.authorizer(pathToConfigFile,options))
The config file is a simple json
file. You can name it whatever you want. The file should be a single object, with one key, routes
, which is an array of arrays.
{
routes:[
[/* route 1 */],
[/* route 2 */],
/* ... */
[/* route n */]
]
}
Each route is an array of 4 or 5 parts, as follows:
[verb,route,[params,][loggedIn,][loader,]condition]
- verb: string, one of GET PUT POST DELETE, and is case-insensitive
- route: string, an express-js compatible route, e.g. "/api/user/:user" or "/post/:post/comment/:comment". Note that ".:format?" is optional. See "format" later.
- params: optional object, which will be checked to match the route, e.g.
{private:true}
or{secret:"true",name:"foo"}
. If the params match, then the route will be applied, else this route is considered to not match and will be ignored. See the examples below and the tests. - loggedIn: optional boolean. If true, user must be logged in via cansecurity before checking authorization. If the user is not logged in, send
401
. - loader: name of a loader in your initializer that should run when the verb/route/params/loggedIn are matched, but before testing the condition
- condition: JavaScript string which should return a condition. If true, then do the opposite of the default behaviour
Here are some examples.
// when GET /api/user, send 403 unless user.roles.admin === true
["GET","/api/user","user.roles.admin === true"],
// when GET /api/user, if not logged in, send 401; if logged in send 403 unless user.roles.admin === true
["GET","/api/user",true,"user.roles.admin === true"],
// when GET /api/user/:user, send 403 unless user.roles.admin === true, OR user.id === req.param('user')
["GET","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
// when GET /api/user/:user AND ?private=true (or in the body), send 403 unless user.roles.admin === true || user.id === req.param('user')
// if private !== true (or is unset or anything else), then this rule is not applied, and access is allowed
["GET","/api/user/:user",{"private":"true"},"user.roles.admin === true || user.id === req.param('user')"],
// same as previous example, but send 401 if !logged in, then continue from previous example
["GET","/api/user/:user",{"private":"true"},true,"user.roles.admin === true || user.id === req.param('user')"],
// when PUT /api/user/:user, send 403 unless user.roles.admin === true || user.id === req.param('user')
["PUT","/api/user/:user","user.roles.admin === true || user.id === req.param('user')"],
// when GET /api/user/:user/roles, send 403 unless user.roles.admin === true || user.id === req.param('user')
["GET","/api/user/:user/roles","user.roles.admin === true || user.id === req.param('user')"],
// when PUT /api/user/:user/roles, send 403 unless user.roles.admin === true
["PUT","/api/user/:user/roles","user.roles.admin === true"]
// when PUT /api/user/:user/roles, run the "roles" loader, then send 403 unless user.roles.admin === true || item.name === 'me'
["PUT","/api/user/:user/roles","roles","(user.roles.admin === true) || (item.name === 'me')"]
If you want to deny access to everything - always send a 403 - unless it is explicitly approved, just add a "deny all" line at the end of your declarative routes.
// when PUT /api/user/:user/roles, send 403 unless user.roles.admin === true
["PUT","/api/user/:user/roles","user.roles.admin === true"]
// when PUT /api/user/:user/roles, run the "roles" loader, then send 403 unless user.roles.admin === true || item.name === 'me'
["PUT","/api/user/:user/roles","roles","(user.roles.admin === true) || (item.name === 'me')"]
// deny everything else
["GET","*","false"]
The condition string is run inside its own new context. Besides the usual nodejs environment, it has the following variable available to it:
req
: the actual expressreq
object, normally found on each route whose signature isfunction(req,res,next)
.request
: an alias forreq
user
: the user object if you used cansecurity authentication. This is the equivalent of callingcansec.getUser(req)
._
: the underscore/lodash utility functions. cansecurity actually uses lodashitem
: the item loaded, if any, by the loader functions passed tocansec.init()
; see below.
You have the option, but not the requirement, to load data before passing your route through the declarative authorizer.
Each loader function has two simple jobs to do:
- Load relevant data into
req.cansecurity.item
- Call
next
req.cansecurity
will already be available as an object. item
can be an object, a string, an array, null, undefined, boolean, or anything at all that you want to pass to your conditions.
The full suite of request
, response
and next
methods is available. Thus, you could easily call res.send(400)
if you have an error and do not wish to proceed, or perhaps next(error)
. Similarly, req.param(someParam)
is also available.
Here is an example:
cansec.init({
loader: {
group: function(req,res,next) {
models.group.find(someParam,function(err,data){
req.cansecurity.item = data;
next();
});
}
}
});
And the declarative:
{
routes: [
["GET","/api/group/:group",true,"group","_.includes(item.members,user.id)"]
]
}
Note: Any route where login is required, login will be validated before running the loader.
So where do you actually define the loader functions? You have two options for where the loader can exist.
If you have or want a single loader, you can pass all of your loader functions into cansec.init()
:
cansec.init({
loader: {
user: function(req,res,next) {
},
group: function(req,res,next) {
}
}
});
And the declarative part:
{
routes: [
["GET","/api/group/:group",true,"group","_.includes(item.members,user.id)"]
]
}
You can define the loader functions in a file local to a certain declarative file:
// in your main server.js
app.use(cansec.authorizer(__dirname+'/path/to/decl.json',{loader: {
fn1: function(req,res,next){},
fn2: function(req,res,next){}
}}))
Of course, you can always separate the loader functions into another file, like with the init file, and require
it yourself:
// in your main server.js
app.use(cansec.authorizer(__dirname+'/path/to/decl.json',{loader: require(pathToLoader)}))
If you can do it globally, why bother with the local? Simple. You can have multiple declarative files. For example, we often separate the security authorization (user Jim is allowed to see his own account) from subscription authorization (user Jim already has 2 accounts and needs to upgrade his plan to get another).
// in your main server.js
app.use(cansec.authorizer(__dirname+'/path/to/security.json',{loader:securityConfig}))
app.use(cansec.authorizer(__dirname+'/path/to/plans.json',{loader:plansConfig}))
If course, you might want to keep them together, in which case just use the global!
What if a particular loader is defined in both lcala and global? The order of priority then is:
- Look for and run the loader in the local; if not found...
- Look for an run the loader in the global.
Simple, right?
The authorizer has one of three possible results:
- Send
401
if authentication is required and the user is not logged in - Send
403
if the route matches the condition fails next()
The logic is as follows:
- Does the route match? If not,
next()
; else - Does the route require authentication? If yes and the user is not logged in, send
401
; else - Does the condition evaluate to
true
and not have any errors? If not, send403
; else next()
Simple:
app.use(cansec.authorizer(pathToConfigFile,options));
app.use(app.router);
Done!
Many REST paths use a "format" extension. In express, it usually looks like this:
app.get("/api/user/:user.:format?",fn);
This allows express to handle both /api/user/10
and /api/user/10.json
.
cansecurity's declarative authorization can handle .:format?
just by putting it in the path:
["GET","/api/user/:user.:format?",{"private":"true"},true,"user.roles.admin === true || user.id === req.param('user')"],
But that can get tedious, if you have a lot of routes. To simplify things, one of the options when setting it up is {format:true}
as follows:
app.use(cansec.authorizer(pathToConfigFile,{format:true}));
If format
is set to true
, then cansecurity will automatically add .:format?
to every path that does not end in .:format?
already, or in /
.
To run the tests, from the root directory, run npm test
.
Note: Tests are set up both for express and for restify. However, running them both causes one to trounce the other. Apparently restify grabs hold of the http module and its server munges the requests that express tries to read. One cannot really blame restify; it never intended to run in the same node instance with another server.
To bypass this issue, run tests twice:
mocha -g '(?:restify)'
mocha -g restify
npm
is set up to do precisely this for you, if you run npm test
.
2.0.0 is a major release with many breaking changes.
The X-CS-Auth
header is deprecated from requests. All credentials should be passed in the Authentication
header, either Basic
for user/pass authentication, or Bearer
for a JSON Web Token that was generated by cansecurity.
The X-CS-User
header is deprecated from responses. All of the user information is now held as a field of the returned JWT.
- Response:
X-CS-User
contains the status and JWT, which includes the user information. - Request:
Authorization
contains the credentials, either user/pass or JWT.
In a successful response, the fields are separated by whitespace instead of :
:
X-CS-Auth: success token username expiry
Instead of
X-CS-Auth: success=token:username:expiry
In keeping with the JWT standard, the expiry field - both in the header and in the JWT - are given in timestamp seconds since epoch, instead of milliseconds.
Express 4 support!
- All cansecurity versions < 1.0.0 support only express < 4.0.0
- All cansecurity versions >= 1.0.0 support only express >= 4.0.0
restify is now fully supported and tested for. However, restify 2.8.0 or higher is required.
Additionally, because the tests trounce each other, run each separately when testing:
mocha -g express
mocha -g restify
Declarative authorization no longer has an option to "allow" or "deny" by default. All rules are "deny" unless the condition passes. It is very easy to invert the condition and make it pass except in certain circumstances.
["GET","/secure/path",true,"allow","a === b"]
Can just as easily be written as
["GET","/secure/path",true,"deny","a !== b"]
Or more simply
["GET","/secure/path",true,"a !== b"]
Prior to version 0.6.0, cansecurity sometimes would send back a 401 or 403 as res.send(401,"unauthenticated")
or res.send(403,"unauthorized")
, and sometimes would just call next({status:401,message:"unauthenticated"})
or next({status:403,message:"unauthorized"})
.
Beginning with version 0.6.0, cansecurity will always return 401 if authentication is required and not present / fails, and will always return a 403 if authorization is required and fails.
This makes the results far more consistent.
These notes apply to anyone using cansecurity prior to v0.5.0. These changes may be breaking, so read carefully.
Prior to version 0.5.0 (and preferably prior to 0.4.8), cansecurity worked with express 2.x and 3.x, although the full testing regimen worked properly only in express 2.x. Beginning with 0.5.0, only express 3.x will work.
In versions of cansecurity prior to 0.5.0, there were two functions passed to init()
:
validatePassword()
was called when the user authenticated with credentials to be checked.getUser()
was called when the user was authenticated already using a token or session, and we just needed the user object.
As of version 0.5.0, these are consolidated into a single validate()
function. Please check the documentation below.
Until version 1.0 of cansecurity, the legacy functions will continue to operate, if deprecated, under the following circumstances:
IF `validate()` is `undefined`, AND (`validatePassword()` and `getUser()`) are present, THEN cansecurity will use the old API.
IF `validate()` is defined, THEN (`validatePassword()` and `getUser()`) will be ignored, whether present or not.
Beginning with cansecurity 1.0, the old API will not function at all.