You may also want to read the recipes to find some use real-world use case, or read the design docs to know more technical details of rrweb.
You are recommended to install rrweb via jsdelivr's CDN service:
<link
rel="stylesheet"
href="https://cdn.jsdelivr.net/npm/rrweb@latest/dist/rrweb.min.css"
/>
<script src="https://cdn.jsdelivr.net/npm/rrweb@latest/dist/rrweb.min.js"></script>Also, you can link to a specific version number that you can update manually:
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/rrweb.min.js"></script>rrweb's code includes both the record and the replay parts. Most of the time you only need to include the record part into your targeted web Apps. This also can be done by using the CDN service:
<script src="https://cdn.jsdelivr.net/npm/rrweb@latest/dist/record/rrweb-record.min.js"></script>Besides the record/rrweb-record.min.js entry, rrweb also provides other bundles for different usage.
# Include record, replay, compression, and decompression.
rrweb-all.js
rrweb-all.min.js
# Include record and replay.
rrweb.js
rrweb.min.js
# Include the styles for replay.
rrweb.min.css
# Record
record/rrweb-record.js
record/rrweb-record.min.js
# Data compression.
record/rrweb-record-pack.js
record/rrweb-record-pack.min.js
# Replay
replay/rrweb-replay.js
replay/rrweb-replay.min.js
# Data decompression.
replay/rrweb-replay-unpack.js
replay/rrweb-replay-unpack.min.jsnpm install --save rrwebrrweb provides both commonJS and ES modules bundles, which are easy to use with the popular bundlers.
rrweb does not support IE11 and below because it uses the MutationObserver API which was supported by these browsers.
If you only included the record code with <script>, then you must use the global variable rrwebRecord instead of rrweb.record.
The following sample code will use the variable rrweb which is the default exporter of this library.
rrweb.record({
emit(event) {
// store the event in any way you like
},
});During recording, the recorder will emit when there is some event incurred, all you need to do is to store the emitted events in any way you like.
The record method returns a function which can be called to stop events from firing:
let stopFn = rrweb.record({
emit(event) {
if (events.length > 100) {
// stop after 100 events
stopFn();
}
},
});A more real-world usage may look like this:
let events = [];
rrweb.record({
emit(event) {
// push event into the events array
events.push(event);
},
});
// this function will send events to the backend and reset the events array
function save() {
const body = JSON.stringify({ events });
events = [];
fetch('http://YOUR_BACKEND_API', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body,
});
}
// save events every 10 seconds
setInterval(save, 10 * 1000);The parameter of rrweb.record accepts the following options.
| key | default | description |
|---|---|---|
| emit | required | the callback function to get emitted events |
| checkoutEveryNth | - | take a full snapshot after every N events refer to the checkout chapter |
| checkoutEveryNms | - | take a full snapshot after every N ms refer to the checkout chapter |
| blockClass | 'rr-block' | Use a string or RegExp to configure which elements should be blocked, refer to the privacy chapter |
| blockSelector | null | Use a string to configure which selector should be blocked, refer to the privacy chapter |
| ignoreClass | 'rr-ignore' | Use a string or RegExp to configure which elements should be ignored, refer to the privacy chapter |
| ignoreCSSAttributes | null | array of CSS attributes that should be ignored |
| maskTextClass | 'rr-mask' | Use a string or RegExp to configure which elements should be masked, refer to the privacy chapter |
| maskTextSelector | null | Use a string to configure which selector should be masked, refer to the privacy chapter |
| maskAllInputs | false | mask all input content as * |
| maskInputOptions | { password: true } | mask some kinds of input * refer to the list |
| maskInputFn | - | customize mask input content recording logic |
| maskTextFn | - | customize mask text content recording logic |
| slimDOMOptions | {} | remove unnecessary parts of the DOM refer to the list |
| dataURLOptions | {} | Canvas image format and quality ,This parameter will be passed to the OffscreenCanvas.convertToBlob(),Using this parameter effectively reduces the size of the recorded data |
| inlineStylesheet | true | whether to inline the stylesheet in the events |
| hooks | {} | hooks for events refer to the list |
| packFn | - | refer to the storage optimization recipe |
| sampling | - | refer to the storage optimization recipe |
| recordCanvas | false | Whether to record the canvas element. Available options:false, true |
| recordCrossOriginIframes | false | Whether to record cross origin iframes. rrweb has to be injected in each child iframe for this to work. Available options:false, true |
| recordAfter | 'load' | If the document is not ready, then the recorder will start recording after the specified event is fired. Available options: DOMContentLoaded, load |
| inlineImages | false | whether to record the image content |
| collectFonts | false | whether to collect fonts in the website |
| userTriggeredOnInput | false | whether to add userTriggered on input events that indicates if this event was triggered directly by the user or not. What is userTriggered? |
| plugins | [] | load plugins to provide extended record functions. What is plugins? |
You may find some contents on the webpage which are not willing to be recorded, then you can use the following approaches:
- An element with the class name
.rr-blockwill not be recorded. Instead, it will replay as a placeholder with the same dimension. - An element with the class name
.rr-ignorewill not record its input events. - All text of elements with the class name
.rr-maskand their children will be masked. input[type="password"]will be masked by default.- Mask options to mask the content in input elements.
By default, all the emitted events are required to replay a session and if you do not want to store all the events, you can use the checkout config.
Most of the time you do not need to configure this. But if you want to do something like capturing just the last N events from when an error has occurred, here is an example:
// We use a two-dimensional array to store multiple events array
const eventsMatrix = [[]];
rrweb.record({
emit(event, isCheckout) {
// isCheckout is a flag to tell you the events has been checkout
if (isCheckout) {
eventsMatrix.push([]);
}
const lastEvents = eventsMatrix[eventsMatrix.length - 1];
lastEvents.push(event);
},
checkoutEveryNth: 200, // checkout every 200 events
});
// send last two events array to the backend
window.onerror = function () {
const len = eventsMatrix.length;
const events = eventsMatrix[len - 2].concat(eventsMatrix[len - 1]);
const body = JSON.stringify({ events });
fetch('http://YOUR_BACKEND_API', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body,
});
};Due to the incremental-snapshot-chain mechanism rrweb used, we can not capture the last N events accurately. With the sample code above, you will finally get the last 200 to 400 events been sent to your backend.
Similarly, you can also configure checkoutEveryNms to capture the last N minutes events:
// We use a two-dimensional array to store multiple events array
const eventsMatrix = [[]];
rrweb.record({
emit(event, isCheckout) {
// isCheckout is a flag to tell you the events has been checkout
if (isCheckout) {
eventsMatrix.push([]);
}
const lastEvents = eventsMatrix[eventsMatrix.length - 1];
lastEvents.push(event);
},
checkoutEveryNms: 5 * 60 * 1000, // checkout every 5 minutes
});
// send last two events array to the backend
window.onerror = function () {
const len = eventsMatrix.length;
const events = eventsMatrix[len - 2].concat(eventsMatrix[len - 1]);
const body = JSON.stringify({ events });
fetch('http://YOUR_BACKEND_API', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body,
});
};With the sample code above, you will finally get the last 5 to 10 minutes of events been sent to your backend.
You need to include the style sheet before replay:
<link
rel="stylesheet"
href="https://cdn.jsdelivr.net/npm/rrweb@latest/dist/rrweb.min.css"
/>And then initialize the replayer with the following code:
const events = YOUR_EVENTS;
const replayer = new rrweb.Replayer(events);
replayer.play();const replayer = new rrweb.Replayer(events);
// play
replayer.play();
// play from the third seconds
replayer.play(3000);
// pause
replayer.pause();
// pause at the fifth seconds
replayer.pause(5000);
// destroy the replayer (hint: this operation is irreversible)
replayer.destroy();The replayer accepts options as its constructor's second parameter, and it has the following options:
| key | default | description |
|---|---|---|
| speed | 1 | replay speed ratio |
| root | document.body | the root element of replayer |
| loadTimeout | 0 | timeout of loading remote style sheet |
| skipInactive | false | whether to skip inactive time |
| showWarning | true | whether to print warning messages during replay |
| showDebug | false | whether to print debug messages during replay |
| blockClass | 'rr-block' | element with the class name will display as a blocked area |
| liveMode | false | whether to enable live mode |
| insertStyleRules | [] | accepts multiple CSS rule string, which will be injected into the replay iframe |
| triggerFocus | true | whether to trigger focus during replay |
| UNSAFE_replayCanvas | false | whether to replay the canvas element. Enable this will remove the sandbox, which is unsafe. |
| pauseAnimation | true | whether to pause CSS animation when the replayer is paused |
| mouseTail | true | whether to show mouse tail during replay. Set to false to disable mouse tail. A complete config can be found in this type |
| unpackFn | - | refer to the storage optimization recipe |
| logConfig | - | configuration of console output playback, refer to the console recipe |
| plugins | [] | load plugins to provide extended replay functions. What is plugins? |
| useVirtualDom | true | whether to use Virtual Dom optimization in the process of skipping to a new point of time |
| logger | console | The logger object used by the replayer to print warnings or errors |
Since rrweb's replayer only provides a basic UI, you can choose rrweb-replayer which is based on rrweb's public APIs but has a feature-rich replayer UI.
rrweb-player can also be included with <script>：
<link
rel="stylesheet"
href="https://cdn.jsdelivr.net/npm/rrweb-player@latest/dist/style.css"
/>
<script src="https://cdn.jsdelivr.net/npm/rrweb-player@latest/dist/index.js"></script>Or installed by using NPM：
npm install --save rrweb-playerimport rrwebPlayer from 'rrweb-player';
import 'rrweb-player/dist/style.css';new rrwebPlayer({
target: document.body, // customizable root element
props: {
events,
},
});| key | default | description |
|---|---|---|
| events | [] | the events for replaying |
| width | 1024 | the width of the replayer |
| height | 576 | the height of the replayer |
| maxScale | 1 | the maximum scale of the replayer (1 = 100%), set to 0 for unlimited |
| autoPlay | true | whether to autoplay |
| speedOption | [1, 2, 4, 8] | speed options in UI |
| showController | true | whether to show the controller UI |
| tags | {} | customize the custom events style with a key-value map |
| ... | - | all the rrweb Replayer options will be bypassed |
Developers may want to extend the rrweb's replayer or respond to its events. Such as giving notification when the replayer starts to skip inactive time.
So rrweb expose a public API on which allow developers to listen to the events and customize the reactions, and it has the following events:
const replayer = new rrweb.Replayer(events);
replayer.on(EVENT_NAME, (payload) => {
...
})The event list:
| Event | Description | Value |
|---|---|---|
| start | started to replay | - |
| pause | paused the replay | - |
| finish | finished the replay | - |
| resize | the viewport has changed | { width, height } |
| fullsnapshot-rebuilded | rebuilded a full snapshot | event |
| load-stylesheet-start | started to load remote stylesheets | - |
| load-stylesheet-end | loaded remote stylesheets | - |
| skip-start | started to skip inactive time | { speed } |
| skip-end | skipped inactive time | { speed } |
| mouse-interaction | mouse interaction has been replayed | { type, target } |
| event-cast | event has been replayed | event |
| custom-event | custom event has been replayed | event |
| destroy | destroyed the replayer | - |
The rrweb-replayer also re-expose the event listener via a component.addEventListener API.
And there are three rrweb-replayer event will be emitted in the same way:
| Event | Description | Value |
|---|---|---|
| ui-update-current-time | current time has changed | { payload } |
| ui-update-player-state | current player state has changed | { payload } |
| ui-update-progress | current progress has changed | { payload } |
You can also play with rrweb by using the REPL testing tool which does not need installation.
Run yarn repl to launch a browser and ask for a URL you want to test on the CLI:
Enter the url you want to record, e.g https://react-redux.realworld.io:
Waiting for the browser to open the specified page and print the following messages on the CLI:
Enter the url you want to record, e.g https://react-redux.realworld.io: https://github.com
Going to open https://github.com...
Ready to record. You can do any interaction on the page.
Once you want to finish the recording, enter 'y' to start replay:
At this point, you can interact on the web page. After the desired operations have been recorded, enter 'y' on the CLI, and the test tool will replay the operations to verify whether the recording was successful.
The following messages will be printed on the CLI during replay:
Enter 'y' to persistently store these recorded events:
At this point, you can enter 'y' again on the CLI. The test tool will save the recorded session into a static HTML file and prompt for the location:
Saved at PATH_TO_YOUR_REPO/temp/replay_2018_11_23T07_53_30.html
This file uses the latest rrweb bundle code, so we can run npm run bundle:browser after patching the code, then refresh the static file to see and debug the impact of the latest code on replay.