[![Build Status](https://github.com/jwtk/jjwt/actions/workflows/ci.yml/badge.svg?branch=master)](https://github.com/jwtk/jjwt/actions/workflows/ci.yml?query=branch%3Amaster)
[![Coverage Status](https://coveralls.io/repos/github/jwtk/jjwt/badge.svg?branch=master)](https://coveralls.io/github/jwtk/jjwt?branch=master)
[![Vuln score](https://snyk-widget.herokuapp.com/badge/mvn/io.jsonwebtoken/jjwt-root/badge.svg)](https://snyk-widget.herokuapp.com/badge/mvn/io.jsonwebtoken/jjwt-root/badge.svg)
[![Known Vulns](https://snyk.io/test/github/jwtk/jjwt/badge.svg)](https://snyk.io/test/github/jwtk/jjwt/badge.svg)
# Java JWT: JSON Web Token for Java and Android
JJWT aims to be the easiest to use and understand library for creating and verifying JSON Web Tokens (JWTs) and
JSON Web Keys (JWKs) on the JVM and Android.
JJWT is a pure Java implementation based exclusively on the
[JOSE Working Group](https://datatracker.ietf.org/wg/jose/documents/) RFC specifications:
* [RFC 7519: JSON Web Token (JWT)](https://tools.ietf.org/html/rfc7519)
* [RFC 7515: JSON Web Signature (JWS)](https://tools.ietf.org/html/rfc7515)
* [RFC 7516: JSON Web Encryption (JWE)](https://tools.ietf.org/html/rfc7516)
* [RFC 7517: JSON Web Key (JWK)](https://tools.ietf.org/html/rfc7517)
* [RFC 7518: JSON Web Algorithms (JWA)](https://tools.ietf.org/html/rfc7518)
* [RFC 7638: JSON Web Key Thumbprint](https://www.rfc-editor.org/rfc/rfc7638.html)
* [RFC 9278: JSON Web Key Thumbprint URI](https://www.rfc-editor.org/rfc/rfc9278.html)
* [RFC 7797: JWS Unencoded Payload Option](https://www.rfc-editor.org/rfc/rfc7797.html)
* [RFC 8037: Edwards Curve algorithms and JWKs](https://www.rfc-editor.org/rfc/rfc8037)
It was created by [Les Hazlewood](https://github.com/lhazlewood)
and is supported and maintained by a [community](https://github.com/jwtk/jjwt/graphs/contributors) of contributors.
JJWT is open source under the terms of the [Apache 2.0 License](http://www.apache.org/licenses/LICENSE-2.0).
## Table of Contents
* [Features](#features)
* [Currently Unsupported Features](#features-unsupported)
* [Community](#community)
* [Getting Help](#help)
* [Questions](#help-questions)
* [Bugs and Feature Requests](#help-issues)
* [Contributing](#contributing)
* [Pull Requests](#contributing-pull-requests)
* [Help Wanted](#contributing-help-wanted)
* [What is a JSON Web Token?](#overview)
* [JWT Example](#overview-example-jwt)
* [JWS Example](#overview-example-jws)
* [JWE Example](#overview-example-jwe)
* [Installation](#install)
* [JDK Projects](#install-jdk)
* [Maven](#install-jdk-maven)
* [Gradle](#install-jdk-gradle)
* [Android Projects](#install-android)
* [Dependencies](#install-android-dependencies)
* [Proguard Exclusions](#install-android-proguard)
* [Bouncy Castle](#install-android-bc)
* [Understanding JJWT Dependencies](#install-understandingdependencies)
* [Quickstart](#quickstart)
* [Create a JWT](#jwt-create)
* [Header](#jwt-header)
* [Header Builder](#jwt-header-builder)
* [Header Parameters](#jwt-header-params)
* [Header Map](#jwt-header-map)
* [Payload](#jwt-payload)
* [Arbitrary Content](#jwt-content)
* [Claims](#jwt-claims)
* [Standard Claims](#jwt-claims-standard)
* [Custom Claims](#jwt-claims-custom)
* [Claims Instance](#jwt-claims-instance)
* [Claims Map](#jwt-claims-map)
* [Compression](#jwt-compression)
* [Read a JWT](#jwt-read)
* [Static Parsing Key](#jwt-read-key)
* [Dynamic Parsing Key Lookup](#key-locator)
* [Custom Key Locator](#key-locator-custom)
* [Key Locator Strategy](#key-locator-strategy)
* [Key Locator Return Values](#key-locator-retvals)
* [Provider-constrained Keys (PKCS11, HSM, etc)](#key-locator-provider)
* [Claim Assertions](#jwt-read-claims)
* [Accounting for Clock Skew](#jwt-read-clock)
* [Custom Clock Support](#jwt-read-clock-custom)
* [Decompression](#jwt-read-decompression)
* [Signed JWTs](#jws)
* [Standard Signature Algorithms](#jws-alg)
* [Signature Algorithm Keys](#jws-key)
* [HMAC-SHA](#jws-key-hmacsha)
* [RSA](#jws-key-rsa)
* [Elliptic Curve](#jws-key-ecdsa)
* [Creating Safe Keys](#jws-key-create)
* [Secret Keys](#jws-key-create-secret)
* [Asymetric Keys](#jws-key-create-asym)
* [Create a JWS](#jws-create)
* [Signing Key](#jws-create-key)
* [SecretKey Formats](#jws-create-key-secret)
* [Signature Algorithm Override](#jws-create-key-algoverride)
* [Compression](#jws-create-compression)
* [Read a JWS](#jws-read)
* [Verification Key](#jws-read-key)
* [Verification Key Locator](#jws-read-key-locator)
* [Decompression](#jws-read-decompression)
* [Unencoded Payload Option](#jws-unencoded)
* [Detached Payload Example](#jws-unencoded-detached)
* [Non-Detached Payload Example](#jws-unencoded-nondetached)
* [Encrypted JWTs](#jwe)
* [JWE Encryption Algorithms](#jwe-enc)
* [JWE Symmetric Encryption](#jwe-enc-symmetric)
* [JWE Key Management Algorithms](#jwe-alg)
* [JWE Standard Key Management Algorithms](#jwe-alg-standard)
* [JWE RSA Key Encryption](#jwe-alg-rsa)
* [JWE AES Key Encryption](#jwe-alg-aes)
* [JWE Direct Key Encryption](#jwe-alg-dir)
* [JWE Password-based Key Encryption](#jwe-alg-pbes2)
* [JWE Elliptic Curve Diffie-Hellman Ephemeral Static Key Agreement](#jwe-alg-ecdhes)
* [Create a JWE](#jwe-create)
* [JWE Compression](#jwe-compression)
* [Read a JWE](#jwe-read)
* [JWE Decryption Key](#jwe-read-key)
* [JWE Decryption Key Locator](#jwe-key-locator)
* [ECDH-ES Decryption with PKCS11 PrivateKeys](#jwe-key-pkcs11)
* [JWE Decompression](#jwe-read-decompression)
* [JSON Web Keys (JWKs)](#jwk)
* [Create a JWK](#jwk-create)
* [Read a JWK](#jwk-read)
* [PrivateKey JWKs](#jwk-private)
* [Private JWK `PublicKey`](#jwk-private-public)
* [Private JWK from `KeyPair`](#jwk-private-keypair)
* [Private JWK Public Conversion](#jwk-private-topub)
* [JWK Thumbprints](#jwk-thumbprint)
* [JWK Thumbprint as Key ID](jwk-thumbprint-kid)
* [JWK Thumbprint URI](#jwk-thumbprint-uri)
* [JWK Security Considerations](#jwk-security)
* [JWK `toString()` Safety](#jwk-tostring)
* [Compression](#compression)
* [Custom Compression Algorithm](#compression-custom)
* [JSON Processor](#json)
* [Custom JSON Processor](#json-custom)
* [Jackson ObjectMapper](#json-jackson)
* [Custom Claim Types](#json-jackson-custom-types)
* [Gson](#json-gson)
* [Base64 Support](#base64)
* [Base64 in Security Contexts](#base64-security)
* [Base64 is not Encryption](#base64-not-encryption)
* [Changing Base64 Characters](#base64-changing-characters)
* [Custom Base64 Codec](#base64-custom)
* [Examples](#examples)
* [JWS Signed with HMAC](#example-jws-hs)
* [JWS Signed with RSA](#example-jws-rsa)
* [JWS Signed with ECDSA](#example-jws-ecdsa)
* [JWE Encrypted Directly with a SecretKey](#example-jwe-dir)
* [JWE Encrypted with RSA](#example-jwe-rsa)
* [JWE Encrypted with AES Key Wrap](#example-jwe-aeskw)
* [JWE Encrypted with ECDH-ES](#example-jwe-ecdhes)
* [JWE Encrypted with a Password](#example-jwe-password)
* [SecretKey JWK](#example-jwk-secret)
* [RSA Public JWK](#example-jwk-rsapub)
* [RSA Private JWK](#example-jwk-rsapriv)
* [Elliptic Curve Public JWK](#example-jwk-ecpub)
* [Elliptic Curve Private JWK](#example-jwk-ecpriv)
* [Edwards Elliptic Curve Public JWK](#example-jwk-edpub)
* [Edwards Elliptic Curve Private JWK](#example-jwk-edpriv)
## Features
* Fully functional on all Java 7+ JDKs and Android
* Automatic security best practices and assertions
* Easy to learn and read API
* Convenient and readable [fluent](http://en.wikipedia.org/wiki/Fluent_interface) interfaces, great for IDE
auto-completion to write code quickly
* Fully RFC specification compliant on all implemented functionality, tested against RFC-specified test vectors
* Stable implementation with almost 1,700 tests and enforced 100% test code coverage. Every single method, statement
and conditional branch variant in the entire codebase is tested and required to pass on every build.
* Creating, parsing and verifying digitally signed compact JWTs (aka JWSs) with all standard JWS algorithms:
| Identifier | Signature Algorithm |
|------------|-------------------------------------------------------------------|
| `HS256` | HMAC using SHA-256 |
| `HS384` | HMAC using SHA-384 |
| `HS512` | HMAC using SHA-512 |
| `ES256` | ECDSA using P-256 and SHA-256 |
| `ES384` | ECDSA using P-384 and SHA-384 |
| `ES512` | ECDSA using P-521 and SHA-512 |
| `RS256` | RSASSA-PKCS-v1_5 using SHA-256 |
| `RS384` | RSASSA-PKCS-v1_5 using SHA-384 |
| `RS512` | RSASSA-PKCS-v1_5 using SHA-512 |
| `PS256` | RSASSA-PSS using SHA-256 and MGF1 with SHA-2561 |
| `PS384` | RSASSA-PSS using SHA-384 and MGF1 with SHA-3841 |
| `PS512` | RSASSA-PSS using SHA-512 and MGF1 with SHA-5121 |
| `EdDSA` | Edwards-curve Digital Signature Algorithm2 |
1. Requires Java 11 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
2. Requires Java 15 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
* Creating, parsing and decrypting encrypted compact JWTs (aka JWEs) with all standard JWE encryption algorithms:
| Identifier | Encryption Algorithm |
|----------------------------------|--------------------------------------------------------------------------------------------------------------------------|
| A128CBC‑HS256
| [AES_128_CBC_HMAC_SHA_256](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.3) authenticated encryption algorithm |
| `A192CBC-HS384` | [AES_192_CBC_HMAC_SHA_384](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.4) authenticated encryption algorithm |
| `A256CBC-HS512` | [AES_256_CBC_HMAC_SHA_512](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.5) authenticated encryption algorithm |
| `A128GCM` | AES GCM using 128-bit key3 |
| `A192GCM` | AES GCM using 192-bit key3 |
| `A256GCM` | AES GCM using 256-bit key3 |
3. Requires Java 8 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
* All Key Management Algorithms for obtaining JWE encryption and decryption keys:
| Identifier | Key Management Algorithm |
|----------------------|-------------------------------------------------------------------------------|
| `RSA1_5` | RSAES-PKCS1-v1_5 |
| `RSA-OAEP` | RSAES OAEP using default parameters |
| `RSA-OAEP-256` | RSAES OAEP using SHA-256 and MGF1 with SHA-256 |
| `A128KW` | AES Key Wrap with default initial value using 128-bit key |
| `A192KW` | AES Key Wrap with default initial value using 192-bit key |
| `A256KW` | AES Key Wrap with default initial value using 256-bit key |
| `dir` | Direct use of a shared symmetric key as the CEK |
| `ECDH-ES` | Elliptic Curve Diffie-Hellman Ephemeral Static key agreement using Concat KDF |
| `ECDH-ES+A128KW` | ECDH-ES using Concat KDF and CEK wrapped with "A128KW" |
| `ECDH-ES+A192KW` | ECDH-ES using Concat KDF and CEK wrapped with "A192KW" |
| `ECDH-ES+A256KW` | ECDH-ES using Concat KDF and CEK wrapped with "A256KW" |
| `A128GCMKW` | Key wrapping with AES GCM using 128-bit key4 |
| `A192GCMKW` | Key wrapping with AES GCM using 192-bit key4 |
| `A256GCMKW` | Key wrapping with AES GCM using 256-bit key4 |
| `PBES2-HS256+A128KW` | PBES2 with HMAC SHA-256 and "A128KW" wrapping4 |
| `PBES2-HS384+A192KW` | PBES2 with HMAC SHA-384 and "A192KW" wrapping4 |
| PBES2‑HS512+A256KW
| PBES2 with HMAC SHA-512 and "A256KW" wrapping4 |
4. Requires Java 8 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
* Creating, parsing and verifying JSON Web Keys (JWKs) in all standard JWA key formats using native Java `Key` types:
| JWK Key Format | Java `Key` Type | JJWT `Jwk` Type |
|----------------------------|------------------------------------|-------------------|
| Symmetric Key | `SecretKey` | `SecretJwk` |
| Elliptic Curve Public Key | `ECPublicKey` | `EcPublicJwk` |
| Elliptic Curve Private Key | `ECPrivateKey` | `EcPrivateJwk` |
| RSA Public Key | `RSAPublicKey` | `RsaPublicJwk` |
| RSA Private Key | `RSAPrivateKey` | `RsaPrivateJwk` |
| XDH Private Key | `XECPublicKey`5 | `OctetPublicJwk` |
| XDH Private Key | `XECPrivateKey`5 | `OctetPrivateJwk` |
| EdDSA Public Key | `EdECPublicKey`6 | `OctetPublicJwk` |
| EdDSA Private Key | `EdECPublicKey`6 | `OctetPrivateJwk` |
5. Requires Java 11 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
6. Requires Java 15 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
* Convenience enhancements beyond the specification such as
* Payload compression for any large JWT, not just JWEs
* Claims assertions (requiring specific values)
* Claim POJO marshaling and unmarshalling when using a compatible JSON parser (e.g. Jackson)
* Secure Key generation based on desired JWA algorithms
* and more...
### Currently Unsupported Features
* [Non-compact](https://tools.ietf.org/html/rfc7515#section-7.2) serialization and parsing.
This feature may be implemented in a future release. Community contributions are welcome!
## Community
### Getting Help
If you have trouble using JJWT, please first read the documentation on this page before asking questions. We try
very hard to ensure JJWT's documentation is robust, categorized with a table of contents, and up to date for each
release.
#### Questions
If the documentation or the API JavaDoc isn't sufficient, and you either have usability questions or are confused
about something, please [ask your question here](https://github.com/jwtk/jjwt/discussions/new?category=q-a). However:
**Please do not create a GitHub issue to ask a question.**
We use GitHub Issues to track actionable work that requires changes to JJWT's design and/or codebase. If you have a
usability question, instead please
[ask your question here](https://github.com/jwtk/jjwt/discussions/new?category=q-a), and we can convert that to an
issue if necessary.
**If a GitHub Issue is created that does not represent actionable work for JJWT's codebase, it will be promptly
closed.**
#### Bugs, Feature Requests, Ideas and General Discussions
If you do not have a usability question and believe you have a legitimate bug or feature request,
please [discuss it here](https://github.com/jwtk/jjwt/discussions) **_FIRST_**. Please do a quick search first to
see if an existing discussion related to yours exist already and join that existing discussion if necesary.
If you feel like you'd like to help fix a bug or implement the new feature yourself, please read the Contributing
section next before starting any work.
### Contributing
#### Pull Requests
Simple Pull Requests that fix anything other than JJWT core code (documentation, JavaDoc, typos, test cases, etc) are
always appreciated and have a high likelihood of being merged quickly. Please send them!
However, if you want or feel the need to change JJWT's functionality or core code, please do not issue a pull request
without [starting a new JJWT discussion](https://github.com/jwtk/jjwt/discussions) and discussing your desired
changes **first**, _before you start working on it_.
It would be a shame to reject your earnest and genuinely-appreciated pull request if it might not align with the
project's goals, design expectations or planned functionality. We've sadly had to reject large PRs in the past because
they were out of sync with project or design expectations - all because the PR author didn't first check in with
the team first before working on a solution.
So, please [create a new JJWT discussion](https://github.com/jwtk/jjwt/discussions) first to discuss, and then we
can see easily convert the discussion to an issue and then see if (or how) a PR is warranted. Thank you!
#### Help Wanted
If you would like to help, but don't know where to start, please visit the
[Help Wanted Issues](https://github.com/jwtk/jjwt/labels/help%20wanted) page and pick any of the
ones there, and we'll be happy to discuss and answer questions in the issue comments.
If any of those don't appeal to you, no worries! Any help you would like to offer would be
appreciated based on the above caveats concerning [contributing pull reqeuests](#contributing-pull-requests). Feel free
to [discuss or ask questions first](https://github.com/jwtk/jjwt/discussions) if you're not sure. :)
## What is a JSON Web Token?
JSON Web Token (JWT) is a _general-purpose_ text-based messaging format for transmitting information in a
compact and secure way. Contrary to popular belief, JWT is not just useful for sending and receiving identity tokens
on the web - even if that is the most common use case. JWTs can be used as messages for _any_ type of data.
A JWT in its simplest form contains two parts:
1. The primary data within the JWT, called the `payload`, and
2. A JSON `Object` with name/value pairs that represent metadata about the `payload` and the
message itself, called the `header`.
A JWT `payload` can be absolutely anything at all - anything that can be represented as a byte array, such as Strings,
images, documents, etc.
But because a JWT `header` is a JSON `Object`, it would make sense that a JWT `payload` could also be a JSON
`Object` as well. In many cases, developers like the `payload` to be JSON that
represents data about a user or computer or similar identity concept. When used this way, the `payload` is called a
JSON `Claims` object, and each name/value pair within that object is called a `claim` - each piece of information
within 'claims' something about an identity.
And while it is useful to 'claim' something about an identity, really anyone can do that. What's important is that you
_trust_ the claims by verifying they come from a person or computer you trust.
A nice feature of JWTs is that they can be secured in various ways. A JWT can be cryptographically signed (making it
what we call a [JWS](https://tools.ietf.org/html/rfc7515)) or encrypted (making it a
[JWE](https://tools.ietf.org/html/rfc7516)). This adds a powerful layer of verifiability to the JWT - a
JWS or JWE recipient can have a high degree of confidence it comes from someone they trust
by verifying a signature or decrypting it. It is this feature of verifiability that makes JWT a good choice
for sending and receiving secure information, like identity claims.
Finally, JSON with whitespace for human readability is nice, but it doesn't make for a very efficient message
format. Therefore, JWTs can be _compacted_ (and even compressed) to a minimal representation - basically
Base64URL-encoded strings - so they can be transmitted around the web more efficiently, such as in HTTP headers or URLs.
### JWT Example
Once you have a `payload` and `header`, how are they compacted for web transmission, and what does the final JWT
actually look like? Let's walk through a simplified version of the process with some pseudocode:
1. Assume we have a JWT with a JSON `header` and a simple text message payload:
**header**
```
{
"alg": "none"
}
```
**payload**
```
The true sign of intelligence is not knowledge but imagination.
```
2. Remove all unnecessary whitespace in the JSON:
```groovy
String header = '{"alg":"none"}'
String payload = 'The true sign of intelligence is not knowledge but imagination.'
```
3. Get the UTF-8 bytes and Base64URL-encode each:
```groovy
String encodedHeader = base64URLEncode( header.getBytes("UTF-8") )
String encodedPayload = base64URLEncode( payload.getBytes("UTF-8") )
```
4. Join the encoded header and claims with period ('.') characters:
```groovy
String compact = encodedHeader + '.' + encodedPayload + '.'
```
The final concatenated `compact` JWT String looks like this:
```
eyJhbGciOiJub25lIn0.VGhlIHRydWUgc2lnbiBvZiBpbnRlbGxpZ2VuY2UgaXMgbm90IGtub3dsZWRnZSBidXQgaW1hZ2luYXRpb24u.
```
This is called an 'unprotected' JWT because no security was involved - no digital signatures or encryption to
'protect' the JWT to ensure it cannot be changed by 3rd parties.
If we wanted to digitally sign the compact form so that we could at least guarantee that no-one changes the data
without us detecting it, we'd have to perform a few more steps, shown next.
### JWS Example
Instead of a plain text payload, the next example will use probably the most common type of payload - a JSON claims
`Object` containing information about a particular identity. We'll also digitally sign the JWT to ensure it
cannot be changed by a 3rd party without us knowing.
1. Assume we have a JSON `header` and a claims `payload`:
**header**
```json
{
"alg": "HS256"
}
```
**payload**
```json
{
"sub": "Joe"
}
```
In this case, the `header` indicates that the `HS256` (HMAC using SHA-256) algorithm will be used to cryptographically sign
the JWT. Also, the `payload` JSON object has a single claim, `sub` with value `Joe`.
There are a number of standard claims, called [Registered Claims](https://tools.ietf.org/html/rfc7519#section-4.1),
in the specification and `sub` (for 'Subject') is one of them.
2. Remove all unnecessary whitespace in both JSON objects:
```groovy
String header = '{"alg":"HS256"}'
String claims = '{"sub":"Joe"}'
```
3. Get their UTF-8 bytes and Base64URL-encode each:
```groovy
String encodedHeader = base64URLEncode( header.getBytes("UTF-8") )
String encodedClaims = base64URLEncode( claims.getBytes("UTF-8") )
```
4. Concatenate the encoded header and claims with a period character '.' delimiter:
```groovy
String concatenated = encodedHeader + '.' + encodedClaims
```
5. Use a sufficiently-strong cryptographic secret or private key, along with a signing algorithm of your choice
(we'll use HMAC-SHA-256 here), and sign the concatenated string:
```groovy
SecretKey key = getMySecretKey()
byte[] signature = hmacSha256( concatenated, key )
```
6. Because signatures are always byte arrays, Base64URL-encode the signature and join it to the `concatenated` string
with a period character '.' delimiter:
```groovy
String compact = concatenated + '.' + base64URLEncode( signature )
```
And there you have it, the final `compact` String looks like this:
```
eyJhbGciOiJIUzI1NiJ9.eyJzdWIiOiJKb2UifQ.1KP0SsvENi7Uz1oQc07aXTL7kpQG5jBNIybqr60AlD4
```
This is called a 'JWS' - short for _signed_ JWT.
Of course, no one would want to do this manually in code, and worse, if you get anything wrong, you could introduce
serious security problems and weaknesses. As a result, JJWT was created to handle all of this for you: JJWT completely
automates both the creation of JWSs and the parsing and verification of JWSs for you.
### JWE Example
So far we have seen an unprotected JWT and a cryptographically signed JWT (called a 'JWS'). One of the things
that is inherent to both of these two is that all the information within them can be seen by anyone - all the data in
both the header and the payload is publicly visible. JWS just ensures the data hasn't been changed by anyone -
it doesn't prevent anyone from seeing it. Many times, this is just fine because the data within them is not
sensitive information.
But what if you needed to represent information in a JWT that _is_ considered sensitive information - maybe someone's
postal address or social security number or bank account number?
In these cases, we'd want a fully-encrypted JWT, called a 'JWE' for short. A JWE uses cryptography to ensure that the
payload remains fully encrypted _and_ authenticated so unauthorized parties cannot see data within, nor change the data
without being detected. Specifically, the JWE specification requires that
[Authenticated Encryption with Associated Data](https://en.wikipedia.org/wiki/Authenticated_encryption#Authenticated_encryption_with_associated_data_(AEAD))
algorithms are used to fully encrypt and protect data.
A full overview of AEAD algorithms are out of scope for this documentation, but here's an example of a final compact
JWE that utilizes these algorithms (line breaks are for readability only):
```
eyJhbGciOiJBMTI4S1ciLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0.
6KB707dM9YTIgHtLvtgWQ8mKwboJW3of9locizkDTHzBC2IlrT1oOQ.
AxY8DCtDaGlsbGljb3RoZQ.
KDlTtXchhZTGufMYmOYGS4HffxPSUrfmqCHXaI9wOGY.
U0m_YmjN04DJvceFICbCVQ
```
Next we'll cover how to install JJWT in your project, and then we'll see how to use JJWT's nice fluent API instead
of risky string manipulation to quickly and safely build JWTs, JWSs, and JWEs.
## Installation
Use your favorite Maven-compatible build tool to pull the dependencies from Maven Central.
The dependencies could differ slightly if you are working with a [JDK project](#install-jdk) or an
[Android project](#install-android).
### JDK Projects
If you're building a (non-Android) JDK project, you will want to define the following dependencies:
#### Maven
```xml
io.jsonwebtoken
jjwt-api
JJWT_RELEASE_VERSION
io.jsonwebtoken
jjwt-impl
JJWT_RELEASE_VERSION
runtime
io.jsonwebtoken
jjwt-jackson
JJWT_RELEASE_VERSION
runtime
```
#### Gradle
```groovy
dependencies {
implementation 'io.jsonwebtoken:jjwt-api:JJWT_RELEASE_VERSION'
runtimeOnly 'io.jsonwebtoken:jjwt-impl:JJWT_RELEASE_VERSION'
runtimeOnly 'io.jsonwebtoken:jjwt-jackson:JJWT_RELEASE_VERSION' // or 'io.jsonwebtoken:jjwt-gson:JJWT_RELEASE_VERSION' for gson
/*
Uncomment this next dependency if you are using:
- JDK 10 or earlier, and you want to use RSASSA-PSS (PS256, PS384, PS512) signature algorithms.
- JDK 10 or earlier, and you want to use EdECDH (X25519 or X448) Elliptic Curve Diffie-Hellman encryption.
- JDK 14 or earlier, and you want to use EdDSA (Ed25519 or Ed448) Elliptic Curve signature algorithms.
It is unnecessary for these algorithms on JDK 15 or later.
*/
// runtimeOnly 'org.bouncycastle:bcprov-jdk18on:1.76' // or bcprov-jdk15to18 on JDK 7
}
```
### Android Projects
Android projects will want to define the following dependencies and Proguard exclusions, and optional
BouncyCastle `Provider`:
#### Dependencies
Add the dependencies to your project:
```groovy
dependencies {
api('io.jsonwebtoken:jjwt-api:JJWT_RELEASE_VERSION')
runtimeOnly('io.jsonwebtoken:jjwt-impl:JJWT_RELEASE_VERSION')
runtimeOnly('io.jsonwebtoken:jjwt-orgjson:JJWT_RELEASE_VERSION') {
exclude(group: 'org.json', module: 'json') //provided by Android natively
}
/*
Uncomment this next dependency if you want to use:
- RSASSA-PSS (PS256, PS384, PS512) signature algorithms.
- EdECDH (X25519 or X448) Elliptic Curve Diffie-Hellman encryption.
- EdDSA (Ed25519 or Ed448) Elliptic Curve signature algorithms.
** AND ALSO ensure you enable the BouncyCastle provider as shown below **
*/
//implementation('org.bouncycastle:bcprov-jdk18on:1.76') // or bcprov-jdk15to18 for JDK 7
}
```
#### Proguard
You can use the following [Android Proguard](https://developer.android.com/studio/build/shrink-code) exclusion rules:
```
-keepattributes InnerClasses
-keep class io.jsonwebtoken.** { *; }
-keepnames class io.jsonwebtoken.* { *; }
-keepnames interface io.jsonwebtoken.* { *; }
-keep class org.bouncycastle.** { *; }
-keepnames class org.bouncycastle.** { *; }
-dontwarn org.bouncycastle.**
```
#### Bouncy Castle
If you want to use JWT RSASSA-PSS algorithms (i.e. `PS256`, `PS384`, and `PS512`), EdECDH (`X25512` or `X448`)
Elliptic Curve Diffie-Hellman encryption, EdDSA (`Ed25519` or `Ed448`) signature algorithms, or you just want to
ensure your Android application is running an updated version of BouncyCastle, you will need to:
1. Uncomment the BouncyCastle dependency as commented above in the [dependencies](#install-android-dependencies) section.
2. Replace the legacy Android custom `BC` provider with the updated one.
Provider registration needs to be done _early_ in the application's lifecycle, preferably in your application's
main `Activity` class as a static initialization block. For example:
```kotlin
class MainActivity : AppCompatActivity() {
companion object {
init {
Security.removeProvider("BC") //remove old/legacy Android-provided BC provider
Security.addProvider(BouncyCastleProvider()) // add 'real'/correct BC provider
}
}
// ... etc ...
}
```
### Understanding JJWT Dependencies
Notice the above JJWT dependency declarations all have only one compile-time dependency and the rest are declared as
_runtime_ dependencies.
This is because JJWT is designed so you only depend on the APIs that are explicitly designed for you to use in
your applications and all other internal implementation details - that can change without warning - are relegated to
runtime-only dependencies. This is an extremely important point if you want to ensure stable JJWT usage and
upgrades over time:
> **Warning**
>
> **JJWT guarantees semantic versioning compatibility for all of its artifacts _except_ the `jjwt-impl` .jar. No such
guarantee is made for the `jjwt-impl` .jar and internal changes in that .jar can happen at any time. Never add the
`jjwt-impl` .jar to your project with `compile` scope - always declare it with `runtime` scope.**
This is done to benefit you: great care goes into curating the `jjwt-api` .jar and ensuring it contains what you need
and remains backwards compatible as much as is possible so you can depend on that safely with compile scope. The
runtime `jjwt-impl` .jar strategy affords the JJWT developers the flexibility to change the internal packages and
implementations whenever and however necessary. This helps us implement features, fix bugs, and ship new releases to
you more quickly and efficiently.
## Quickstart
Most complexity is hidden behind a convenient and readable builder-based
[fluent interface](http://en.wikipedia.org/wiki/Fluent_interface), great for relying on IDE auto-completion to write
code quickly. Here's an example:
```java
import io.jsonwebtoken.Jwts;
import io.jsonwebtoken.security.Keys;
import java.security.Key;
// We need a signing key, so we'll create one just for this example. Usually
// the key would be read from your application configuration instead.
SecretKey key = Jwts.SIG.HS256.key().build();
String jws = Jwts.builder().subject("Joe").signWith(key).compact();
```
How easy was that!?
In this case, we are:
1. *building* a JWT that will have the
[registered claim](https://tools.ietf.org/html/rfc7519#section-4.1) `sub` (Subject) set to `Joe`. We are then
2. *signing* the JWT using a key suitable for the HMAC-SHA-256 algorithm. Finally, we are
3. *compacting* it into its final `String` form. A signed JWT is called a 'JWS'.
The resultant `jws` String looks like this:
```
eyJhbGciOiJIUzI1NiJ9.eyJzdWIiOiJKb2UifQ.1KP0SsvENi7Uz1oQc07aXTL7kpQG5jBNIybqr60AlD4
```
Now let's verify the JWT (you should always discard JWTs that don't match an expected signature):
```java
assert Jwts.parser().verifyWith(key).build().parseClaimsJws(jws).getPayload().getSubject().equals("Joe");
```
There are two things going on here. The `key` from before is being used to verify the signature of the JWT. If it
fails to verify the JWT, a `SignatureException` (which extends `JwtException`) is thrown. Assuming the JWT is
verified, we parse the claims and assert that that subject is set to `Joe`. You have to love code one-liners
that pack a punch!
> **Note**
>
> **Type-safe JWTs:** To get a type-safe `Claims` JWT result, call the `parseClaimsJws` method (since there are many
similar methods available). You will get an `UnsupportedJwtException` if you parse your JWT with wrong method.
But what if parsing or signature validation failed? You can catch `JwtException` and react accordingly:
```java
try {
Jwts.parser().verifyWith(key).build().parseClaimsJws(compactJws);
//OK, we can trust this JWT
} catch (JwtException e) {
//don't trust the JWT!
}
```
Now that we've had a quickstart 'taste' of how to create and parse JWTs, let's cover JJWT's API in-depth.
## Creating a JWT
You create a JWT as follows:
1. Use the `Jwts.builder()` method to create a `JwtBuilder` instance.
2. Optionally set any [`header` parameters](#jwt-header) as desired.
3. Call builder methods to set the payload [content](#jwt-content) or [claims](#jwt-claims).
4. Optionally call `signWith` or `encryptWith` methods if you want to digitally sign or encrypt the JWT.
5. Call the `compact()` method to produce the resulting compact JWT string.
For example:
```java
String jwt = Jwts.builder() // (1)
.header() // (2) optional
.keyId("aKeyId")
.and()
.subject("Bob") // (3) JSON Claims, or
//.content(aByteArray, "text/plain") // any byte[] content, with media type
.signWith(signingKey) // (4) if signing, or
//.encryptWith(key, keyAlg, encryptionAlg) // if encrypting
.compact(); // (5)
```
* The JWT `payload` may be either `byte[]` content (via `content`) _or_ JSON Claims
(such as `subject`, `claims`, etc), but not both.
* Either digital signatures (`signWith`) or encryption (`encryptWith`) may be used, but not both.
> **Warning**
>
> **Unprotected JWTs**: If you do not use the `signWith` or `encryptWith` builder methods, **an Unprotected JWT will be
> created, which offers no security protection at all**. If you need security protection, consider either
> [digitally signing](#jws) or [encrypting](#jwe) the JWT before calling the `compact()` builder method.
### JWT Header
A JWT header is a JSON `Object` that provides metadata about the contents, format, and any cryptographic operations
relevant to the JWT `payload`. JJWT provides a number of ways of setting the entire header and/or multiple individual
header parameters (name/value pairs).
#### JwtBuilder Header
The easiest and recommended way to set one or more JWT header parameters (name/value pairs) is to use the
`JwtBuilder`'s `header()` builder as desired, and then call its `and()` method to return back
to the `JwtBuilder` for further configuration. For example:
```java
String jwt = Jwts.builder()
.header() // <----
.keyId("aKeyId")
.x509Url(aUri)
.add("someName", anyValue)
.add(mapValues)
// ... etc ...
.and() // go back to the JwtBuilder
.subject("Joe") // resume JwtBuilder calls...
// ... etc ...
.compact();
```
The `JwtBuilder` `header()` builder also supports automatically calculating X.509 thumbprints and other builder-style benefits that
a simple property getter/setter object would not do.
> **Note**
>
> **Automatic Headers**: You do not need to set the `alg`, `enc` or `zip` headers - JJWT will always set them
> automatically as needed.
##### Custom Header Parameters
In addition to type-safe builder methods for standard header parameters, `JwtBuilder.header()` can also support
arbitrary name/value pairs via the `add` method:
```java
Jwts.builder()
.header()
.add("aHeaderName", aValue)
// ... etc ...
.and() // return to the JwtBuilder
// ... etc ...
```
##### Header Parameter Map
The `add` method is also overloaded to support multiple parameters in a `Map`:
```java
Jwts.builder()
.header()
.add(multipleHeaderParamsMap)
// ... etc ...
.and() // return to the JwtBuilder
// ... etc ...
```
#### Jwts HeaderBuilder
Using `Jwts.builder().header()` shown above is the preferred way to modify a header when using the `JwtBuilder`.
However, if you would like to create a 'standalone' `Header` outside of the context of using the `JwtBuilder`, you
can use `Jwts.header()` instead to return an independent `Header` builder. For example:
```java
Header header = Jwts.header()
.keyId("aKeyId")
.x509Url(aUri)
.add("someName", anyValue)
.add(mapValues)
// ... etc ...
.build() // <---- not 'and()'
```
There are only two differences between `Jwts.header()` and `Jwts.builder().header()`:
1. `Jwts.header()` builds a 'detached' `Header` that is not associated with any particular JWT, whereas
`Jwts.builder().header()` always modifies the header of the immediate JWT being constructed by its parent
`JwtBuilder`.
2. `Jwts.header()` has a `build()` method to produce an explicit `Header` instance and
`Jwts.builder().header()` does not (it has an `and()` method instead) because its parent `JwtBuilder` will implicitly
create the header instance when necessary.
A standalone header might be useful if you want to aggregate common header parameters in a single 'template'
instance so you don't have to repeat them for each `JwtBuilder` usage. Then this 'template' `Header` can be used to
populate `JwtBuilder` usages by just appending it to the `JwtBuilder` header, for example:
```java
// perhaps somewhere in application configuration:
Header commonHeaders = Jwts.header()
.issuer("My Company")
// ... etc ...
.build();
// --------------------------------
// somewhere else during actual Jwt construction:
String jwt = Jwts.builder()
.header()
.add(commonHeaders) // <----
.add("specificHeader", specificValue) // jwt-specific headers...
.and()
.subject("whatever")
// ... etc ...
.compact();
```
### JWT Payload
A JWT `payload` can be anything at all - anything that can be represented as a byte array, such as text, images,
documents, and more. But since a JWT `header` is always JSON, it makes sense that the `payload` could also be JSON,
especially for representing identity claims.
As a result, the `JwtBuilder` supports two distinct payload options:
* `content` if you would like the payload to be arbitrary byte array content, or
* `claims` (and supporting helper methods) if you would like the payload to be a JSON Claims `Object`.
Either option may be used, but not both. Using both will cause `compact()` to throw an exception.
#### Arbitrary Content
You can set the JWT payload to be any arbitrary byte array content by using the `JwtBuilder` `content` method.
For example:
```java
byte[] content = "Hello World".getBytes(StandardCharsets.UTF_8);
String jwt = Jwts.builder()
.content(content, "text/plain") // <---
// ... etc ...
.build();
```
Notice this particular example of `content` uses the two-argument convenience variant:
1. The first argument is the actual byte content to set as the JWT payload
2. The second argument is a String identifier of an IANA Media Type.
The second argument will cause the `JwtBuilder` to automatically set the `cty` (Content Type) header according to the
JWT specification's [recommended compact format](https://www.rfc-editor.org/rfc/rfc7515.html#section-4.1.10).
This two-argument variant is typically recommended over the single-argument `content(byte[])` method because it
guarantees the JWT recipient can inspect the `cty` header to determine how to convert the `payload` byte array into
a final form that the application can use.
Without setting the `cty` header, the JWT recipient _must_ know via out-of-band (external) information how to process
the byte array, which is usually less convenient and always requires code changes if the content format ever changes.
For these reasons, it is strongly recommended to use the two-argument `content` method variant.
#### JWT Claims
Instead of a content byte array, a JWT payload may contain assertions or claims for a JWT recipient. In
this case, the payload is a `Claims` JSON `Object`, and JJWT supports claims creation with type-safe
builder methods.
##### Standard Claims
The `JwtBuilder` provides convenient builder methods for standard registered Claim names defined in the JWT
specification. They are:
* `issuer`: sets the [`iss` (Issuer) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.1)
* `subject`: sets the [`sub` (Subject) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.2)
* `audience`: sets the [`aud` (Audience) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.3)
* `expiration`: sets the [`exp` (Expiration Time) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.4)
* `notBefore`: sets the [`nbf` (Not Before) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.5)
* `issuedAt`: sets the [`iat` (Issued At) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.6)
* `id`: sets the [`jti` (JWT ID) Claim](https://tools.ietf.org/html/rfc7519#section-4.1.7)
For example:
```java
String jws = Jwts.builder()
.issuer("me")
.subject("Bob")
.audience("you")
.expiration(expiration) //a java.util.Date
.notBefore(notBefore) //a java.util.Date
.issuedAt(new Date()) // for example, now
.id(UUID.randomUUID().toString()) //just an example id
/// ... etc ...
```
##### Custom Claims
If you need to set one or more custom claims that don't match the standard setter method claims shown above, you
can simply call the `JwtBuilder` `claim` method one or more times as needed:
```java
String jws = Jwts.builder()
.claim("hello", "world")
// ... etc ...
```
Each time `claim` is called, it simply appends the key-value pair to an internal `Claims` builder, potentially
overwriting any existing identically-named key/value pair.
Obviously, you do not need to call `claim` for any [standard claim name](#jws-create-claims-standard), and it is
recommended instead to call the standard respective type-safe named builder method as this enhances readability.
##### Claims Map
If you want to add multiple claims at once, you can use `JwtBuilder` `claims(Map)` method:
```java
Map claims = getMyClaimsMap(); //implement me
String jws = Jwts.builder()
.claims(claims)
// ... etc ...
```
### JWT Compression
If your JWT payload is large (contains a lot of data), you might want to compress the JWT to reduce its size. Note
that this is *not* a standard feature for all JWTs - only JWEs - and is not likely to be supported by other JWT
libraries for non-JWE tokens. JJWT supports compression for both JWSs and JWEs, however.
Please see the main [Compression](#compression) section to see how to compress and decompress JWTs.
## Reading a JWT
You read (parse) a JWT as follows:
1. Use the `Jwts.parser()` method to create a `JwtParserBuilder` instance.
2. Optionally call `keyLocator`, `verifyWith` or `decryptWith` methods if you expect to parse [signed](#jws) or [encrypted](#jwe) JWTs.
3. Call the `build()` method on the `JwtParserBuilder` to create and return a thread-safe `JwtParser`.
4. Call one of the various `parse*` methods with your compact JWT string, depending on the type of JWT you expect.
5. Wrap the `parse*` call in a try/catch block in case parsing, signature verification, or decryption fails.
For example:
```java
Jwt,?> jwt;
try {
jwt = Jwts.parser() // (1)
.keyLocator(keyLocator) // (2) dynamically locate signing or encryption keys
//.verifyWith(key) // or a constant key used to verify all signed JWTs
//.decryptWith(key) // or a constant key used to decrypt all encrypted JWTs
.build() // (3)
.parse(compact); // (4) or parseClaimsJws, parseClaimsJwe, parseContentJws, etc
// we can safely trust the JWT
catch (JwtException ex) { // (5)
// we *cannot* use the JWT as intended by its creator
}
```
> **Note**
>
> **Type-safe JWTs:** If you are certain your parser will only ever encounter a specific kind of JWT (for example, you only
> ever use signed JWTs with `Claims` payloads, or encrypted JWTs with `byte[]` content payloads, etc), you can call the
> associated type-safe `parseClaimsJws`, `parseClaimsJwe`, (etc) method variant instead of the generic `parse` method.
>
> These `parse*` methods will return the type-safe JWT you are expecting, for example, a `Jws` or `Jwe`
> instead of a generic `Jwt,?>` instance.
### Constant Parsing Key
If the JWT parsed is a JWS or JWE, a key will be necessary to verify the signature or decrypt it. If a JWS and
signature verification fails, or if a JWE and decryption fails, the JWT cannot be safely trusted and should be
discarded.
So which key do we use?
* If parsing a JWS and the JWS was signed with a `SecretKey`, the same `SecretKey` should be specified on the
`JwtParserBuilder`. For example:
```java
Jwts.parser()
.verifyWith(secretKey) // <----
.build()
.parseClaimsJws(jwsString);
```
* If parsing a JWS and the JWS was signed with a `PrivateKey`, that key's corresponding `PublicKey` (not the
`PrivateKey`) should be specified on the `JwtParserBuilder`. For example:
```java
Jwts.parser()
.verifyWith(publicKey) // <---- publicKey, not privateKey
.build()
.parseClaimsJws(jwsString);
```
* If parsing a JWE and the JWE was encrypted with direct encryption using a `SecretKey`, the same `SecretKey` should be
specified on the `JwtParserBuilder`. For example:
```java
Jwts.parser()
.decryptWith(secretKey) // <---- or a Password from Keys.password(charArray)
.build()
.parseClaimsJwe(jweString);
```
* If parsing a JWE and the JWE was encrypted with a key algorithm using with a `PublicKey`, that key's corresponding
`PrivateKey` (not the `PublicKey`) should be specified on the `JwtParserBuilder`. For example:
```java
Jwts.parser()
.decryptWith(privateKey) // <---- privateKey, not publicKey
.build()
.parseClaimsJwe(jweString);
```
#### Multiple Keys?
But you might have noticed something - what if your application doesn't use just a single `SecretKey` or `KeyPair`? What
if JWSs and JWEs can be created with different `SecretKey`s or public/private keys, or a combination of both? How do
you know which key to specify if you don't inspect the JWT first?
In these cases, you can't call the `JwtParserBuilder`'s `verifyWith` or `decryptWith` methods with a single key -
instead, you'll need to configure a parsing Key Locator, discussed next.
### Dynamic Key Lookup
It is common in many applications to receive JWTs that can be encrypted or signed by different cryptographic keys. For
example, maybe a JWT created to assert a specific user identity uses a Key specific to that exact user. Or perhaps JWTs
specific to a particular customer all use that customer's Key. Or maybe your application creates JWTs that are
encrypted with a key specific to your application for your own use (e.g. a user session token).
In all of these and similar scenarios, you won't know which key was used to sign or encrypt a JWT until the JWT is
received, at parse time, so you can't 'hard code' any verification or decryption key using the `JwtParserBuilder`'s
`verifyWith` or `decryptWith` methods. Those are only to be used when the same key is used to verify or decrypt
*all* JWSs or JWEs, which won't work for dynamically signed or encrypted JWTs.
#### Key Locator
If you need to support dynamic key lookup when encountering JWTs, you'll need to implement
the `Locator` interface and specify an instance on the `JwtParserBuilder` via the `keyLocator` method. For
example:
```java
Locator keyLocator = getMyKeyLocator();
Jwts.parser()
.keyLocator(keyLocator) // <----
.build()
// ... etc ...
```
A `Locator` is used to lookup _both_ JWS signature verification keys _and_ JWE decryption keys. You need to
determine which key to return based on information in the JWT `header`, for example:
```java
public class MyKeyLocator extends LocatorAdapter {
@Override
public Key locate(ProtectedHeader> header) { // a JwsHeader or JweHeader
// implement me
}
}
```
The `JwtParser` will invoke the `locate` method after parsing the JWT `header`, but _before parsing the `payload`,
or verifying any JWS signature or decrypting any JWE ciphertext_. This allows you to inspect the `header` argument
for any information that can help you look up the `Key` to use for verifying _that specific jwt_. This is very
powerful for applications with more complex security models that might use different keys at different times or for
different users or customers.
#### Key Locator Strategy
What data might you inspect to determine how to lookup a signature verification or decryption key?
The JWT specifications' preferred approach is to set a `kid` (Key ID) header value when the JWT is being created,
for example:
```java
Key key = getSigningKey(); // or getEncryptionKey() for JWE
String keyId = getKeyId(key); //any mechanism you have to associate a key with an ID is fine
String jws = Jwts.builder()
.header().keyId(keyId).and() // <--- add `kid` header
.signWith(key) // for JWS
//.encryptWith(key, keyAlg, encryptionAlg) // for JWE
.compact();
```
Then during parsing, your `Locator` implementation can inspect the `header` to get the `kid` value and then use it
to look up the verification or decryption key from somewhere, like a database, keystore or Hardware Security Module
(HSM). For example:
```java
public class MyKeyLocator extends LocatorAdapter {
@Override
public Key locate(ProtectedHeader> header) { // both JwsHeader and JweHeader extend ProtectedHeader
//inspect the header, lookup and return the verification key
String keyId = header.getKeyId(); //or any other parameter that you need to inspect
Key key = lookupKey(keyId); //implement me
return key;
}
}
```
Note that inspecting the `header.getKeyId()` is just the most common approach to look up a key - you could inspect any
number of header parameters to determine how to lookup the verification or decryption key. It is all based on how
the JWT was created.
If you extend `LocatorAdapter` as shown above, but for some reason have different lookup strategies for
signature verification keys versus decryption keys, you can forego overriding the `locate(ProtectedHeader>)` method
in favor of two respective `locate(JwsHeader)` and `locate(JweHeader)` methods:
```java
public class MyKeyLocator extends LocatorAdapter {
@Override
public Key locate(JwsHeader header) {
String keyId = header.getKeyId(); //or any other parameter that you need to inspect
return lookupSignatureVerificationKey(keyId); //implement me
}
@Override
public Key locate(JweHeader header) {
String keyId = header.getKeyId(); //or any other parameter// that you need to inspect
return lookupDecryptionKey(keyId); //implement me
}
}
```
> **Note**
>
> **Simpler Lookup**: If possible, try to keep the key lookup strategy the same between JWSs and JWEs (i.e. using
> only `locate(ProtectedHeader>)`), preferably using only
> the `kid` (Key ID) header value or perhaps a public key thumbprint. You will find the implementation is much
> simpler and easier to maintain over time, and also creates smaller headers for compact transmission.
#### Key Locator Return Values
Regardless of which implementation strategy you choose, remember to return the appropriate type of key depending
on the type of JWS or JWE algorithm used. That is:
* For JWS:
* For HMAC-based signature algorithms, the returned verification key should be a `SecretKey`, and,
* For asymmetric signature algorithms, the returned verification key should be a `PublicKey` (not a `PrivateKey`).
* For JWE:
* For JWE direct encryption, the returned decryption key should be a `SecretKey`.
* For password-based key derivation algorithms, the returned decryption key should be a
`io.jsonwebtoken.security.Password`. You can create a `Password` instance by calling
`Keys.password(char[] passwordCharacters)`.
* For asymmetric key management algorithms, the returned decryption key should be a `PrivateKey` (not a `PublicKey`).
#### Provider-constrained Keys
If any verification or decryption key returned from a Key `Locator` must be used with a specific security `Provider`
(such as for PKCS11 or Hardware Security Module (HSM) keys), you must make that `Provider` available for JWT parsing
in one of 3 ways, listed in order of recommendation and simplicity:
1. [Configure the Provider in the JVM](https://docs.oracle.com/en/java/javase/17/security/howtoimplaprovider.html#GUID-831AA25F-F702-442D-A2E4-8DA6DEA16F33),
either by modifying the `java.security` file or by registering the `Provider` dynamically via
[Security.addProvider(Provider)](https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/security/Security.html#addProvider(java.security.Provider)).
This is the recommended approach so you do not need to modify code anywhere that may need to parse JWTs.
2. Set the `Provider` as the parser default by calling `JwtParserBuilder#provider(Provider)`. This will
ensure the provider is used by default with _all_ located keys unless overridden by a key-specific Provider. This
is only recommended when you are confident that all JWTs encountered by the parser instance will use keys
attributed to the same `Provider`, unless overridden by a specific key.
3. Associate the `Provider` with a specific key using `Keys.builder` so it is used for that key only. This option is
useful if some located keys require a specific provider, while other located keys can assume a default provider. For
example:
```java
public Key locate(Header> header) {
PrivateKey /* or SecretKey */ key = findKey(header); // implement me
Provider keySpecificProvider = findKeyProvider(key); // implement me
if (keySpecificProvider != null) {
// Ensure the key-specific provider (e.g. for PKCS11 or HSM) will be used
// during decryption with the KeyAlgorithm in the JWE 'alg' header
return Keys.builder(key).provider(keySpecificProvider).build();
}
// otherwise default provider is fine:
return key;
}
```
### Claim Assertions
You can enforce that the JWT you are parsing conforms to expectations that you require and are important for your
application.
For example, let's say that you require that the JWT you are parsing has a specific `sub` (subject) value,
otherwise you may not trust the token. You can do that by using one of the various `require`* methods on the
`JwtParserBuilder`:
```java
try {
Jwts.parser().requireSubject("jsmith")/* etc... */.build().parse(s);
} catch (InvalidClaimException ice) {
// the sub claim was missing or did not have a 'jsmith' value
}
```
If it is important to react to a missing vs an incorrect value, instead of catching `InvalidClaimException`,
you can catch either `MissingClaimException` or `IncorrectClaimException`:
```java
try {
Jwts.parser().requireSubject("jsmith")/* etc... */.build().parse(s);
} catch(MissingClaimException mce) {
// the parsed JWT did not have the sub claim
} catch(IncorrectClaimException ice) {
// the parsed JWT had a sub claim, but its value was not equal to 'jsmith'
}
```
You can also require custom claims by using the `require(claimName, requiredValue)` method - for example:
```java
try {
Jwts.parser().require("myClaim", "myRequiredValue")/* etc... */.build().parse(s);
} catch(InvalidClaimException ice) {
// the 'myClaim' claim was missing or did not have a 'myRequiredValue' value
}
```
(or, again, you could catch either `MissingClaimException` or `IncorrectClaimException` instead).
Please see the `JwtParserBuilder` class and/or JavaDoc for a full list of the various `require`* methods you may use
for claims assertions.
### Accounting for Clock Skew
When parsing a JWT, you might find that `exp` or `nbf` claim assertions fail (throw exceptions) because the clock on
the parsing machine is not perfectly in sync with the clock on the machine that created the JWT. This can cause
obvious problems since `exp` and `nbf` are time-based assertions, and clock times need to be reliably in sync for shared
assertions.
You can account for these differences (usually no more than a few minutes) when parsing using the `JwtParserBuilder`'s
`clockSkewSeconds`. For example:
```java
long seconds = 3 * 60; //3 minutes
Jwts.parser()
.clockSkewSeconds(seconds) // <----
// ... etc ...
.build()
.parse(jwt);
```
This ensures that minor clock differences between the machines can be ignored. Two or three minutes should be more than
enough; it would be fairly strange if a production machine's clock was more than 5 minutes difference from most
atomic clocks around the world.
#### Custom Clock Support
If the above `clockSkewSeconds` isn't sufficient for your needs, the timestamps created
during parsing for timestamp comparisons can be obtained via a custom time source. Call the `JwtParserBuilder`'s
`clock` method with an implementation of the `io.jsonwebtoken.Clock` interface. For example:
```java
Clock clock = new MyClock();
Jwts.parser().clock(myClock) //... etc ...
```
The `JwtParser`'s default `Clock` implementation simply returns `new Date()` to reflect the time when parsing occurs,
as most would expect. However, supplying your own clock could be useful, especially when writing test cases to
guarantee deterministic behavior.
### JWT Decompression
If you used JJWT to compress a JWT and you used a custom compression algorithm, you will need to tell the
`JwtParserBuilder` how to resolve your `CompressionAlgorithm` to decompress the JWT.
Please see the [Compression](#compression) section below to see how to decompress JWTs during parsing.
## Signed JWTs
The JWT specification provides for the ability to
[cryptographically _sign_](https://en.wikipedia.org/wiki/Digital_signature) a JWT. Signing a JWT:
1. guarantees the JWT was created by someone we know (it is authentic) as well as
2. guarantees that no-one has manipulated or changed the JWT after it was created (its integrity is maintained).
These two properties - authenticity and integrity - assure us that a JWT contains information we can trust. If a
JWT fails authenticity or integrity checks, we should always reject that JWT because we can't trust it.
But before we dig in to showing you how to create a JWS using JJWT, let's briefly discuss Signature Algorithms and
Keys, specifically as they relate to the JWT specifications. Understanding them is critical to being able to create a
JWS properly.
### Standard Signature Algorithms
The JWT specifications identify 13 standard signature algorithms - 3 secret key algorithms and 10 asymmetric
key algorithms:
| Identifier | Signature Algorithm |
| --- | --- |
| `HS256` | HMAC using SHA-256 |
| `HS384` | HMAC using SHA-384 |
| `HS512` | HMAC using SHA-512 |
| `ES256` | ECDSA using P-256 and SHA-256 |
| `ES384` | ECDSA using P-384 and SHA-384 |
| `ES512` | ECDSA using P-521 and SHA-512 |
| `RS256` | RSASSA-PKCS-v1_5 using SHA-256 |
| `RS384` | RSASSA-PKCS-v1_5 using SHA-384 |
| `RS512` | RSASSA-PKCS-v1_5 using SHA-512 |
| `PS256` | RSASSA-PSS using SHA-256 and MGF1 with SHA-2561 |
| `PS384` | RSASSA-PSS using SHA-384 and MGF1 with SHA-3841 |
| `PS512` | RSASSA-PSS using SHA-512 and MGF1 with SHA-5121 |
| `EdDSA` | Edwards-Curve Digital Signature Algorithm (EdDSA)2 |
1. Requires Java 11 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
2. Requires Java 15 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
These are all represented as constants in the `io.jsonwebtoken.Jwts.SIG` registry class.
### Signature Algorithms Keys
What's really important about the above standard signature algorithms - other than their security properties - is that
the JWT specification [RFC 7518, Sections 3.2 through 3.5](https://tools.ietf.org/html/rfc7518#section-3)
_requires_ (mandates) that you MUST use keys that are sufficiently strong for a chosen algorithm.
This means that JJWT - a specification-compliant library - will also enforce that you use sufficiently strong keys
for the algorithms you choose. If you provide a weak key for a given algorithm, JJWT will reject it and throw an
exception.
This is not because we want to make your life difficult, we promise! The reason why the JWT specification, and
consequently JJWT, mandates key lengths is that the security model of a particular algorithm can completely break
down if you don't adhere to the mandatory key properties of the algorithm, effectively having no security at all. No
one wants completely insecure JWTs, right? Right!
So what are the key strength requirements?
#### HMAC-SHA
JWT HMAC-SHA signature algorithms `HS256`, `HS384`, and `HS512` require a secret key that is _at least_ as many bits as
the algorithm's signature (digest) length per [RFC 7512 Section 3.2](https://tools.ietf.org/html/rfc7518#section-3.2).
This means:
* `HS256` is HMAC-SHA-256, and that produces digests that are 256 bits (32 bytes) long, so `HS256` _requires_ that you
use a secret key that is at least 32 bytes long.
* `HS384` is HMAC-SHA-384, and that produces digests that are 384 bits (48 bytes) long, so `HS384` _requires_ that you
use a secret key that is at least 48 bytes long.
* `HS512` is HMAC-SHA-512, and that produces digests that are 512 bits (64 bytes) long, so `HS512` _requires_ that you
use a secret key that is at least 64 bytes long.
#### RSA
JWT RSA signature algorithms `RS256`, `RS384`, `RS512`, `PS256`, `PS384` and `PS512` all require a minimum key length
(aka an RSA modulus bit length) of `2048` bits per RFC 7512 Sections
[3.3](https://tools.ietf.org/html/rfc7518#section-3.3) and [3.5](https://tools.ietf.org/html/rfc7518#section-3.5).
Anything smaller than this (such as 1024 bits) will be rejected with an `WeakKeyException`.
That said, in keeping with best practices and increasing key lengths for security longevity, JJWT
recommends that you use:
* at least 2048 bit keys with `RS256` and `PS256`
* at least 3072 bit keys with `RS384` and `PS384`
* at least 4096 bit keys with `RS512` and `PS512`
These are only JJWT suggestions and not requirements. JJWT only enforces JWT specification requirements and
for any RSA key, the requirement is the RSA key (modulus) length in bits MUST be >= 2048 bits.
#### Elliptic Curve
JWT Elliptic Curve signature algorithms `ES256`, `ES384`, and `ES512` all require a key length
(aka an Elliptic Curve order bit length) equal to the algorithm signature's individual
`R` and `S` components per [RFC 7512 Section 3.4](https://tools.ietf.org/html/rfc7518#section-3.4). This means:
* `ES256` requires that you use a private key that is exactly 256 bits (32 bytes) long.
* `ES384` requires that you use a private key that is exactly 384 bits (48 bytes) long.
* `ES512` requires that you use a private key that is exactly 521 bits (65 or 66 bytes) long (depending on format).
#### Edwards Curve
The JWT Edwards Curve signature algorithm `EdDSA` supports two sizes of private and public `EdECKey`s (these types
were introduced in Java 15):
* `Ed25519` algorithm keys must be 256 bits (32 bytes) long and produce signatures 512 bits (64 bytes) long.
* `Ed448` algorithm keys must be 456 bits (57 bytes) long and produce signatures 912 bits (114 bytes) long.
#### Creating Safe Keys
If you don't want to think about bit length requirements or just want to make your life easier, JJWT has
provided convenient builder classes that can generate sufficiently secure keys for any given
JWT signature algorithm you might want to use.
##### Secret Keys
If you want to generate a sufficiently strong `SecretKey` for use with the JWT HMAC-SHA algorithms, use the respective
algorithm's `key()` builder method:
```java
SecretKey key = Jwts.SIG.HS256.key().build(); //or HS384.key() or HS512.key()
```
Under the hood, JJWT uses the JCA default provider's `KeyGenerator` to create a secure-random key with the correct
minimum length for the given algorithm.
If you want to specify a specific JCA `Provider` or `SecureRandom` to use during key generation, you may specify those
as builder arguments. For example:
```java
SecretKey key = Jwts.SIG.HS256.key().provider(aProvider).random(aSecureRandom).build();
```
If you need to save this new `SecretKey`, you can Base64 (or Base64URL) encode it:
```java
String secretString = Encoders.BASE64.encode(key.getEncoded());
```
Ensure you save the resulting `secretString` somewhere safe -
[Base64-encoding is not encryption](#base64-not-encryption), so it's still considered sensitive information. You can
further encrypt it, etc, before saving to disk (for example).
##### Asymmetric Keys
If you want to generate sufficiently strong Elliptic Curve or RSA asymmetric key pairs for use with JWT ECDSA or RSA
algorithms, use an algorithm's respective `keyPair()` builder method:
```java
KeyPair keyPair = Jwts.SIG.RS256.keyPair().build(); //or RS384, RS512, PS256, etc...
```
Once you've generated a `KeyPair`, you can use the private key (`keyPair.getPrivate()`) to create a JWS and the
public key (`keyPair.getPublic()`) to parse/verify a JWS.
> **Note**
>
> * **The `PS256`, `PS384`, and `PS512` algorithms require JDK 11 or a compatible JCA Provider
> (like BouncyCastle) in the runtime classpath.**
>
>
> * **The `EdDSA` algorithms requires JDK 15 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.**
>
> If you want to use either set of algorithms, and you are on an earlier JDK that does not support them,
> see the [Installation](#Installation) section to see how to enable BouncyCastle. All other algorithms are
> natively supported by the JDK.
### Creating a JWS
You create a JWS as follows:
1. Use the `Jwts.builder()` method to create a `JwtBuilder` instance.
2. Call `JwtBuilder` methods to set the `payload` content or claims and any header parameters as desired.
3. Specify the `SecretKey` or asymmetric `PrivateKey` you want to use to sign the JWT.
4. Finally, call the `compact()` method to compact and sign, producing the final jws.
For example:
```java
String jws = Jwts.builder() // (1)
.subject("Bob") // (2)
.signWith(key) // (3) <---
.compact(); // (4)
```
#### Signing Key
It is usually recommended to specify the signing key by calling the `JwtBuilder`'s `signWith` method and let JJWT
determine the most secure algorithm allowed for the specified key.:
```java
String jws = Jwts.builder()
// ... etc ...
.signWith(key) // <---
.compact();
```
For example, if you call `signWith` with a `SecretKey` that is 256 bits (32 bytes) long, it is not strong enough for
`HS384` or `HS512`, so JJWT will automatically sign the JWT using `HS256`.
When using `signWith` JJWT will also automatically set the required `alg` header with the associated algorithm
identifier.
Similarly, if you called `signWith` with an RSA `PrivateKey` that was 4096 bits long, JJWT will use the `RS512`
algorithm and automatically set the `alg` header to `RS512`.
The same selection logic applies for Elliptic Curve `PrivateKey`s.
> **Note**
>
> **You cannot sign JWTs with `PublicKey`s as this is always insecure.** JJWT will reject any specified
> `PublicKey` for signing with an `InvalidKeyException`.
##### SecretKey Formats
If you want to sign a JWS using HMAC-SHA algorithms, and you have a secret key `String` or
[encoded byte array](https://docs.oracle.com/javase/8/docs/api/java/security/Key.html#getEncoded--), you will need
to convert it into a `SecretKey` instance to use as the `signWith` method argument.
If your secret key is:
* An [encoded byte array](https://docs.oracle.com/javase/8/docs/api/java/security/Key.html#getEncoded--):
```java
SecretKey key = Keys.hmacShaKeyFor(encodedKeyBytes);
```
* A Base64-encoded string:
```java
SecretKey key = Keys.hmacShaKeyFor(Decoders.BASE64.decode(secretString));
```
* A Base64URL-encoded string:
```java
SecretKey key = Keys.hmacShaKeyFor(Decoders.BASE64URL.decode(secretString));
```
* A raw (non-encoded) string (e.g. a password String):
```java
Password key = Keys.password(secretString.toCharArray());
```
> **Warning**
>
> It is almost always incorrect to call any variant of `secretString.getBytes` in any cryptographic context.
> Safe cryptographic keys are never represented as direct (unencoded) strings. If you have a password that should
> be represented as a `Key` for `HMAC-SHA` algorithms, it is _strongly_ recommended to use a key derivation
> algorithm to derive a cryptographically-strong `Key` from the password, and never use the password directly.
##### SignatureAlgorithm Override
In some specific cases, you might want to override JJWT's default selected signature algorithm for a given key.
For example, if you have an RSA `PrivateKey` that is 2048 bits, JJWT would automatically choose the `RS256` algorithm.
If you wanted to use `RS384` or `RS512` instead, you could manually specify it with the overloaded `signWith` method
that accepts the `SignatureAlgorithm` as an additional argument:
```java
.signWith(privateKey, Jwts.SIG.RS512) // <---
.compact();
```
This is allowed because the JWT specification allows any RSA algorithm strength for any RSA key >= 2048 bits. JJWT just
prefers `RS512` for keys >= 4096 bits, followed by `RS384` for keys >= 3072 bits and finally `RS256` for keys >= 2048
bits.
**In all cases however, regardless of your chosen algorithms, JJWT will assert that the specified key is allowed to be
used for that algorithm when possible according to the JWT specification requirements.**
#### JWS Compression
If your JWT payload is large (contains a lot of data), and you are certain that JJWT will also be the same library
that reads/parses your JWS, you might want to compress the JWS to reduce its size.
> **Warning**
>
> **Not Standard for JWS**: JJWT supports compression for JWS, but it is not a standard feature for JWS. The
> JWT RFC specifications standardize this _only_ for JWEs, and it is not likely to be supported by other JWT libraries
> for JWS. Use JWS compression only if you are certain that JJWT (or another library that supports JWS compression)
> will be parsing the JWS
Please see the main [Compression](#compression) section to see how to compress and decompress JWTs.
### Reading a JWS
You read (parse) a JWS as follows:
1. Use the `Jwts.parser()` method to create a `JwtParserBuilder` instance.
2. Call either [keyLocator](#key-locator) or `verifyWith` methods to determine the key used to verify the JWS signature.
3. Call the `build()` method on the `JwtParserBuilder` to return a thread-safe `JwtParser`.
4. Finally, call the `parseClaimsJws(String)` method with your jws `String`, producing the original JWS.
5. The entire call is wrapped in a try/catch block in case parsing or signature validation fails. We'll cover
exceptions and causes for failure later.
For example:
```java
Jws jws;
try {
jws = Jwts.parser() // (1)
.keyLocator(keyLocator) // (2) dynamically lookup verification keys based on each JWS
//.verifyWith(key) // or a static key used to verify all encountered JWSs
.build() // (3)
.parseClaimsJws(jwsString); // (4) or parseContentJws(jwsString)
// we can safely trust the JWT
catch (JwtException ex) { // (5)
// we *cannot* use the JWT as intended by its creator
}
```
> **Note**
>
> **Type-safe JWSs:**
> * If you are expecting a JWS with a Claims `payload`, call the `JwtParser`'s `parseClaimsJws` method.
> * If you are expecting a JWS with a content `payload`, call the `JwtParser`'s `parseContentJws` method.
#### Verification Key
The most important thing to do when reading a JWS is to specify the key used to verify the JWS's
cryptographic signature. If signature verification fails, the JWT cannot be safely trusted and should be
discarded.
So which key do we use for verification?
* If the jws was signed with a `SecretKey`, the same `SecretKey` should be specified on the `JwtParserBuilder`.
For example:
```java
Jwts.parser()
.verifyWith(secretKey) // <----
.build()
.parseClaimsJws(jwsString);
```
* If the jws was signed with a `PrivateKey`, that key's corresponding `PublicKey` (not the `PrivateKey`) should be
specified on the `JwtParserBuilder`. For example:
```java
Jwts.parser()
.verifyWith(publicKey) // <---- publicKey, not privateKey
.build()
.parseClaimsJws(jwsString);
```
#### Verification Key Locator
But you might have noticed something - what if your application doesn't use just a single `SecretKey` or `KeyPair`? What
if JWSs can be created with different `SecretKey`s or public/private keys, or a combination of both? How do you
know which key to specify if you can't inspect the JWT first?
In these cases, you can't call the `JwtParserBuilder`'s `verifyWith` method with a single key - instead, you'll need a
Key Locator. Please see the [Key Lookup](#key-locator) section to see how to dynamically obtain different keys when
parsing JWSs or JWEs.
#### JWS Decompression
If you used JJWT to compress a JWS and you used a custom compression algorithm, you will need to tell the
`JwtParserBuilder` how to resolve your `CompressionAlgorithm` to decompress the JWT.
Please see the [Compression](#compression) section below to see how to decompress JWTs during parsing.
### Unencoded Payload Option
In some cases, especially if a JWS payload is large, it could be desirable to _not_ Base64URL-encode the JWS payload,
or even exclude the payload from the compact JWS string entirely. The JWT RFC specifications provide support
for these use cases via the
[JSON Web Signature (JWS) Unencoded Payload Option](https://www.rfc-editor.org/rfc/rfc7797.html) specification,
which JJWT supports.
This option comes with both benefits and disadvantages:
#### Benefits
A JWS producer can still create a JWS string to use for payload integrity verification without having to either:
1. Base64URL-encode the (potentially very large) payload, saving the time that could take.
2. Include the payload in the compact JWS string at all. Omitting the payload from the JWS compact string
entirely produces smaller JWSs that can be more efficient to transfer.
#### Disadvantages
1. Your application, and not JJWT, incurs the responsibility to ensure the payload is not modified during transmission
so the recipient can verify the JWS signature. For example, by using a sufficiently strong TLS (https) cipher
suite as well as any additional care before and after transmission, since
[TLS does not guarantee end-to-end security](https://tozny.com/blog/end-to-end-encryption-vs-https/).
2. If you choose to include the unencoded payload in the JWS compact string, your application
[MUST](https://www.rfc-editor.org/rfc/rfc7797.html#section-5.2) ensure that the payload does not contain a
period (`.`) character anywhere in the payload. The JWS recipient will experience parsing errors otherwise.
Before attempting to use this option, one should be aware of the RFC's
[security considerations](https://www.rfc-editor.org/rfc/rfc7797.html#section-8) first.
> **Note**
>
> **Protected JWS Only**
>
> The RFC specification defines the Unencoded Payload option for use only with JWSs. It may not be used with
> with unprotected JWTs or encrypted JWEs.
#### Detached Payload Example
This example shows creating and parsing a compact JWS using an unencoded payload that is detached, i.e. where the
payload is not embedded in the compact JWS string at all.
We need to do three things during creation:
1. Specify the JWS signing key; it's a JWS and still needs to be signed.
2. Specify the raw payload bytes via the `JwtBuilder`'s `content` method.
3. Indicate that the payload should _not_ be Base64Url-encoded using the `JwtBuilder`'s `encodePayload(false)` method.
```java
// create a test key for this example:
SecretKey testKey = Jwts.SIG.HS512.key().build();
String message = "Hello World. It's a Beautiful Day!";
byte[] content = message.getBytes(StandardCharsets.UTF_8);
String jws = Jwts.builder().signWith(testKey) // #1
.content(content) // #2
.encodePayload(false) // #3
.compact();
```
To parse the resulting `jws` string, we need to do two things when creating the `JwtParser`:
1. Specify the signature verification key.
3. Specify the externally-transmitted unencoded payload bytes, required for signature verification.
```java
Jws parsed = Jwts.parser().verifyWith(testKey) // 1
.build()
.parseContentJws(jws, content); // 2
assertArrayEquals(content, parsed.getPayload());
```
#### Non-Detached Payload Example
This example shows creating and parsing a compact JWS with what the RFC calls a 'non-detached' unencoded payload, i.e.
a raw string directly embedded as the payload in the compact JWS string.
We need to do three things during creation:
1. Specify the JWS signing key; it's a JWS and still needs to be signed.
2. Specify the raw payload string via the `JwtBuilder`'s `content` method. Per
[the RFC](https://www.rfc-editor.org/rfc/rfc7797.html#section-5.2), the payload string **_MUST NOT contain any
period (`.`) characters_**.
3. Indicate that the payload should _not_ be Base64Url-encoded using the `JwtBuilder`'s `encodePayload(false)` method.
```java
// create a test key for this example:
SecretKey testKey = Jwts.SIG.HS512.key().build();
String claimsString = "{\"sub\":\"joe\",\"iss\":\"me\"}";
String jws = Jwts.builder().signWith(testKey) // #1
.content(claimsString) // #2
.encodePayload(false) // #3
.compact();
```
If you were to print the `jws` string, you'd see something like this:
```
eyJhbGciOiJIUzUxMiIsImI2NCI6ZmFsc2UsImNyaXQiOlsiYjY0Il19.{"sub":"joe","iss":"me"}.wkoxYEd//...etc...
```
See how the `claimsString` is embedded directly as the center `payload` token instead of a standard Base64URL value?
This is why no period (`.`) characters can exist in the payload. If they did, any standard JWT parser would see more
than two periods total, which is required for parsing standard JWSs.
To parse the resulting `jws` string, we need to do two things when creating the `JwtParser`:
1. Specify the signature verification key.
2. Indicate that we want to support Unencoded Payload Option JWSs by enabling the `b64` `crit` header parameter.
```java
Jws parsed = Jwts.parser().verifyWith(testKey) // 1
.critical().add("b64").and() // 2
.build()
.parseClaimsJws(jws);
assert "joe".equals(parsed.getPayload().getSubject());
assert "me".equals(parsed.getPayload().getIssuer());
```
Did you notice we used the `.parseClaimsJws(String)` method instead of `.parseClaimsJws(String, byte[])`? This is
because the non-detached payload is already present and JJWT has what it needs for signature verification.
Additionally, we needed to specify the `b64` critical value: because we're not using the two-argument
`parseClaimsJws(jws, content)` method, the parser has no way of knowing if you wish to allow or support unencoded
payloads. Unencoded payloads have additional security considerations as described above, so they are disabled by
the parser by default unless you indicate you want to support them by using `critical().add("b64")`.
Finally, even if the payload contains a non-detached String, you could still use the two-argument method using the
payload String's UTF-8 bytes instead:
```java
parsed = Jwts.parser().verifyWith(testKey)
.build()
.parseClaimsJws(jws, claimsString.getBytes(StandardCharsets.UTF_8)); // <---
```
## Encrypted JWTs
The JWT specification also provides for the ability to encrypt and decrypt a JWT. Encrypting a JWT:
1. guarantees that no-one other than the intended JWT recipient can see the JWT `payload` (it is confidential), and
2. guarantees that no-one has manipulated or changed the JWT after it was created (its integrity is maintained).
These two properties - confidentiality and integrity - assure us that an encrypted JWT contains a `payload` that
no-one else can see, _nor_ has anyone changed or altered the data in transit.
Encryption and confidentiality seem somewhat obvious: if you encrypt a message, it is confidential by the notion that
random 3rd parties cannot make sense of the encrypted message. But some might be surprised to know that **_general
encryption does _not_ guarantee that someone hasn't tampered/altered an encrypted message in transit_**. Most of us
assume that if a message can be decrypted, then the message would be authentic and unchanged - after all, if you can
decrypt it, it must not have been tampered with, right? Because if it was changed, decryption would surely fail, right?
Unfortunately, this is not actually guaranteed in all cryptographic ciphers. There are certain attack vectors where
it is possible to change an encrypted payload (called 'ciphertext'), and the message recipient is still able to
successfully decrypt the (modified) payload. In these cases, the ciphertext integrity was not maintained - a
malicious 3rd party could intercept a message and change the payload content, even if they don't understand what is
inside the payload, and the message recipient could never know.
To combat this, there is a category of encryption algorithms that ensures both confidentiality _and_ integrity of the
ciphertext data. These types of algorithms are called
[Authenticated Encryption](https://en.wikipedia.org/wiki/Authenticated_encryption) algorithms.
As a result, to ensure JWTs do not suffer from this problem, the JWE RFC specifications require that any encryption
algorithm used to encrypt a JWT _MUST_ be an Authenticated Encryption algorithm. JWT users can be sufficiently
confident their encrypted JWTs maintain the properties of both confidentiality and integrity.
### JWE Encryption Algorithms
The JWT specification defines 6 standard Authenticated Encryption algorithms used to encrypt a JWT `payload`:
| Identifier | Required Key Bit Length | Encryption Algorithm |
|--------------------------------- | ----------------------- | -------------------- |
| A128CBC‑HS256
| 256 | [AES_128_CBC_HMAC_SHA_256](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.3) authenticated encryption algorithm |
| `A192CBC-HS384` | 384 | [AES_192_CBC_HMAC_SHA_384](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.4) authenticated encryption algorithm |
| `A256CBC-HS512` | 512 | [AES_256_CBC_HMAC_SHA_512](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.5) authenticated encryption algorithm |
| `A128GCM` | 128 | AES GCM using 128-bit key1 |
| `A192GCM` | 192 | AES GCM using 192-bit key1 |
| `A256GCM` | 256 | AES GCM using 256-bit key1 |
1. Requires Java 8+ or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
These are all represented as constants in the `io.jsonwebtoken.Jwts.ENC` registry singleton as
implementations of the `io.jsonwebtoken.security.AeadAlgorithm` interface.
As shown in the table above, each algorithm requires a key of sufficient length. The JWT specification
[RFC 7518, Sections 5.2.3 through 5.3](https://www.rfc-editor.org/rfc/rfc7518.html#section-5.2.3)
_requires_ (mandates) that you MUST use keys that are sufficiently strong for a chosen algorithm. This means that
JJWT - a specification-compliant library - will also enforce that you use sufficiently strong keys
for the algorithms you choose. If you provide a weak key for a given algorithm, JJWT will reject it and throw an
exception.
The reason why the JWT specification, and consequently JJWT, mandates key lengths is that the security model of a
particular algorithm can completely break down if you don't adhere to the mandatory key properties of the algorithm,
effectively having no security at all.
#### Symmetric Ciphers
You might have noticed something about the above Authenticated Encryption algorithms: they're all variants of the
AES algorithm, and AES always uses a symmetric (secret) key to perform encryption and decryption. That's kind of
strange, isn't it?
What about RSA and Elliptic Curve asymmetric key cryptography? And Diffie-Hellman key exchange? What about
password-based key derivation algorithms? Surely any of those could be desirable depending on the use case, no?
Yes, they definitely can, and the JWT specifications do support them, albeit indirectly: those other
algorithms _are_ indeed supported and used, but they aren't used to encrypt the JWT `payload` directly. They are
used to _produce_ the actual key used to encrypt the `JWT` payload.
This is all done via the JWT specification's concept of a Key Management Algorithm, covered next. After we cover that,
we'll show you how to encrypt and parse your own JWTs with the `JwtBuilder` and `JwtParserBuilder`.
### JWE Key Management Algorithms
As stated above, all standard JWA Encryption Algorithms are AES-based authenticated encryption algorithms. So what
about RSA and Elliptic Curve cryptography? And password-based key derivation, or Diffie-Hellman exchange?
All of those are supported as well, but they are not used directly for encryption. They are used to _produce_ the
key that will be used to directly encrypt the JWT `payload`.
That is, JWT encryption can be thought of as a two-step process, shown in the following pseudocode:
```groovy
Key algorithmKey = getKeyManagementAlgorithmKey(); // PublicKey, SecretKey, or Password
SecretKey contentEncryptionKey = keyManagementAlgorithm.produceEncryptionKey(algorithmKey); // 1
byte[] ciphertext = encryptionAlgorithm.encrypt(payload, contentEncryptionKey); // 2
```
Steps:
1. Use the `algorithmKey` to produce the actual key that will be used to encrypt the payload. The JWT specifications
call this result the 'Content Encryption Key'.
2. Take the resulting Content Encryption Key and use it directly with the Authenticated Encryption algorithm to
actually encrypt the JWT `payload`.
So why the indirection? Why not just use any `PublicKey`, `SecretKey` or `Password` to encrypt the `payload`
_directly_ ?
There are quite a few reasons for this.
1. Asymmetric key encryption (like RSA and Elliptic Curve) tends to be slow. Like _really_ slow. Symmetric key
cipher algorithms in contrast are _really fast_. This matters a lot in production applications that could be
handling a JWT on every HTTP request, which could be thousands per second.
2. RSA encryption (for example) can only encrypt a relatively small amount of data. A 2048-bit RSA key can only
encrypt up to a maximum of 245 bytes. A 4096-bit RSA key can only encrypt up to a maximum of 501 bytes. There are
plenty of JWTs that can exceed 245 bytes, and that would make RSA unusable.
3. Passwords usually make for very poor encryption keys - they often have poor entropy, or they themselves are
often too short to be used directly with algorithms that mandate minimum key lengths to help ensure safety.
For these reasons and more, using one secure algorithm to generate or encrypt a key used for another (very fast) secure
algorithm has been proven to be a great way to increase security through many more secure algorithms while
also still resulting in very fast and secure output. This is after all how TLS (for https encryption) works -
two parties can use more complex cryptography (like RSA or Elliptic Curve) to negotiate a small, fast encryption key.
This fast encryption key is produced during the 'TLS handshake' and is called the TLS 'session key'.
So the JWT specifications work much in the same way: one key from any number of various algorithm types can be used
to produce a final symmetric key, and that symmetric key is used to encrypt the JWT `payload`.
#### JWE Standard Key Management Algorithms
The JWT specification defines 17 standard Key Management Algorithms used to produce the JWE
Content Encryption Key (CEK):
| Identifier | Key Management Algorithm |
| --- |-------------------------------------------------------------------------------|
| `RSA1_5` | RSAES-PKCS1-v1_5 |
| `RSA-OAEP` | RSAES OAEP using default parameters |
| `RSA-OAEP-256` | RSAES OAEP using SHA-256 and MGF1 with SHA-256 |
| `A128KW` | AES Key Wrap with default initial value using 128-bit key |
| `A192KW` | AES Key Wrap with default initial value using 192-bit key |
| `A256KW` | AES Key Wrap with default initial value using 256-bit key |
| `dir` | Direct use of a shared symmetric key as the Content Encryption Key |
| `ECDH-ES` | Elliptic Curve Diffie-Hellman Ephemeral Static key agreement using Concat KDF |
| `ECDH-ES+A128KW` | ECDH-ES using Concat KDF and CEK wrapped with "A128KW" |
| `ECDH-ES+A192KW` | ECDH-ES using Concat KDF and CEK wrapped with "A192KW" |
| `ECDH-ES+A256KW` | ECDH-ES using Concat KDF and CEK wrapped with "A256KW" |
| `A128GCMKW` | Key wrapping with AES GCM using 128-bit key3 |
| `A192GCMKW` | Key wrapping with AES GCM using 192-bit key3 |
| `A256GCMKW` | Key wrapping with AES GCM using 256-bit key3 |
| `PBES2-HS256+A128KW` | PBES2 with HMAC SHA-256 and "A128KW" wrapping3 |
| `PBES2-HS384+A192KW` | PBES2 with HMAC SHA-384 and "A192KW" wrapping3 |
| PBES2‑HS512+A256KW
| PBES2 with HMAC SHA-512 and "A256KW" wrapping3 |
3. Requires Java 8 or a compatible JCA Provider (like BouncyCastle) in the runtime classpath.
These are all represented as constants in the `io.jsonwebtoken.Jwts.KEY` registry singleton as
implementations of the `io.jsonwebtoken.security.KeyAlgorithm` interface.
But 17 algorithms are a lot to choose from. When would you use them? The sections below describe when you might
choose each category of algorithms and how they behave.
##### RSA Key Encryption
The JWT RSA key management algorithms `RSA1_5`, `RSA-OAEP`, and `RSA-OAEP-256` are used when you want to use the
JWE recipient's RSA _public_ key during encryption. This ensures that only the JWE recipient can decrypt
and read the JWE (using their RSA `private` key).
During JWE creation, these algorithms:
* Generate a new secure-random Content Encryption Key (CEK) suitable for the desired [encryption algorithm](#jwe-enc).
* Encrypt the JWE payload with the desired encryption algorithm using the new CEK, producing the JWE payload ciphertext.
* Encrypt the CEK itself with the specified RSA key wrap algorithm using the JWE recipient's RSA public key.
* Embed the payload ciphertext and encrypted CEK in the resulting JWE.
During JWE decryption, these algorithms:
* Retrieve the encrypted Content Encryption Key (CEK) embedded in the JWE.
* Decrypt the encrypted CEK with the discovered RSA key unwrap algorithm using the JWE recipient's RSA private key,
producing the decrypted Content Encryption Key (CEK).
* Decrypt the JWE ciphertext payload with the JWE's identified [encryption algorithm](#jwe-enc) using the decrypted CEK.
> **Warning**
>
> RFC 7518 Sections [4.2](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.2) and
> [4.3](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.3) _require_ (mandate) that RSA keys >= 2048 bits
> MUST be used with these algorithms. JJWT will throw an exception if it detects weaker keys being used.
##### AES Key Encryption
The JWT AES key management algorithms `A128KW`, `A192KW`, `A256KW`, `A128GCMKW`, `A192GCMKW`, and `A256GCMKW` are
used when you have a symmetric secret key, but you don't want to use that secret key to directly
encrypt/decrypt the JWT.
Instead, a new secure-random key is generated each time a JWE is created, and that new/random key is used to directly
encrypt/decrypt the JWT payload. The secure-random key is itself encrypted with your symmetric secret key
using the AES Wrap algorithm, and the encrypted key is embedded in the resulting JWE.
This allows the JWE to be encrypted with a random short-lived key, reducing material exposure of the potentially
longer-lived symmetric secret key.
Because these particular algorithms use a symmetric secret key, they are best suited when the JWE creator and
receiver are the same, ensuring the secret key does not need to be shared with multiple parties.
During JWE creation, these algorithms:
* Generate a new secure-random Content Encryption Key (CEK) suitable for the desired [encryption algorithm](#jwe-enc).
* Encrypt the JWE payload with the desired encryption algorithm using the new CEK, producing the JWE payload ciphertext.
* Encrypt the CEK itself with the specified AES key algorithm (either AES Key Wrap or AES with GCM encryption),
producing the encrypted CEK.
* Embed the payload ciphertext and encrypted CEK in the resulting JWE.
During JWE decryption, these algorithms:
* Retrieve the encrypted Content Encryption Key (CEK) embedded in the JWE.
* Decrypt the encrypted CEK with the discovered AES key algorithm using the symmetric secret key.
* Decrypt the JWE ciphertext payload with the JWE's identified [encryption algorithm](#jwe-enc) using the decrypted CEK.
> **Warning**
>
> The symmetric key used for the AES key algorithms MUST be 128, 192 or 256 bits as required by the specific AES
> key algorithm. JJWT will throw an exception if it detects weaker keys than what is required.
##### Direct Key Encryption
The JWT `dir` (direct) key management algorithm is used when you have a symmetric secret key, and you want to use it
to directly encrypt the JWT payload.
Because this algorithm uses a symmetric secret key, it is best suited when the JWE creator and receiver are the
same, ensuring the secret key does not need to be shared with multiple parties.
This is the simplest key algorithm for direct encryption that does not perform any key encryption. It is essentially
a 'no op' key algorithm, allowing the shared key to be used to directly encrypt the JWT payload.
During JWE creation, this algorithm:
* Encrypts the JWE payload with the desired encryption algorithm directly using the symmetric secret key,
producing the JWE payload ciphertext.
* Embeds the payload ciphertext in the resulting JWE.
Note that because this algorithm does not produce an encrypted key value, an encrypted CEK is _not_ embedded in the
resulting JWE.
During JWE decryption, this algorithm decrypts the JWE ciphertext payload with the JWE's
identified [encryption algorithm](#jwe-enc) directly using the symmetric secret key. No encrypted CEK is used.
> **Warning**
>
> The symmetric secret key MUST be 128, 192 or 256 bits as required by the associated
> [AEAD encryption algorithm](#jwe-enc) used to encrypt the payload. JJWT will throw an exception if it detects
> weaker keys than what is required.
##### Password-Based Key Encryption
The JWT password-based key encryption algorithms `PBES2-HS256+A128KW`, `PBES2-HS384+A192KW`, and `PBES2-HS512+A256KW`
are used when you want to use a password (character array) to encrypt and decrypt a JWT.
However, because passwords are usually too weak or problematic to use directly in cryptographic contexts, these
algorithms utilize key derivation techniques with work factors (e.g. computation iterations) and secure-random salts
to produce stronger cryptographic keys suitable for cryptographic operations.
This allows the payload to be encrypted with a random short-lived cryptographically-stronger key, reducing the need to
expose the longer-lived (and potentially weaker) password.
Because these algorithms use a secret password, they are best suited when the JWE creator and receiver are the
same, ensuring the secret password does not need to be shared with multiple parties.
During JWE creation, these algorithms:
* Generate a new secure-random Content Encryption Key (CEK) suitable for the desired [encryption algorithm](#jwe-enc).
* Encrypt the JWE payload with the desired encryption algorithm using the new CEK, producing the JWE payload ciphertext.
* Derive a 'key encryption key' (KEK) with the desired "PBES2 with HMAC SHA" algorithm using the password, a suitable
number of computational iterations, and a secure-random salt value.
* Encrypt the generated CEK with the corresponding AES Key Wrap algorithm using the password-derived KEK.
* Embed the payload ciphertext and encrypted CEK in the resulting JWE.
> **Note**
>
> **Secure defaults**: When using these algorithms, if you do not specify a work factor (i.e. number of computational
> iterations), JJWT will automatically use an
> [OWASP PBKDF2 recommended](https://cheatsheetseries.owasp.org/cheatsheets/Password_Storage_Cheat_Sheet.html#pbkdf2)
> default appropriate for the specified `PBES2` algorithm.
During JWE decryption, these algorithms:
* Retrieve the encrypted Content Encryption Key (CEK) embedded in the JWE.
* Derive the 'key encryption key' (KEK) with the discovered "PBES2 with HMAC SHA" algorithm using the password and the
number of computational iterations and secure-random salt value discovered in the JWE header.
* Decrypt the encrypted CEK with the corresponding AES Key Unwrap algorithm using the password-derived KEK.
* Decrypt the JWE ciphertext payload with the JWE's identified [encryption algorithm](#jwe-enc) using the decrypted CEK.
##### Elliptic Curve Diffie-Hellman Ephemeral Static Key Agreement (ECDH-ES)
The JWT Elliptic Curve Diffie-Hellman Ephemeral Static key agreement algorithms `ECDH-ES`, `ECDH-ES+A128KW`,
`ECDH-ES+A192KW`, and `ECDH-ES+A256KW` are used when you want to use the JWE recipient's Elliptic Curve _public_ key
during encryption. This ensures that only the JWE recipient can decrypt and read the JWE (using their Elliptic Curve
_private_ key).
During JWE creation, these algorithms:
* Obtain the Content Encryption Key (CEK) used to encrypt the JWE payload as follows:
* Inspect the JWE recipient's Elliptic Curve public key and determine its Curve.
* Generate a new secure-random ephemeral Ellipic Curve public/private key pair on this same Curve.
* Add the ephemeral EC public key to the JWE
[epk header](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.1) for inclusion in the final JWE.
* Produce an ECDH shared secret with the ECDH Key Agreement algorithm using the JWE recipient's EC public key
and the ephemeral EC private key.
* Derive a symmetric secret key with the Concat Key Derivation Function
([NIST.800-56A](https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-56Ar2.pdf), Section 5.8.1) using
this ECDH shared secret and any provided
[PartyUInfo](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.2) and/or
[PartyVInfo](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.3).
* If the key algorithm is `ECDH-ES`:
* Use the Concat KDF-derived symmetric secret key directly as the Content Encryption Key (CEK). No encrypted key
is created, nor embedded in the resulting JWE.
* Otherwise, if the key algorithm is `ECDH-ES+A128KW`, `ECDH-ES+A192KW`, or `ECDH-ES+A256KW`:
* Generate a new secure-random Content Encryption Key (CEK) suitable for the desired [encryption algorithm](#jwe-enc).
* Encrypt this new CEK with the corresponding AES Key Wrap algorithm using the Concat KDF-derived secret key,
producing the encrypted CEK.
* Embed the encrypted CEK in the resulting JWE.
* Encrypt the JWE payload with the desired encryption algorithm using the obtained CEK, producing the JWE payload
ciphertext.
* Embed the payload ciphertext in the resulting JWE.
During JWE decryption, these algorithms:
* Obtain the Content Encryption Key (CEK) used to decrypt the JWE payload as follows:
* Retrieve the required ephemeral Elliptic Curve public key from the JWE's
[epk header](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.1).
* Ensure the ephemeral EC public key exists on the same curve as the JWE recipient's EC private key.
* Produce the ECDH shared secret with the ECDH Key Agreement algorithm using the JWE recipient's EC private key
and the ephemeral EC public key.
* Derive a symmetric secret key with the Concat Key Derivation Function
([NIST.800-56A](https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-56Ar2.pdf), Section 5.8.1) using
this ECDH shared secret and any
[PartyUInfo](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.2) and/or
[PartyVInfo](https://www.rfc-editor.org/rfc/rfc7518.html#section-4.6.1.3) found in the JWE header.
* If the key algorithm is `ECDH-ES`:
* Use the Concat KDF-derived secret key directly as the Content Encryption Key (CEK). No encrypted key is used.
* Otherwise, if the key algorithm is `ECDH-ES+A128KW`, `ECDH-ES+A192KW`, or `ECDH-ES+A256KW`:
* Obtain the encrypted key ciphertext embedded in the JWE.
* Decrypt the encrypted key ciphertext with the associated AES Key Unwrap algorithm using the Concat KDF-derived
secret key, producing the unencrypted Content Encryption Key (CEK).
* Decrypt the JWE payload ciphertext with the JWE's discovered encryption algorithm using the obtained CEK.
### Creating a JWE
Now that we know the difference between a JWE Encryption Algorithm and a JWE Key Management Algorithm, how do we use
them to encrypt a JWT?
You create an encrypted JWT (called a 'JWE') as follows:
1. Use the `Jwts.builder()` method to create a `JwtBuilder` instance.
2. Call `JwtBuilder` methods to set the `payload` content or claims and any [header](#jws-create-header) parameters as desired.
3. Call the `encryptWith` method, specifying the Key, Key Algorithm, and Encryption Algorithm you want to use.
4. Finally, call the `compact()` method to compact and encrypt, producing the final jwe.
For example:
```java
String jwe = Jwts.builder() // (1)
.subject("Bob") // (2)
.encryptWith(key, keyAlgorithm, encryptionAlgorithm) // (3)
.compact(); // (4)
```
Before calling `compact()`, you may set any [header](#jws-create-header) parameters and [claims](#jws-create-claims)
exactly the same way as described for JWS.
#### JWE Compression
If your JWT payload or Claims set is large (contains a lot of data), you might want to compress the JWE to reduce
its size. Please see the main [Compression](#compression) section to see how to compress and decompress JWTs.
### Reading a JWE
You read (parse) a JWE as follows:
1. Use the `Jwts.parser()` method to create a `JwtParserBuilder` instance.
2. Call either [keyLocator](#key-locator) or `decryptWith` methods to determine the key used to decrypt the JWE.
4. Call the `JwtParserBuilder`'s `build()` method to create a thread-safe `JwtParser`.
5. Parse the jwe string with the `JwtParser`'s `parseClaimsJwe` or `parseContentJwe` method.
6. Wrap the entire call is in a try/catch block in case decryption or integrity verification fails.
For example:
```java
Jwe jwe;
try {
jwe = Jwts.parser() // (1)
.keyLocator(keyLocator) // (2) dynamically lookup decryption keys based on each JWE
//.decryptWith(key) // or a static key used to decrypt all encountered JWEs
.build() // (3)
.parseClaimsJwe(jweString); // (4) or parseContentJwe(jweString);
// we can safely trust the JWT
catch (JwtException ex) { // (5)
// we *cannot* use the JWT as intended by its creator
}
```
> **Note**
>
> **Type-safe JWEs:**
> * If you are expecting a JWE with a Claims `payload`, call the `JwtParser`'s `parseClaimsJwe` method.
> * If you are expecting a JWE with a content `payload`, call the `JwtParser`'s `parseContentJwe` method.
#### Decryption Key
The most important thing to do when reading a JWE is to specify the key used during decryption. If decryption or
integrity protection checks fail, the JWT cannot be safely trusted and should be discarded.
So which key do we use for decryption?
* If the jwe was encrypted _directly_ with a `SecretKey`, the same `SecretKey` must be specified on the
`JwtParserBuilder`. For example:
```java
Jwts.parser()
.decryptWith(secretKey) // <----
.build()
.parseClaimsJwe(jweString);
```
* If the jwe was encrypted using a key produced by a Password-based key derivation `KeyAlgorithm`, the same
`Password` must be specified on the `JwtParserBuilder`. For example:
```java
Password password = Keys.password(passwordChars);
Jwts.parser()
.decryptWith(password) // <---- an `io.jsonwebtoken.security.Password` instance
.build()
.parseClaimsJwe(jweString);
```
* If the jwe was encrypted with a key produced by an asymmetric `KeyAlgorithm`, the corresponding `PrivateKey` (not
the `PublicKey`) must be specified on the `JwtParserBuilder`. For example:
```java
Jwts.parser()
.decryptWith(privateKey) // <---- a `PrivateKey`, not a `PublicKey`
.build()
.parseClaimsJws(jweString);
```
#### Decryption Key Locator
What if your application doesn't use just a single `SecretKey` or `KeyPair`? What
if JWEs can be created with different `SecretKey`s, `Password`s or public/private keys, or a combination of all of
them? How do you know which key to specify if you can't inspect the JWT first?
In these cases, you can't call the `JwtParserBuilder`'s `decryptWith` method with a single key - instead, you'll need
to use a Key `Locator`. Please see the [Key Lookup](#key-locator) section to see how to dynamically obtain different
keys when parsing JWSs or JWEs.
#### ECDH-ES Decryption with PKCS11 PrivateKeys
The JWT `ECDH-ES`, `ECDH-ES+A128KW`, `ECDH-ES+A192KW`, and `ECDH-ES+A256KW` key algorithms validate JWE input using
public key information, even when using `PrivateKey`s to decrypt. Ordinarily this is automatically performed
by JJWT when your `PrivateKey` instances implement the
[ECKey](https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/security/interfaces/ECKey.html) or
[EdECKey](https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/security/interfaces/EdECKey.html)
(or BouncyCastle equivalent) interfaces, which is the case for most JCA `Provider` implementations.
However, if your decryption `PrivateKey`s are stored in a Hardware Security Module (HSM) and/or you use the
[SunPKCS11 Provider](https://docs.oracle.com/en/java/javase/17/security/pkcs11-reference-guide1.html#GUID-6DA72F34-6C6A-4F7D-ADBA-5811576A9331),
it is likely that your `PrivateKey` instances _do not_ implement `ECKey`.
In these cases, you need to provide both the PKCS11 `PrivateKey` and it's companion `PublicKey` during decryption
by using the `Keys.builder` method. For example:
```java
KeyPair pair = getMyPkcs11KeyPair();
PrivateKey jwtParserDecryptionKey = Keys.builder(pair.getPrivate())
.publicKey(pair.getPublic()) // PublicKey must implement ECKey or EdECKey or BouncyCastle equivalent
.build();
```
You then use the resulting `jwtParserDecryptionKey` (not `pair.getPrivate()`) with the `JwtParserBuilder` or as
the return value from a custom [Key Locator](#key-locator) implementation. For example:
```java
PrivateKey decryptionKey = Keys.builder(pkcs11PrivateKey).publicKey(pkcs11PublicKey).build();
Jwts.parser()
.decryptWith(decryptionKey) // <----
.build()
.parseClaimsJwe(jweString);
```
Or as the return value from your key locator:
```java
Jwts.parser()
.keyLocator(keyLocator) // your keyLocator.locate(header) would return Keys.builder...
.build()
.parseClaimsJwe(jweString);
```
Please see the [Provider-constrained Keys](#key-locator-provider) section for more information, as well as
code examples of how to implement a Key `Locator` using the `Keys.builder` technique.
#### JWE Decompression
If a JWE is compressed using the `DEF` ([DEFLATE](https://www.rfc-editor.org/rfc/rfc1951)) or `GZIP`
([GZIP](https://www.rfc-editor.org/rfc/rfc1952.html)) compression algorithms, it will automatically be decompressed
after decryption, and there is nothing you need to configure.
If, however, a custom compression algorithm was used to compress the JWE, you will need to tell the
`JwtParserBuilder` how to resolve your `CompressionAlgorithm` to decompress the JWT.
Please see the [Compression](#compression) section below to see how to decompress JWTs during parsing.
## JSON Web Keys (JWKs)
[JSON Web Keys](https://www.rfc-editor.org/rfc/rfc7517.html) (JWKs) are JSON serializations of cryptographic keys,
allowing key material to be embedded in JWTs or transmitted between parties in a standard JSON-based text format. They
are essentially a JSON-based alternative to other text-based key formats, such as the
[DER, PEM and PKCS12](https://serverfault.com/a/9717) text strings or files commonly used when configuring TLS on web
servers, for example.
For example, an identity web service may expose its RSA or Elliptic Curve Public Keys to 3rd parties in the JWK format.
A client may then parse the public key JWKs to verify the service's [JWS](#jws) tokens, as well as send encrypted
information to the service using [JWE](#jwe)s.
JWKs can be converted to and from standard Java `Key` types as expected using the same builder/parser patterns we've
seen for JWTs.
### Create a JWK
You create a JWK as follows:
1. Use the `Jwks.builder()` method to create a `JwkBuilder` instance.
2. Call the `key` method with the Java key you wish to represent as a JWK.
3. Call builder methods to set any additional key parameters or metadata, such as a `kid` (Key ID), X509 Certificates,
etc as desired.
4. Call the `build()` method to produce the resulting JWK.
For example:
```java
SecretKey key = getSecretKey(); // or RSA or EC PublicKey or PrivateKey
SecretJwk = Jwks.builder().key(key) // (1) and (2)
.id("mySecretKeyId") // (3)
// ... etc ...
.build(); // (4)
```
#### JWK from a Map
If you have a `Map` of name/value pairs that reflect an existing JWK, you add them and build a type-safe
`Jwk` instance:
```java
Map jwkValues = getMyJwkMap();
Jwk> jwk = Jwks.builder().add(jwkValues).build();
```
### Read a JWK
You can read/parse a JWK by building a `JwkParser` and parsing the JWK JSON string with its `parse` method:
```java
String json = getJwkJsonString();
Jwk> jwk = Jwks.parser()
//.provider(aJcaProvider) // optional
//.deserializer(deserializer) // optional
.build() // create the parser
.parse(json); // actually parse the JSON
Key key = jwk.toKey(); // convert to a Java Key instance
```
As shown above you can specify a custom JCA Provider or [JSON deserializer](#json) in the same way as the `JwtBuilder`.
### PrivateKey JWKs
Unlike Java, the JWA specification requires a private JWKs to contain _both_ public key _and_ private key material
(see [RFC 7518, Section 6.1.1](https://www.rfc-editor.org/rfc/rfc7518.html#section-6.2.2) and
[RFC 7518, Section 6.3.2](https://www.rfc-editor.org/rfc/rfc7518.html#section-6.3.2)).
In this sense, a private JWK (represented as a `PrivateJwk` or a subtype, such as `RsaPrivateJwk`, `EcPrivateJwk`, etc)
can be thought of more like a Java `KeyPair` instance. Consequently, when creating a `PrivateJwk` instance,
the `PrivateKey`'s corresponding `PublicKey` is required.
#### Private JWK `PublicKey`
If you do not provide a `PublicKey` when creating a `PrivateJwk`, JJWT will automatically derive the `PublicKey` from
the `PrivateKey` instance if possible. However, because this can add
some computing time, it is typically recommended to provide the `PublicKey` when possible to avoid this extra work.
For example:
```java
RSAPrivateKey rsaPrivateKey = getRSAPrivateKey(); // or ECPrivateKey
RsaPrivateJwk jwk = Jwks.builder().key(rsaPrivateKey)
//.publicKey(rsaPublicKey) // optional, but recommended to avoid extra computation work
.build();
```
#### Private JWK from KeyPair
If you have a Java `KeyPair` instance, then you have both the public and private key material necessary to create a
`PrivateJwk`. For example:
```java
KeyPair rsaKeyPair = getRSAKeyPair();
RsaPrivateJwk rsaPrivJwk = Jwks.builder().rsaKeyPair(rsaKeyPair).build();
KeyPair ecKeyPair = getECKeyPair();
EcPrivateJwk ecPrivJwk = Jwks.builder().ecKeyPair(ecKeyPair).build();
KeyPair edEcKeyPair = getEdECKeyPair();
OctetPrivateJwk edEcPrivJwk = Jwks.builder().octetKeyPair(edEcKeyPair).build();
```
Note that:
* An exception will thrown when calling `rsaKeyPair` if the specified `KeyPair` instance does not contain
`RSAPublicKey` and `RSAPrivateKey` instances.
* Similarly, an exception will be thrown when calling `ecKeyPair` if
the `KeyPair` instance does not contain `ECPublicKey` and `ECPrivateKey` instances.
* Finally, an exception will be
thrown when calling `octetKeyPair` if the `KeyPair` instance does not contain X25519, X448, Ed25519, or Ed448 keys
(introduced in JDK 11 and 15 or when using BouncyCastle).
#### Private JWK Public Conversion
Because private JWKs contain public key material, you can always obtain the private JWK's corresponding public JWK and
Java `PublicKey` or `KeyPair`. For example:
```java
RsaPrivateJwk privateJwk = Jwks.builder().key(rsaPrivateKey).build(); // or ecPrivateKey or edEcPrivateKey
// Get the matching public JWK and/or PublicKey:
RsaPublicJwk pubJwk = privateJwk.toPublicJwk(); // JWK instance
RSAPublicKey pubKey = pubJwk.toKey(); // Java PublicKey instance
KeyPair pair = privateJwk.toKeyPair(); // io.jsonwebtoken.security.KeyPair retains key types
java.security.KeyPair jdkPair = pair.toJavaKeyPair(); // does not retain pub/private key types
```
### JWK Thumbprints
A [JWK Thumbprint](https://www.rfc-editor.org/rfc/rfc7638.html) is a digest (aka hash) of a canonical JSON
representation of a JWK's public properties. 'Canonical' in this case means that only RFC-specified values in any JWK
are used in an exact order thumbprint calculation. This ensures that anyone can calculate a JWK's same exact
thumbprint, regardless of custom parameters or JSON key/value ordering differences in a JWK.
All `Jwk` instances support [JWK Thumbprint](https://www.rfc-editor.org/rfc/rfc7638.html)s via the
`thumbprint()` and `thumbprint(HashAlgorithm)` methods:
```java
HashAlgorithm hashAlg = Jwks.HASH.SHA256; // or SHA384, SHA512, etc.
Jwk> jwk = Jwks.builder(). /* ... */ .build();
JwkThumbprint sha256Thumbprint = jwk.thumbprint(); // SHA-256 thumbprint by default
JwkThumbprint anotherThumbprint = jwk.thumbprint(Jwks.HASH.SHA512); // or a specified hash algorithm
```
The resulting `JwkThumbprint` instance provides some useful methods:
* `jwkThumbprint.toByteArray()`: the thumbprint's actual digest bytes - i.e. the raw output from the hash algorithm
* `jwkThumbprint.toString()`: the digest bytes as a Base64URL-encoded string
* `jwkThumbprint.getHashAlgorithm()`: the specific `HashAlgorithm` used to compute the thumbprint. Many standard IANA
hash algorithms are available as constants in the `Jwks.HASH` utility class.
* `jwkThumbprint.toURI()`: the thumbprint's canonical URI as defined by the [JWK Thumbprint URI](https://www.rfc-editor.org/rfc/rfc9278.html) specification
#### JWK Thumbprint as a Key ID
Because a thumbprint is an order-guaranteed unique digest of a JWK, JWK thumbprints are often used as convenient
unique identifiers for a JWK (e.g. the JWK's `kid` (Key ID) value). These identifiers can be useful when
[locating keys](#key-locator) for JWS signature verification or JWE decryption, for example.
For example:
```java
String kid = jwk.thumbprint().toString(); // Thumbprint bytes as a Base64URL-encoded string
Key key = findKey(kid);
assert jwk.toKey().equals(key);
```
However, because `Jwk` instances are immutable, you can't set the key id after the JWK is created. For example, the
following is not possible:
```java
String kid = jwk.thumbprint().toString();
jwk.setId(kid) // Jwks are immutable - there is no `setId` method
```
Instead, you may use the `idFromThumbprint` methods on the `JwkBuilder` when creating a `Jwk`:
```java
Jwk> jwk = Jwks.builder().key(aKey)
.idFromThumbprint() // or idFromThumbprint(HashAlgorithm)
.build();
```
Calling either `idFromThumbprint` method will ensure that calling `jwk.getId()` equals `thumbprint.toString()`
(which is `Encoders.BASE64URL.encode(thumbprint.toByteArray())`).
#### JWK Thumbprint URI
A JWK's thumbprint's canonical URI as defined by the [JWK Thumbprint URI](https://www.rfc-editor.org/rfc/rfc9278.html)
specification may be obtained by calling the thumbprint's `toURI()` method:
```java
URI canonicalThumbprintURI = jwk.thumbprint().toURI();
```
Per the RFC specification, if you call `canonicalThumbprintURI.toString()`, you would see a string that looks like this:
```text
urn:ietf:params:oauth:jwk-thumbprint:HASH_ALG_ID:BASE64URL_DIGEST
```
where:
* `urn:ietf:params:oauth:jwk-thumbprint:` is the URI scheme+prefix
* `HASH_ALG_ID` is the standard identifier used to compute the thumbprint as defined in the
[IANA Named Information Hash Algorithm Registry](https://www.iana.org/assignments/named-information/named-information.xhtml).
This is the same as `thumbprint.getHashAlgorithm().getId()`.
* `BASE64URL_DIGEST` is the Base64URL-encoded thumbprint bytes, equal to `jwkThumbprint.toString()`.
### JWK Security Considerations
Because they contain secret or private key material, `SecretJwk` and `PrivateJwk` (e.g. `RsaPrivateJwk`,
`EcPrivateJwk`, etc) instances should be used with great care and never accidentally transmitted to 3rd parties.
Even so, JJWT's `Jwk` implementations will suppress certain values in `toString()` output for safety as described
next.
#### JWK `toString()` Safety
Because it would be incredibly easy to accidentally print key material to `System.out.println()` or application
logs, all `Jwk` implementations will print redacted values instead of actual secret or private key material.
For example, consider the following Secret JWK JSON example from
[RFC 7515, Appendix A.1.1](https://www.rfc-editor.org/rfc/rfc7515#appendix-A.1.1):
```json
{
"kty": "oct",
"k": "AyM1SysPpbyDfgZld3umj1qzKObwVMkoqQ-EstJQLr_T-1qS0gZH75aKtMN3Yj0iPS4hcgUuTwjAzZr1Z9CAow",
"kid": "HMAC key used in https://www.rfc-editor.org/rfc/rfc7515#appendix-A.1.1 example."
}
```
The `k` value (`AyAyM1SysPpby...`) reflects secure key material and should never be accidentially
exposed.
If you were to parse this JSON as a `Jwk`, calling `toString()` will _NOT_ print this value. It will
instead print the string literal `` for any secret or private key data value. For example:
```java
String json = getExampleSecretKeyJson();
Jwk> jwk = Jwks.parser().build().parse(json);
System.out.printn(jwk);
```
This code would print the following string literal to the System console:
```text
{kty=oct, k=, kid=HMAC key used in https://www.rfc-editor.org/rfc/rfc7515#appendix-A.1.1 example.}
```
This is true for all secret or private key members in `SecretJwk` and `PrivateJwk` (e.g. `RsaPrivateJwk`,
`EcPrivateJwk`, etc) instances.
## Compression
> **Warning**
>
> **The JWT specification standardizes compression for JWEs (Encrypted JWTs) ONLY, however JJWT supports it for JWS
> (Signed JWTs) as well**.
>
> If you are positive that a JWS you create with JJWT will _also_ be parsed with JJWT,
> you can use this feature with both JWEs and JWSs, otherwise it is best to only use it for JWEs.
If a JWT's `payload` is sufficiently large - that is, it is a large content byte array or JSON with a lot of
name/value pairs (or individual values are very large or verbose) - you can reduce the size of the compact JWT by
compressing the payload.
This might be important to you if the resulting JWT is used in a URL for example, since URLs are best kept under
4096 characters due to browser, user mail agent, or HTTP gateway compatibility issues. Smaller JWTs also help reduce
bandwidth utilization, which may or may not be important depending on your application's volume or needs.
If you want to compress your JWT, you can use the `JwtBuilder`'s `compressWith(CompressionAlgorithm)` method. For
example:
```java
Jwts.builder()
.compressWith(Jwts.ZIP.DEF) // DEFLATE compression algorithm
// .. etc ...
```
If you use any of the algorithm constants in the `Jwts.ZIP` class, that's it, you're done. You don't have to
do anything during parsing or configure the `JwtParserBuilder` for compression - JJWT will automatically decompress
the payload as expected.
### Custom Compression Algorithm
If the default `Jwts.ZIP` compression algorithms are not suitable for your needs, you can create your own
`CompressionAlgorithm` implementation(s).
Just as you would with the default algorithms, you may specify that you want a JWT compressed by calling the
`JwtBuilder`'s `compressWith` method, supplying your custom implementation instance. For example:
```java
CompressionAlgorithm myAlg = new MyCompressionAlgorithm();
Jwts.builder()
.compressWith(myAlg) // <----
// .. etc ...
```
When you call `compressWith`, the JWT `payload` will be compressed with your algorithm, and the
[`zip` (Compression Algorithm)](https://www.rfc-editor.org/rfc/rfc7516.html#section-4.1.3)
header will automatically be set to the value returned by your algorithm's `algorithm.getId()` method as
required by the JWT specification.
However, the `JwtParser` needs to be aware of this custom algorithm as well, so it can use it while parsing. You do this
by modifying the `JwtParserBuilder`'s `zip()` collection. For example:
```java
CompressionAlgorithm myAlg = new MyCompressionAlgorithm();
Jwts.parser()
.zip().add(myAlg).and() // <----
// .. etc ...
```
This adds additional `CompressionAlgorithm` implementations to the parser's overall total set of supported compression
algorithms (which already includes all of the `Jwts.ZIP` algorithms by default).
The parser will then automatically check to see if the JWT `zip` header has been set to see if a compression
algorithm has been used to compress the JWT. If set, the parser will automatically look up your
`CompressionAlgorithm` by its `getId()` value, and use it to decompress the JWT.
## JSON Support
A `JwtBuilder` will serialize the `Header` and `Claims` maps (and potentially any Java objects they
contain) to JSON with a `Serializer