Skip to content

mathieucarbou/ESPAsyncWebServer

BranchesTags
 
 

Repository files navigation

ESPAsyncWebServer

Latest Release PlatformIO Registry

License: LGPL 3.0 Contributor Covenant

Build GitHub latest commit Gitpod Ready-to-Code

Asynchronous HTTP and WebSocket Server Library for ESP32, ESP8266 and RP2040 Supports: WebSocket, SSE, Authentication, Arduino Json 7, File Upload, Static File serving, URL Rewrite, URL Redirect, etc.

This fork is based on yubox-node-org/ESPAsyncWebServer and includes all the concurrency fixes.

Changes in this fork

  • (bug) A lot of bug fixes
  • (ci) Better CI with a complete matrix of Arduino versions and boards
  • (ci) Deployed in PlatformIO registry and Arduino IDE library manager
  • (feat) Arduino 3 / ESP-IDF 5 compatibility
  • (feat) ArduinoJson 7 compatibility
  • (feat) ESP32 / ESP8266 / RP2040 support
  • (feat) MessagePack support
  • (feat) Middleware support with pre-built middlewares for authentication, authorization, rate limiting, logging, cors, etc.
  • (feat) Request attributes to store data on the request object
  • (feat) Response header control and override
  • (feat) Response override: support the ability to replace a previously sent response by another one
  • (feat) Resumable download support using HEAD and bytes range
  • (feat) StreamConcat example to show how to stream multiple files in one response
  • (feat) Removed ESPIDF Editor (this is not the role of a web server library to do that - get the source files from the original repos if required)
  • (perf) AsyncTCPSock support: AsyncTCP can be ignored and AsyncTCPSock used instead
  • (perf) char* overloads to avoid using String
  • (perf) DEFAULT_MAX_WS_CLIENTS to change the number of allows WebSocket clients and use cleanupClients() to help cleanup resources about dead clients
  • (perf) setCloseClientOnQueueFull(bool) which can be set on a client to either close the connection or discard messages but not close the connection when the queue is full
  • (perf) SSE_MAX_QUEUED_MESSAGES to control the maximum number of messages that can be queued for a SSE client
  • (perf) WS_MAX_QUEUED_MESSAGES: control the maximum number of messages that can be queued for a Websocket client
  • (perf) Code size improvements
  • (perf) Lot of code cleanup and optimizations
  • (perf) Performance improvements in terms of memory, speed and size

WARNING: Important notes about future version 4.x

This ESPAsyncWebServer fork is now at version 3.x.

Next version 4.x will:

  1. Drop support for ESP8266, which goes EOL in a few years. All ESP8266 boards can be replaced by equivalent ESP32 boards.
  2. Drop support for Arduino 2.x and ESP-IDF 4.x. The library will be compatible with Arduino 3.x and ESP-IDF 5.x.
  3. Drop support for ArduinoJson 5.x and 6.x. The library will be compatible with ArduinoJson 7.x.

So if you need one of these feature, you will have to stick with 3.x or another fork.

Dependencies

WARNING The library name was changed from ESP Async WebServer to ESPAsyncWebServer as per the Arduino Lint recommendations, but its name had to stay ESP Async WebServer in Arduino Registry.

PlatformIO / pioarduino:

lib_compat_mode = strict
lib_ldf_mode = chain
lib_deps = mathieucarbou/ESPAsyncWebServer @ 3.3.23

Dependencies:

AsyncTCPSock

AsyncTCPSock can be used instead of AsyncTCP by excluding AsyncTCP from the library dependencies and adding AsyncTCPSock instead:

lib_compat_mode = strict
lib_ldf_mode = chain
lib_deps =
  ; mathieucarbou/AsyncTCP @ 3.2.14
  https://github.com/mathieucarbou/AsyncTCPSock/archive/refs/tags/v1.0.3-dev.zip
  mathieucarbou/ESPAsyncWebServer @ 3.3.23
lib_ignore =
  AsyncTCP
  mathieucarbou/AsyncTCP

Performance

Performance of mathieucarbou/ESPAsyncWebServer @ 3.3.23:

> brew install autocannon
> autocannon -c 10 -w 10 -d 20 http://192.168.4.1

With mathieucarbou/AsyncTCP @ 3.2.14

With https://github.com/mathieucarbou/AsyncTCPSock/archive/refs/tags/v1.0.3-dev.zip:

SSE performance

In the example, there is an endpoint /events with some comments showing how these metrics are calculated.

Test is running for 20 seconds with 10 connections.

// With AsyncTCP, with 10 workers: no message discarded from the queue
//
// Total: 1875 events, 468.75000000000000000000 events / second
// Total: 1870 events, 467.50000000000000000000 events / second
// Total: 1871 events, 467.75000000000000000000 events / second
// Total: 1875 events, 468.75000000000000000000 events / second
// Total: 1871 events, 467.75000000000000000000 events / second
// Total: 1805 events, 451.25000000000000000000 events / second
// Total: 1803 events, 450.75000000000000000000 events / second
// Total: 1873 events, 468.25000000000000000000 events / second
// Total: 1872 events, 468.00000000000000000000 events / second
// Total: 1805 events, 451.25000000000000000000 events / second
//
// With AsyncTCPSock, with 10 workers: no message discarded from the queue
//
// Total: 1242 events, 310.50000000000000000000 events / second
// Total: 1242 events, 310.50000000000000000000 events / second
// Total: 1242 events, 310.50000000000000000000 events / second
// Total: 1242 events, 310.50000000000000000000 events / second
// Total: 1181 events, 295.25000000000000000000 events / second
// Total: 1182 events, 295.50000000000000000000 events / second
// Total: 1240 events, 310.00000000000000000000 events / second
// Total: 1181 events, 295.25000000000000000000 events / second
// Total: 1181 events, 295.25000000000000000000 events / second
// Total: 1183 events, 295.75000000000000000000 events / second

Important recommendations

Most of the crashes are caused by improper configuration of the library for the project. Here are some recommendations to avoid them.

  1. Set the running core to be on the same core of your application (usually core 1) -D CONFIG_ASYNC_TCP_RUNNING_CORE=1
  2. Set the stack size appropriately with -D CONFIG_ASYNC_TCP_STACK_SIZE=16384. The default value of 16384 might be too much for your project. You can look at the MycilaTaskMonitor project to monitor the stack usage.
  3. You can change if you know what you are doing the task priority with -D CONFIG_ASYNC_TCP_PRIORITY=10. Default is 10.
  4. You can increase the queue size with -D CONFIG_ASYNC_TCP_QUEUE_SIZE=128. Default is 64.
  5. You can decrease the maximum ack time -D CONFIG_ASYNC_TCP_MAX_ACK_TIME=3000. Default is 5000.

I personally use the following configuration in my projects because my WS messages can be big (up to 4k). If you have smaller messages, you can increase WS_MAX_QUEUED_MESSAGES to 128.

  -D CONFIG_ASYNC_TCP_MAX_ACK_TIME=3000
  -D CONFIG_ASYNC_TCP_PRIORITY=10
  -D CONFIG_ASYNC_TCP_QUEUE_SIZE=128
  -D CONFIG_ASYNC_TCP_RUNNING_CORE=1
  -D CONFIG_ASYNC_TCP_STACK_SIZE=4096
  -D WS_MAX_QUEUED_MESSAGES=64

AsyncWebSocketMessageBuffer and makeBuffer()

The fork from yubox-node-org introduces some breaking API changes compared to the original library, especially regarding the use of std::shared_ptr<std::vector<uint8_t>> for WebSocket.

This fork is compatible with the original library from me-no-dev regarding WebSocket, and wraps the optimizations done by yubox-node-org in the AsyncWebSocketMessageBuffer class. So you have the choice of which API to use.

Here are examples for serializing a Json document in a websocket message buffer:

void send(JsonDocument& doc) {
  const size_t len = measureJson(doc);

  // original API from me-no-dev
  AsyncWebSocketMessageBuffer* buffer = _ws->makeBuffer(len);
  assert(buffer); // up to you to keep or remove this
  serializeJson(doc, buffer->get(), len);
  _ws->textAll(buffer);
}
void send(JsonDocument& doc) {
  const size_t len = measureJson(doc);

  // this fork (originally from yubox-node-org), uses another API with shared pointer
  auto buffer = std::make_shared<std::vector<uint8_t>>(len);
  assert(buffer); // up to you to keep or remove this
  serializeJson(doc, buffer->data(), len);
  _ws->textAll(std::move(buffer));
}

I recommend to use the official API AsyncWebSocketMessageBuffer to retain further compatibility.

How to replace a response

  // It is possible to replace a response.
  // The previous one will be deleted.
  // Response sending happens when the handler returns.
  server.on("/replace", HTTP_GET, [](AsyncWebServerRequest* request) {
    request->send(200, "text/plain", "Hello, world");
    // oups! finally we want to send a different response
    request->send(400, "text/plain", "validation error");
  });

This will send error 400 instead of 200.

How to use Middleware

Middleware is a way to intercept requests to perform some operations on them, like authentication, authorization, logging, etc and also act on the response headers.

Middleware can either be attached to individual handlers, attached at the server level (thus applied to all handlers), or both. They will be executed in the order they are attached, and they can stop the request processing by sending a response themselves.

You can have a look at the SimpleServer.ino example for some use cases.

For example, such middleware would handle authentication and set some attributes on the request to make them available for the next middleware and for the handler which will process the request.

AsyncMiddlewareFunction complexAuth([](AsyncWebServerRequest* request, ArMiddlewareNext next) {
  if (!request->authenticate("user", "password")) {
    return request->requestAuthentication();
  }

  request->setAttribute("user", "Mathieu");
  request->setAttribute("role", "staff");

  next(); // continue processing

  // you can act one the response object
  request->getResponse()->addHeader("X-Rate-Limit", "200");
});

Here are the list of available middlewares:

  • AsyncMiddlewareFunction: can convert a lambda function (ArMiddlewareCallback) to a middleware
  • AuthenticationMiddleware: to handle basic/digest authentication globally or per handler
  • AuthorizationMiddleware: to handle authorization globally or per handler
  • CorsMiddleware: to handle CORS preflight request globally or per handler
  • HeaderFilterMiddleware: to filter out headers from the request
  • HeaderFreeMiddleware: to only keep some headers from the request, and remove the others
  • LoggerMiddleware: to log requests globally or per handler with the same pattern as curl. Will also record request processing time
  • RateLimitMiddleware: to limit the number of requests on a windows of time globally or per handler

How to use authentication with AuthenticationMiddleware

Do not use the setUsername() and setPassword() methods on the hanlders anymore. They are deprecated. These methods were causing a copy of the username and password for each handler, which is not efficient.

Now, you can use the AuthenticationMiddleware to handle authentication globally or per handler.

AuthenticationMiddleware authMiddleware;

// [...]

authMiddleware.setAuthType(AsyncAuthType::AUTH_DIGEST);
authMiddleware.setRealm("My app name");
authMiddleware.setUsername("admin");
authMiddleware.setPassword("admin");
authMiddleware.setAuthFailureMessage("Authentication failed");
authMiddleware.generateHash(); // optimization to avoid generating the hash at each request

// [...]

server.addMiddleware(&authMiddleware); // globally add authentication to the server

// [...]

myHandler.addMiddleware(&authMiddleware); // add authentication to a specific handler

Migration to Middleware to improve performance and memory usage

  • AsyncEventSource.authorizeConnect(...) => do not use this method anymore: add a common AuthorizationMiddleware to the handler or server, and make sure to add it AFTER the AuthenticationMiddleware if you use authentication.
  • AsyncWebHandler.setAuthentication(...) => do not use this method anymore: add a common AuthenticationMiddleware to the handler or server
  • ArUploadHandlerFunction and ArBodyHandlerFunction => these callbacks receiving body data and upload and not calling anymore the authentication code for performance reasons. These callbacks can be called multiple times during request parsing, so this is up to the user to now call the AuthenticationMiddleware.allowed(request) if needed and ideally when the method is called for the first time. These callbacks are also not triggering the whole middleware chain since they are not part of the request processing workflow (they are not the final handler).

Original Documentation

Why should you care

  • Using asynchronous network means that you can handle more than one connection at the same time
  • You are called once the request is ready and parsed
  • When you send the response, you are immediately ready to handle other connections while the server is taking care of sending the response in the background
  • Speed is OMG
  • Easy to use API, HTTP Basic and Digest MD5 Authentication (default), ChunkedResponse
  • Easily extendible to handle any type of content
  • Supports Continue 100
  • Async WebSocket plugin offering different locations without extra servers or ports
  • Async EventSource (Server-Sent Events) plugin to send events to the browser
  • URL Rewrite plugin for conditional and permanent url rewrites
  • ServeStatic plugin that supports cache, Last-Modified, default index and more
  • Simple template processing engine to handle templates

Important things to remember

  • This is fully asynchronous server and as such does not run on the loop thread.
  • You can not use yield or delay or any function that uses them inside the callbacks
  • The server is smart enough to know when to close the connection and free resources
  • You can not send more than one response to a single request

Principles of operation

The Async Web server

  • Listens for connections
  • Wraps the new clients into Request
  • Keeps track of clients and cleans memory
  • Manages Rewrites and apply them on the request url
  • Manages Handlers and attaches them to Requests

Request Life Cycle

  • TCP connection is received by the server
  • The connection is wrapped inside Request object
  • When the request head is received (type, url, get params, http version and host), the server goes through all Rewrites (in the order they were added) to rewrite the url and inject query parameters, next, it goes through all attached Handlers(in the order they were added) trying to find one that canHandle the given request. If none are found, the default(catch-all) handler is attached.
  • The rest of the request is received, calling the handleUpload or handleBody methods of the Handler if they are needed (POST+File/Body)
  • When the whole request is parsed, the result is given to the handleRequest method of the Handler and is ready to be responded to
  • In the handleRequest method, to the Request is attached a Response object (see below) that will serve the response data back to the client
  • When the Response is sent, the client is closed and freed from the memory

Rewrites and how do they work

  • The Rewrites are used to rewrite the request url and/or inject get parameters for a specific request url path.
  • All Rewrites are evaluated on the request in the order they have been added to the server.
  • The Rewrite will change the request url only if the request url (excluding get parameters) is fully match the rewrite url, and when the optional Filter callback return true.
  • Setting a Filter to the Rewrite enables to control when to apply the rewrite, decision can be based on request url, http version, request host/port/target host, get parameters or the request client's localIP or remoteIP.
  • Two filter callbacks are provided: ON_AP_FILTER to execute the rewrite when request is made to the AP interface, ON_STA_FILTER to execute the rewrite when request is made to the STA interface.
  • The Rewrite can specify a target url with optional get parameters, e.g. /to-url?with=params

Handlers and how do they work

  • The Handlers are used for executing specific actions to particular requests
  • One Handler instance can be attached to any request and lives together with the server
  • Setting a Filter to the Handler enables to control when to apply the handler, decision can be based on request url, http version, request host/port/target host, get parameters or the request client's localIP or remoteIP.
  • Two filter callbacks are provided: ON_AP_FILTER to execute the rewrite when request is made to the AP interface, ON_STA_FILTER to execute the rewrite when request is made to the STA interface.
  • The canHandle method is used for handler specific control on whether the requests can be handled and for declaring any interesting headers that the Request should parse. Decision can be based on request method, request url, http version, request host/port/target host and get parameters
  • Once a Handler is attached to given Request (canHandle returned true) that Handler takes care to receive any file/data upload and attach a Response once the Request has been fully parsed
  • Handlers are evaluated in the order they are attached to the server. The canHandle is called only if the Filter that was set to the Handler return true.
  • The first Handler that can handle the request is selected, not further Filter and canHandle are called.

Responses and how do they work

  • The Response objects are used to send the response data back to the client
  • The Response object lives with the Request and is freed on end or disconnect
  • Different techniques are used depending on the response type to send the data in packets returning back almost immediately and sending the next packet when this one is received. Any time in between is spent to run the user loop and handle other network packets
  • Responding asynchronously is probably the most difficult thing for most to understand
  • Many different options exist for the user to make responding a background task

Template processing

  • ESPAsyncWebserver contains simple template processing engine.
  • Template processing can be added to most response types.
  • Currently it supports only replacing template placeholders with actual values. No conditional processing, cycles, etc.
  • Placeholders are delimited with % symbols. Like this: %TEMPLATE_PLACEHOLDER%.
  • It works by extracting placeholder name from response text and passing it to user provided function which should return actual value to be used instead of placeholder.
  • Since it's user provided function, it is possible for library users to implement conditional processing and cycles themselves.
  • Since it's impossible to know the actual response size after template processing step in advance (and, therefore, to include it in response headers), the response becomes chunked.

Libraries and projects that use AsyncWebServer

  • WebSocketToSerial - Debug serial devices through the web browser
  • Sattrack - Track the ISS with ESP8266
  • ESP Radio - Icecast radio based on ESP8266 and VS1053
  • VZero - the Wireless zero-config controller for volkszaehler.org
  • ESPurna - ESPurna ("spark" in Catalan) is a custom C firmware for ESP8266 based smart switches. It was originally developed with the ITead Sonoff in mind.
  • fauxmoESP - Belkin WeMo emulator library for ESP8266.
  • ESP-RFID - MFRC522 RFID Access Control Management project for ESP8266.

Request Variables

Common Variables

request->version();       // uint8_t: 0 = HTTP/1.0, 1 = HTTP/1.1
request->method();        // enum:    HTTP_GET, HTTP_POST, HTTP_DELETE, HTTP_PUT, HTTP_PATCH, HTTP_HEAD, HTTP_OPTIONS
request->url();           // String:  URL of the request (not including host, port or GET parameters)
request->host();          // String:  The requested host (can be used for virtual hosting)
request->contentType();   // String:  ContentType of the request (not avaiable in Handler::canHandle)
request->contentLength(); // size_t:  ContentLength of the request (not avaiable in Handler::canHandle)
request->multipart();     // bool:    True if the request has content type "multipart"

Headers

//List all collected headers
int headers = request->headers();
int i;
for(i=0;i<headers;i++){
  AsyncWebHeader* h = request->getHeader(i);
  Serial.printf("HEADER[%s]: %s\n", h->name().c_str(), h->value().c_str());
}

//get specific header by name
if(request->hasHeader("MyHeader")){
  AsyncWebHeader* h = request->getHeader("MyHeader");
  Serial.printf("MyHeader: %s\n", h->value().c_str());
}

//List all collected headers (Compatibility)
int headers = request->headers();
int i;
for(i=0;i<headers;i++){
  Serial.printf("HEADER[%s]: %s\n", request->headerName(i).c_str(), request->header(i).c_str());
}

//get specific header by name (Compatibility)
if(request->hasHeader("MyHeader")){
  Serial.printf("MyHeader: %s\n", request->header("MyHeader").c_str());
}

GET, POST and FILE parameters

//List all parameters
int params = request->params();
for(int i=0;i<params;i++){
  AsyncWebParameter* p = request->getParam(i);
  if(p->isFile()){ //p->isPost() is also true
    Serial.printf("FILE[%s]: %s, size: %u\n", p->name().c_str(), p->value().c_str(), p->size());
  } else if(p->isPost()){
    Serial.printf("POST[%s]: %s\n", p->name().c_str(), p->value().c_str());
  } else {
    Serial.printf("GET[%s]: %s\n", p->name().c_str(), p->value().c_str());
  }
}

//Check if GET parameter exists
if(request->hasParam("download"))
  AsyncWebParameter* p = request->getParam("download");

//Check if POST (but not File) parameter exists
if(request->hasParam("download", true))
  AsyncWebParameter* p = request->getParam("download", true);

//Check if FILE was uploaded
if(request->hasParam("download", true, true))
  AsyncWebParameter* p = request->getParam("download", true, true);

//List all parameters (Compatibility)
int args = request->args();
for(int i=0;i<args;i++){
  Serial.printf("ARG[%s]: %s\n", request->argName(i).c_str(), request->arg(i).c_str());
}

//Check if parameter exists (Compatibility)
if(request->hasArg("download"))
  String arg = request->arg("download");

FILE Upload handling

void handleUpload(AsyncWebServerRequest *request, String filename, size_t index, uint8_t *data, size_t len, bool final){
  if(!index){
    Serial.printf("UploadStart: %s\n", filename.c_str());
  }
  for(size_t i=0; i<len; i++){
    Serial.write(data[i]);
  }
  if(final){
    Serial.printf("UploadEnd: %s, %u B\n", filename.c_str(), index+len);
  }
}

Body data handling

void handleBody(AsyncWebServerRequest *request, uint8_t *data, size_t len, size_t index, size_t total){
  if(!index){
    Serial.printf("BodyStart: %u B\n", total);
  }
  for(size_t i=0; i<len; i++){
    Serial.write(data[i]);
  }
  if(index + len == total){
    Serial.printf("BodyEnd: %u B\n", total);
  }
}

If needed, the _tempObject field on the request can be used to store a pointer to temporary data (e.g. from the body) associated with the request. If assigned, the pointer will automatically be freed along with the request.

JSON body handling with ArduinoJson

Endpoints which consume JSON can use a special handler to get ready to use JSON data in the request callback:

#include "AsyncJson.h"
#include "ArduinoJson.h"

AsyncCallbackJsonWebHandler* handler = new AsyncCallbackJsonWebHandler("/rest/endpoint", [](AsyncWebServerRequest *request, JsonVariant &json) {
  JsonObject jsonObj = json.as<JsonObject>();
  // ...
});
server.addHandler(handler);

Responses

Redirect to another URL

//to local url
request->redirect("/login");

//to external url
request->redirect("http://esp8266.com");

Basic response with HTTP Code

request->send(404); //Sends 404 File Not Found

Basic response with HTTP Code and extra headers

AsyncWebServerResponse *response = request->beginResponse(404); //Sends 404 File Not Found
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Basic response with string content

request->send(200, "text/plain", "Hello World!");

Basic response with string content and extra headers

AsyncWebServerResponse *response = request->beginResponse(200, "text/plain", "Hello World!");
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Send large webpage from PROGMEM

const char index_html[] PROGMEM = "..."; // large char array, tested with 14k
request->send_P(200, "text/html", index_html);

Send large webpage from PROGMEM and extra headers

const char index_html[] PROGMEM = "..."; // large char array, tested with 14k
AsyncWebServerResponse *response = request->beginResponse_P(200, "text/html", index_html);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Send large webpage from PROGMEM containing templates

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

const char index_html[] PROGMEM = "..."; // large char array, tested with 14k
request->send_P(200, "text/html", index_html, processor);

Send large webpage from PROGMEM containing templates and extra headers

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

const char index_html[] PROGMEM = "..."; // large char array, tested with 14k
AsyncWebServerResponse *response = request->beginResponse_P(200, "text/html", index_html, processor);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Send binary content from PROGMEM

//File: favicon.ico.gz, Size: 726
#define favicon_ico_gz_len 726
const uint8_t favicon_ico_gz[] PROGMEM = {
 0x1F, 0x8B, 0x08, 0x08, 0x0B, 0x87, 0x90, 0x57, 0x00, 0x03, 0x66, 0x61, 0x76, 0x69, 0x63, 0x6F,
 0x6E, 0x2E, 0x69, 0x63, 0x6F, 0x00, 0xCD, 0x53, 0x5F, 0x48, 0x9A, 0x51, 0x14, 0xBF, 0x62, 0x6D,
 0x86, 0x96, 0xA9, 0x64, 0xD3, 0xFE, 0xA8, 0x99, 0x65, 0x1A, 0xB4, 0x8A, 0xA8, 0x51, 0x54, 0x23,
 0xA8, 0x11, 0x49, 0x51, 0x8A, 0x34, 0x62, 0x93, 0x85, 0x31, 0x58, 0x44, 0x12, 0x45, 0x2D, 0x58,
 0xF5, 0x52, 0x41, 0x10, 0x23, 0x82, 0xA0, 0x20, 0x98, 0x2F, 0xC1, 0x26, 0xED, 0xA1, 0x20, 0x89,
 0x04, 0xD7, 0x83, 0x58, 0x20, 0x28, 0x04, 0xAB, 0xD1, 0x9B, 0x8C, 0xE5, 0xC3, 0x60, 0x32, 0x64,
 0x0E, 0x56, 0xBF, 0x9D, 0xEF, 0xF6, 0x30, 0x82, 0xED, 0xAD, 0x87, 0xDD, 0x8F, 0xF3, 0xDD, 0x8F,
 0x73, 0xCF, 0xEF, 0x9C, 0xDF, 0x39, 0xBF, 0xFB, 0x31, 0x26, 0xA2, 0x27, 0x37, 0x97, 0xD1, 0x5B,
 0xCF, 0x9E, 0x67, 0x30, 0xA6, 0x66, 0x8C, 0x99, 0xC9, 0xC8, 0x45, 0x9E, 0x6B, 0x3F, 0x5F, 0x74,
 0xA6, 0x94, 0x5E, 0xDB, 0xFF, 0xB2, 0xE6, 0xE7, 0xE7, 0xF9, 0xDE, 0xD6, 0xD6, 0x96, 0xDB, 0xD8,
 0xD8, 0x78, 0xBF, 0xA1, 0xA1, 0xC1, 0xDA, 0xDC, 0xDC, 0x2C, 0xEB, 0xED, 0xED, 0x15, 0x9B, 0xCD,
 0xE6, 0x4A, 0x83, 0xC1, 0xE0, 0x2E, 0x29, 0x29, 0x99, 0xD6, 0x6A, 0xB5, 0x4F, 0x75, 0x3A, 0x9D,
 0x61, 0x75, 0x75, 0x95, 0xB5, 0xB7, 0xB7, 0xDF, 0xC8, 0xD1, 0xD4, 0xD4, 0xF4, 0xB0, 0xBA, 0xBA,
 0xFA, 0x83, 0xD5, 0x6A, 0xFD, 0x5A, 0x5E, 0x5E, 0x9E, 0x28, 0x2D, 0x2D, 0x0D, 0x10, 0xC6, 0x4B,
 0x98, 0x78, 0x5E, 0x5E, 0xDE, 0x95, 0x42, 0xA1, 0x40, 0x4E, 0x4E, 0xCE, 0x65, 0x76, 0x76, 0xF6,
 0x47, 0xB5, 0x5A, 0x6D, 0x4F, 0x26, 0x93, 0xA2, 0xD6, 0xD6, 0x56, 0x8E, 0x6D, 0x69, 0x69, 0xD1,
 0x11, 0x36, 0x62, 0xB1, 0x58, 0x60, 0x32, 0x99, 0xA0, 0xD7, 0xEB, 0x51, 0x58, 0x58, 0x88, 0xFC,
 0xFC, 0x7C, 0x10, 0x16, 0x02, 0x56, 0x2E, 0x97, 0x43, 0x2A, 0x95, 0x42, 0x2C, 0x16, 0x23, 0x33,
 0x33, 0x33, 0xAE, 0x52, 0xA9, 0x1E, 0x64, 0x65, 0x65, 0x71, 0x7C, 0x7D, 0x7D, 0xBD, 0x93, 0xEA,
 0xFE, 0x30, 0x1A, 0x8D, 0xE8, 0xEC, 0xEC, 0xC4, 0xE2, 0xE2, 0x22, 0x6A, 0x6A, 0x6A, 0x40, 0x39,
 0x41, 0xB5, 0x38, 0x4E, 0xC8, 0x33, 0x3C, 0x3C, 0x0C, 0x87, 0xC3, 0xC1, 0x6B, 0x54, 0x54, 0x54,
 0xBC, 0xE9, 0xEB, 0xEB, 0x93, 0x5F, 0x5C, 0x5C, 0x30, 0x8A, 0x9D, 0x2E, 0x2B, 0x2B, 0xBB, 0xA2,
 0x3E, 0x41, 0xBD, 0x21, 0x1E, 0x8F, 0x63, 0x6A, 0x6A, 0x0A, 0x81, 0x40, 0x00, 0x94, 0x1B, 0x3D,
 0x3D, 0x3D, 0x42, 0x3C, 0x96, 0x96, 0x96, 0x70, 0x7E, 0x7E, 0x8E, 0xE3, 0xE3, 0x63, 0xF8, 0xFD,
 0xFE, 0xB4, 0xD7, 0xEB, 0xF5, 0x8F, 0x8F, 0x8F, 0x5B, 0x68, 0x5E, 0x6F, 0x05, 0xCE, 0xB4, 0xE3,
 0xE8, 0xE8, 0x08, 0x27, 0x27, 0x27, 0xD8, 0xDF, 0xDF, 0xC7, 0xD9, 0xD9, 0x19, 0x6C, 0x36, 0x1B,
 0x36, 0x36, 0x36, 0x38, 0x9F, 0x85, 0x85, 0x05, 0xAC, 0xAF, 0xAF, 0x23, 0x1A, 0x8D, 0x22, 0x91,
 0x48, 0x20, 0x16, 0x8B, 0xFD, 0xDA, 0xDA, 0xDA, 0x7A, 0x41, 0x33, 0x7E, 0x57, 0x50, 0x50, 0x80,
 0x89, 0x89, 0x09, 0x84, 0xC3, 0x61, 0x6C, 0x6F, 0x6F, 0x23, 0x12, 0x89, 0xE0, 0xE0, 0xE0, 0x00,
 0x43, 0x43, 0x43, 0x58, 0x5E, 0x5E, 0xE6, 0x9C, 0x7D, 0x3E, 0x1F, 0x46, 0x47, 0x47, 0x79, 0xBE,
 0xBD, 0xBD, 0x3D, 0xE1, 0x3C, 0x1D, 0x0C, 0x06, 0x9F, 0x10, 0xB7, 0xC7, 0x84, 0x4F, 0xF6, 0xF7,
 0xF7, 0x63, 0x60, 0x60, 0x00, 0x83, 0x83, 0x83, 0x18, 0x19, 0x19, 0xC1, 0xDC, 0xDC, 0x1C, 0x8F,
 0x17, 0x7C, 0xA4, 0x27, 0xE7, 0x34, 0x39, 0x39, 0x89, 0x9D, 0x9D, 0x1D, 0x6E, 0x54, 0xE3, 0x13,
 0xE5, 0x34, 0x11, 0x37, 0x49, 0x51, 0x51, 0xD1, 0x4B, 0xA5, 0x52, 0xF9, 0x45, 0x26, 0x93, 0x5D,
 0x0A, 0xF3, 0x92, 0x48, 0x24, 0xA0, 0x6F, 0x14, 0x17, 0x17, 0xA3, 0xB6, 0xB6, 0x16, 0x5D, 0x5D,
 0x5D, 0x7C, 0x1E, 0xBB, 0xBB, 0xBB, 0x9C, 0xD7, 0xE1, 0xE1, 0x21, 0x42, 0xA1, 0xD0, 0x6B, 0xD2,
 0x45, 0x4C, 0x33, 0x12, 0x34, 0xCC, 0xA0, 0x19, 0x54, 0x92, 0x56, 0x0E, 0xD2, 0xD9, 0x43, 0xF8,
 0xCF, 0x82, 0x56, 0xC2, 0xDC, 0xEB, 0xEA, 0xEA, 0x38, 0x7E, 0x6C, 0x6C, 0x4C, 0xE0, 0xFE, 0x9D,
 0xB8, 0xBF, 0xA7, 0xFA, 0xAF, 0x56, 0x56, 0x56, 0xEE, 0x6D, 0x6E, 0x6E, 0xDE, 0xB8, 0x47, 0x55,
 0x55, 0x55, 0x6C, 0x66, 0x66, 0x46, 0x44, 0xDA, 0x3B, 0x34, 0x1A, 0x4D, 0x94, 0xB0, 0x3F, 0x09,
 0x7B, 0x45, 0xBD, 0xA5, 0x5D, 0x2E, 0x57, 0x8C, 0x7A, 0x73, 0xD9, 0xED, 0xF6, 0x3B, 0x84, 0xFF,
 0xE7, 0x7D, 0xA6, 0x3A, 0x2C, 0x95, 0x4A, 0xB1, 0x8E, 0x8E, 0x0E, 0x6D, 0x77, 0x77, 0xB7, 0xCD,
 0xE9, 0x74, 0x3E, 0x73, 0xBB, 0xDD, 0x8F, 0x3C, 0x1E, 0x8F, 0xE6, 0xF4, 0xF4, 0x94, 0xAD, 0xAD,
 0xAD, 0xDD, 0xDE, 0xCF, 0x73, 0x0B, 0x0B, 0xB8, 0xB6, 0xE0, 0x5D, 0xC6, 0x66, 0xC5, 0xE4, 0x10,
 0x4C, 0xF4, 0xF7, 0xD8, 0x59, 0xF2, 0x7F, 0xA3, 0xB8, 0xB4, 0xFC, 0x0F, 0xEE, 0x37, 0x70, 0xEC,
 0x16, 0x4A, 0x7E, 0x04, 0x00, 0x00
};

AsyncWebServerResponse *response = request->beginResponse_P(200, "image/x-icon", favicon_ico_gz, favicon_ico_gz_len);
response->addHeader("Content-Encoding", "gzip");
request->send(response);

Respond with content coming from a Stream

//read 12 bytes from Serial and send them as Content Type text/plain
request->send(Serial, "text/plain", 12);

Respond with content coming from a Stream and extra headers

//read 12 bytes from Serial and send them as Content Type text/plain
AsyncWebServerResponse *response = request->beginResponse(Serial, "text/plain", 12);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Respond with content coming from a Stream containing templates

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

//read 12 bytes from Serial and send them as Content Type text/plain
request->send(Serial, "text/plain", 12, processor);

Respond with content coming from a Stream containing templates and extra headers

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

//read 12 bytes from Serial and send them as Content Type text/plain
AsyncWebServerResponse *response = request->beginResponse(Serial, "text/plain", 12, processor);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Respond with content coming from a File

//Send index.htm with default content type
request->send(SPIFFS, "/index.htm");

//Send index.htm as text
request->send(SPIFFS, "/index.htm", "text/plain");

//Download index.htm
request->send(SPIFFS, "/index.htm", String(), true);

Respond with content coming from a File and extra headers

//Send index.htm with default content type
AsyncWebServerResponse *response = request->beginResponse(SPIFFS, "/index.htm");

//Send index.htm as text
AsyncWebServerResponse *response = request->beginResponse(SPIFFS, "/index.htm", "text/plain");

//Download index.htm
AsyncWebServerResponse *response = request->beginResponse(SPIFFS, "/index.htm", String(), true);

response->addHeader("Server","ESP Async Web Server");
request->send(response);

Respond with content coming from a File containing templates

Internally uses Chunked Response.

Index.htm contents:

%HELLO_FROM_TEMPLATE%

Somewhere in source files:

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

//Send index.htm with template processor function
request->send(SPIFFS, "/index.htm", String(), false, processor);

Respond with content using a callback

//send 128 bytes as plain text
request->send("text/plain", 128, [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will not be asked for more bytes once the content length has been reached.
  //Keep in mind that you can not delay or yield waiting for more data!
  //Send what you currently have and you will be asked for more again
  return mySource.read(buffer, maxLen);
});

Respond with content using a callback and extra headers

//send 128 bytes as plain text
AsyncWebServerResponse *response = request->beginResponse("text/plain", 128, [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will not be asked for more bytes once the content length has been reached.
  //Keep in mind that you can not delay or yield waiting for more data!
  //Send what you currently have and you will be asked for more again
  return mySource.read(buffer, maxLen);
});
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Respond with file content using a callback and extra headers

With this code your ESP is able to serve even large (large in terms of ESP, e.g. 100kB) files without memory problems.

You need to create a file handler in outer function (to have a single one for request) but use it in a lambda. The catch is that the lambda has it's own lifecycle which may/will cause it's called after the original function is over thus the original file handle is destroyed. Using the captured &file in the lambda then causes segfault (Hello, Exception 9!) and the whole ESP crashes.
By using this code, you tell the compiler to move the handle into the lambda so it won't be destroyed when outer function (that one where you call request->send(response)) ends.

const File file = ... // e.g. SPIFFS.open(path, "r");

const contentType = "application/javascript";

AsyncWebServerResponse *response = request->beginResponse(
  contentType,
  file.size(),
  [file](uint8_t *buffer, size_t maxLen, size_t total) mutable -> size_t {
     int bytes = file.read(buffer, maxLen);

     // close file at the end
     if (bytes + total == file.size()) file.close();

     return max(0, bytes); // return 0 even when no bytes were loaded
   }
);

if (gzipped) {
  response->addHeader(F("Content-Encoding"), F("gzip"));
}

request->send(response);

Respond with content using a callback containing templates

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

//send 128 bytes as plain text
request->send("text/plain", 128, [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will not be asked for more bytes once the content length has been reached.
  //Keep in mind that you can not delay or yield waiting for more data!
  //Send what you currently have and you will be asked for more again
  return mySource.read(buffer, maxLen);
}, processor);

Respond with content using a callback containing templates and extra headers

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

//send 128 bytes as plain text
AsyncWebServerResponse *response = request->beginResponse("text/plain", 128, [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will not be asked for more bytes once the content length has been reached.
  //Keep in mind that you can not delay or yield waiting for more data!
  //Send what you currently have and you will be asked for more again
  return mySource.read(buffer, maxLen);
}, processor);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Chunked Response

Used when content length is unknown. Works best if the client supports HTTP/1.1

AsyncWebServerResponse *response = request->beginChunkedResponse("text/plain", [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will be asked for more data until 0 is returned
  //Keep in mind that you can not delay or yield waiting for more data!
  return mySource.read(buffer, maxLen);
});
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Chunked Response containing templates

Used when content length is unknown. Works best if the client supports HTTP/1.1

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

AsyncWebServerResponse *response = request->beginChunkedResponse("text/plain", [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //index equals the amount of bytes that have been already sent
  //You will be asked for more data until 0 is returned
  //Keep in mind that you can not delay or yield waiting for more data!
  return mySource.read(buffer, maxLen);
}, processor);
response->addHeader("Server","ESP Async Web Server");
request->send(response);

Print to response

AsyncResponseStream *response = request->beginResponseStream("text/html");
response->addHeader("Server","ESP Async Web Server");
response->printf("<!DOCTYPE html><html><head><title>Webpage at %s</title></head><body>", request->url().c_str());

response->print("<h2>Hello ");
response->print(request->client()->remoteIP());
response->print("</h2>");

response->print("<h3>General</h3>");
response->print("<ul>");
response->printf("<li>Version: HTTP/1.%u</li>", request->version());
response->printf("<li>Method: %s</li>", request->methodToString());
response->printf("<li>URL: %s</li>", request->url().c_str());
response->printf("<li>Host: %s</li>", request->host().c_str());
response->printf("<li>ContentType: %s</li>", request->contentType().c_str());
response->printf("<li>ContentLength: %u</li>", request->contentLength());
response->printf("<li>Multipart: %s</li>", request->multipart()?"true":"false");
response->print("</ul>");

response->print("<h3>Headers</h3>");
response->print("<ul>");
int headers = request->headers();
for(int i=0;i<headers;i++){
  AsyncWebHeader* h = request->getHeader(i);
  response->printf("<li>%s: %s</li>", h->name().c_str(), h->value().c_str());
}
response->print("</ul>");

response->print("<h3>Parameters</h3>");
response->print("<ul>");
int params = request->params();
for(int i=0;i<params;i++){
  AsyncWebParameter* p = request->getParam(i);
  if(p->isFile()){
    response->printf("<li>FILE[%s]: %s, size: %u</li>", p->name().c_str(), p->value().c_str(), p->size());
  } else if(p->isPost()){
    response->printf("<li>POST[%s]: %s</li>", p->name().c_str(), p->value().c_str());
  } else {
    response->printf("<li>GET[%s]: %s</li>", p->name().c_str(), p->value().c_str());
  }
}
response->print("</ul>");

response->print("</body></html>");
//send the response last
request->send(response);

ArduinoJson Basic Response

This way of sending Json is great for when the result is below 4KB

#include "AsyncJson.h"
#include "ArduinoJson.h"


AsyncResponseStream *response = request->beginResponseStream("application/json");
DynamicJsonBuffer jsonBuffer;
JsonObject &root = jsonBuffer.createObject();
root["heap"] = ESP.getFreeHeap();
root["ssid"] = WiFi.SSID();
root.printTo(*response);
request->send(response);

ArduinoJson Advanced Response

This response can handle really large Json objects (tested to 40KB) There isn't any noticeable speed decrease for small results with the method above Since ArduinoJson does not allow reading parts of the string, the whole Json has to be passed every time a chunks needs to be sent, which shows speed decrease proportional to the resulting json packets

#include "AsyncJson.h"
#include "ArduinoJson.h"


AsyncJsonResponse * response = new AsyncJsonResponse();
response->addHeader("Server","ESP Async Web Server");
JsonObject& root = response->getRoot();
root["heap"] = ESP.getFreeHeap();
root["ssid"] = WiFi.SSID();
response->setLength();
request->send(response);

Serving static files

In addition to serving files from SPIFFS as described above, the server provide a dedicated handler that optimize the performance of serving files from SPIFFS - AsyncStaticWebHandler. Use server.serveStatic() function to initialize and add a new instance of AsyncStaticWebHandler to the server. The Handler will not handle the request if the file does not exists, e.g. the server will continue to look for another handler that can handle the request. Notice that you can chain setter functions to setup the handler, or keep a pointer to change it at a later time.

Serving specific file by name

// Serve the file "/www/page.htm" when request url is "/page.htm"
server.serveStatic("/page.htm", SPIFFS, "/www/page.htm");

Serving files in directory

To serve files in a directory, the path to the files should specify a directory in SPIFFS and ends with "/".

// Serve files in directory "/www/" when request url starts with "/"
// Request to the root or none existing files will try to server the defualt
// file name "index.htm" if exists
server.serveStatic("/", SPIFFS, "/www/");

// Server with different default file
server.serveStatic("/", SPIFFS, "/www/").setDefaultFile("default.html");

Serving static files with authentication

server
    .serveStatic("/", SPIFFS, "/www/")
    .setDefaultFile("default.html")
    .setAuthentication("user", "pass");

Specifying Cache-Control header

It is possible to specify Cache-Control header value to reduce the number of calls to the server once the client loaded the files. For more information on Cache-Control values see Cache-Control

// Cache responses for 10 minutes (600 seconds)
server.serveStatic("/", SPIFFS, "/www/").setCacheControl("max-age=600");

//*** Change Cache-Control after server setup ***

// During setup - keep a pointer to the handler
AsyncStaticWebHandler* handler = &server.serveStatic("/", SPIFFS, "/www/").setCacheControl("max-age=600");

// At a later event - change Cache-Control
handler->setCacheControl("max-age=30");

Specifying Date-Modified header

It is possible to specify Date-Modified header to enable the server to return Not-Modified (304) response for requests with "If-Modified-Since" header with the same value, instead of responding with the actual file content.

// Update the date modified string every time files are updated
server.serveStatic("/", SPIFFS, "/www/").setLastModified("Mon, 20 Jun 2016 14:00:00 GMT");

//*** Chage last modified value at a later stage ***

// During setup - read last modified value from config or EEPROM
String date_modified = loadDateModified();
AsyncStaticWebHandler* handler = &server.serveStatic("/", SPIFFS, "/www/");
handler->setLastModified(date_modified);

// At a later event when files are updated
String date_modified = getNewDateModfied();
saveDateModified(date_modified); // Save for next reset
handler->setLastModified(date_modified);

Specifying Template Processor callback

It is possible to specify template processor for static files. For information on template processor see Respond with content coming from a File containing templates.

String processor(const String& var)
{
  if(var == "HELLO_FROM_TEMPLATE")
    return F("Hello world!");
  return String();
}

// ...

server.serveStatic("/", SPIFFS, "/www/").setTemplateProcessor(processor);

Serving static files by custom handling

It may happen your static files are too big and the ESP will crash the request before it sends the whole file.
In that case, you can handle static files with custom file serving through not found handler.

This code below is more-or-less equivalent to this:

webServer.serveStatic("/", SPIFFS, STATIC_FILES_PREFIX).setDefaultFile("index.html")

First, declare the handling function:

bool handleStaticFile(AsyncWebServerRequest *request) {
  String path = STATIC_FILES_PREFIX + request->url();

  if (path.endsWith("/")) path += F("index.html");

  String contentType = getContentType(path);
  String pathWithGz = path + ".gz";

  if (SPIFFS.exists(pathWithGz) || SPIFFS.exists(path)) {
    bool gzipped = false;
    if (SPIFFS.exists(pathWithGz)) {
        gzipped = true;
        path += ".gz";
    }

    // TODO serve the file

    return true;
  }

  return false;
}

And then configure your webserver:

webServer.onNotFound([](AsyncWebServerRequest *request) {
  if (handleStaticFile(request)) return;

  request->send(404);
});

You may want to try Respond with file content using a callback and extra headers For actual serving the file.

Param Rewrite With Matching

It is possible to rewrite the request url with parameter match. Here is an example with one parameter: Rewrite for example "/radio/{frequence}" -> "/radio?f={frequence}"

class OneParamRewrite : public AsyncWebRewrite
{
  protected:
    String _urlPrefix;
    int _paramIndex;
    String _paramsBackup;

  public:
  OneParamRewrite(const char* from, const char* to)
    : AsyncWebRewrite(from, to) {

      _paramIndex = _from.indexOf('{');

      if( _paramIndex >=0 && _from.endsWith("}")) {
        _urlPrefix = _from.substring(0, _paramIndex);
        int index = _params.indexOf('{');
        if(index >= 0) {
          _params = _params.substring(0, index);
        }
      } else {
        _urlPrefix = _from;
      }
      _paramsBackup = _params;
  }

  bool match(AsyncWebServerRequest *request) override {
    if(request->url().startsWith(_urlPrefix)) {
      if(_paramIndex >= 0) {
        _params = _paramsBackup + request->url().substring(_paramIndex);
      } else {
        _params = _paramsBackup;
      }
    return true;

    } else {
      return false;
    }
  }
};

Usage:

  server.addRewrite( new OneParamRewrite("/radio/{frequence}", "/radio?f={frequence}") );

Using filters

Filters can be set to Rewrite or Handler in order to control when to apply the rewrite and consider the handler. A filter is a callback function that evaluates the request and return a boolean true to include the item or false to exclude it. Two filter callback are provided for convince:

  • ON_STA_FILTER - return true when requests are made to the STA (station mode) interface.
  • ON_AP_FILTER - return true when requests are made to the AP (access point) interface.

Serve different site files in AP mode

server.serveStatic("/", SPIFFS, "/www/").setFilter(ON_STA_FILTER);
server.serveStatic("/", SPIFFS, "/ap/").setFilter(ON_AP_FILTER);

Rewrite to different index on AP

// Serve the file "/www/index-ap.htm" in AP, and the file "/www/index.htm" on STA
server.rewrite("/", "index.htm");
server.rewrite("/index.htm", "index-ap.htm").setFilter(ON_AP_FILTER);
server.serveStatic("/", SPIFFS, "/www/");

Serving different hosts

// Filter callback using request host
bool filterOnHost1(AsyncWebServerRequest *request) { return request->host() == "host1"; }

// Server setup: server files in "/host1/" to requests for "host1", and files in "/www/" otherwise.
server.serveStatic("/", SPIFFS, "/host1/").setFilter(filterOnHost1);
server.serveStatic("/", SPIFFS, "/www/");

Determine interface inside callbacks

  String RedirectUrl = "http://";
  if (ON_STA_FILTER(request)) {
    RedirectUrl += WiFi.localIP().toString();
  } else {
    RedirectUrl += WiFi.softAPIP().toString();
  }
  RedirectUrl += "/index.htm";
  request->redirect(RedirectUrl);

Bad Responses

Some responses are implemented, but you should not use them, because they do not conform to HTTP. The following example will lead to unclean close of the connection and more time wasted than providing the length of the content

Respond with content using a callback without content length to HTTP/1.0 clients

//This is used as fallback for chunked responses to HTTP/1.0 Clients
request->send("text/plain", 0, [](uint8_t *buffer, size_t maxLen, size_t index) -> size_t {
  //Write up to "maxLen" bytes into "buffer" and return the amount written.
  //You will be asked for more data until 0 is returned
  //Keep in mind that you can not delay or yield waiting for more data!
  return mySource.read(buffer, maxLen);
});

Async WebSocket Plugin

The server includes a web socket plugin which lets you define different WebSocket locations to connect to without starting another listening service or using different port

Async WebSocket Event

void onEvent(AsyncWebSocket * server, AsyncWebSocketClient * client, AwsEventType type, void * arg, uint8_t *data, size_t len){
  if(type == WS_EVT_CONNECT){
    //client connected
    os_printf("ws[%s][%u] connect\n", server->url(), client->id());
    client->printf("Hello Client %u :)", client->id());
    client->ping();
  } else if(type == WS_EVT_DISCONNECT){
    //client disconnected
    os_printf("ws[%s][%u] disconnect: %u\n", server->url(), client->id());
  } else if(type == WS_EVT_ERROR){
    //error was received from the other end
    os_printf("ws[%s][%u] error(%u): %s\n", server->url(), client->id(), *((uint16_t*)arg), (char*)data);
  } else if(type == WS_EVT_PONG){
    //pong message was received (in response to a ping request maybe)
    os_printf("ws[%s][%u] pong[%u]: %s\n", server->url(), client->id(), len, (len)?(char*)data:"");
  } else if(type == WS_EVT_DATA){
    //data packet
    AwsFrameInfo * info = (AwsFrameInfo*)arg;
    if(info->final && info->index == 0 && info->len == len){
      //the whole message is in a single frame and we got all of it's data
      os_printf("ws[%s][%u] %s-message[%llu]: ", server->url(), client->id(), (info->opcode == WS_TEXT)?"text":"binary", info->len);
      if(info->opcode == WS_TEXT){
        data[len] = 0;
        os_printf("%s\n", (char*)data);
      } else {
        for(size_t i=0; i < info->len; i++){
          os_printf("%02x ", data[i]);
        }
        os_printf("\n");
      }
      if(info->opcode == WS_TEXT)
        client->text("I got your text message");
      else
        client->binary("I got your binary message");
    } else {
      //message is comprised of multiple frames or the frame is split into multiple packets
      if(info->index == 0){
        if(info->num == 0)
          os_printf("ws[%s][%u] %s-message start\n", server->url(), client->id(), (info->message_opcode == WS_TEXT)?"text":"binary");
        os_printf("ws[%s][%u] frame[%u] start[%llu]\n", server->url(), client->id(), info->num, info->len);
      }

      os_printf("ws[%s][%u] frame[%u] %s[%llu - %llu]: ", server->url(), client->id(), info->num, (info->message_opcode == WS_TEXT)?"text":"binary", info->index, info->index + len);
      if(info->message_opcode == WS_TEXT){
        data[len] = 0;
        os_printf("%s\n", (char*)data);
      } else {
        for(size_t i=0; i < len; i++){
          os_printf("%02x ", data[i]);
        }
        os_printf("\n");
      }

      if((info->index + len) == info->len){
        os_printf("ws[%s][%u] frame[%u] end[%llu]\n", server->url(), client->id(), info->num, info->len);
        if(info->final){
          os_printf("ws[%s][%u] %s-message end\n", server->url(), client->id(), (info->message_opcode == WS_TEXT)?"text":"binary");
          if(info->message_opcode == WS_TEXT)
            client->text("I got your text message");
          else
            client->binary("I got your binary message");
        }
      }
    }
  }
}

Methods for sending data to a socket client

//Server methods
AsyncWebSocket ws("/ws");
//printf to a client
ws.printf((uint32_t)client_id, arguments...);
//printf to all clients
ws.printfAll(arguments...);
//printf_P to a client
ws.printf_P((uint32_t)client_id, PSTR(format), arguments...);
//printfAll_P to all clients
ws.printfAll_P(PSTR(format), arguments...);
//send text to a client
ws.text((uint32_t)client_id, (char*)text);
ws.text((uint32_t)client_id, (uint8_t*)text, (size_t)len);
//send text from PROGMEM to a client
ws.text((uint32_t)client_id, PSTR("text"));
const char flash_text[] PROGMEM = "Text to send"
ws.text((uint32_t)client_id, FPSTR(flash_text));
//send text to all clients
ws.textAll((char*)text);
ws.textAll((uint8_t*)text, (size_t)len);
//send binary to a client
ws.binary((uint32_t)client_id, (char*)binary);
ws.binary((uint32_t)client_id, (uint8_t*)binary, (size_t)len);
//send binary from PROGMEM to a client
const uint8_t flash_binary[] PROGMEM = { 0x01, 0x02, 0x03, 0x04 };
ws.binary((uint32_t)client_id, flash_binary, 4);
//send binary to all clients
ws.binaryAll((char*)binary);
ws.binaryAll((uint8_t*)binary, (size_t)len);
//HTTP Authenticate before switch to Websocket protocol
ws.setAuthentication("user", "pass");

//client methods
AsyncWebSocketClient * client;
//printf
client->printf(arguments...);
//printf_P
client->printf_P(PSTR(format), arguments...);
//send text
client->text((char*)text);
client->text((uint8_t*)text, (size_t)len);
//send text from PROGMEM
client->text(PSTR("text"));
const char flash_text[] PROGMEM = "Text to send";
client->text(FPSTR(flash_text));
//send binary
client->binary((char*)binary);
client->binary((uint8_t*)binary, (size_t)len);
//send binary from PROGMEM
const uint8_t flash_binary[] PROGMEM = { 0x01, 0x02, 0x03, 0x04 };
client->binary(flash_binary, 4);

Direct access to web socket message buffer

When sending a web socket message using the above methods a buffer is created. Under certain circumstances you might want to manipulate or populate this buffer directly from your application, for example to prevent unnecessary duplications of the data. This example below shows how to create a buffer and print data to it from an ArduinoJson object then send it.

void sendDataWs(AsyncWebSocketClient * client)
{
    DynamicJsonBuffer jsonBuffer;
    JsonObject& root = jsonBuffer.createObject();
    root["a"] = "abc";
    root["b"] = "abcd";
    root["c"] = "abcde";
    root["d"] = "abcdef";
    root["e"] = "abcdefg";
    size_t len = root.measureLength();
    AsyncWebSocketMessageBuffer * buffer = ws.makeBuffer(len); //  creates a buffer (len + 1) for you.
    if (buffer) {
        root.printTo((char *)buffer->get(), len + 1);
        if (client) {
            client->text(buffer);
        } else {
            ws.textAll(buffer);
        }
    }
}

Limiting the number of web socket clients

Browsers sometimes do not correctly close the websocket connection, even when the close() function is called in javascript. This will eventually exhaust the web server's resources and will cause the server to crash. Periodically calling the cleanClients() function from the main loop() function limits the number of clients by closing the oldest client when the maximum number of clients has been exceeded. This can called be every cycle, however, if you wish to use less power, then calling as infrequently as once per second is sufficient.

void loop(){
  ws.cleanupClients();
}

Async Event Source Plugin

The server includes EventSource (Server-Sent Events) plugin which can be used to send short text events to the browser. Difference between EventSource and WebSockets is that EventSource is single direction, text-only protocol.

Setup Event Source on the server

AsyncWebServer server(80);
AsyncEventSource events("/events");

void setup(){
  // setup ......
  events.onConnect([](AsyncEventSourceClient *client){
    if(client->lastId()){
      Serial.printf("Client reconnected! Last message ID that it gat is: %u\n", client->lastId());
    }
    //send event with message "hello!", id current millis
    // and set reconnect delay to 1 second
    client->send("hello!",NULL,millis(),1000);
  });
  //HTTP Basic authentication
  events.setAuthentication("user", "pass");
  server.addHandler(&events);
  // setup ......
}

void loop(){
  if(eventTriggered){ // your logic here
    //send event "myevent"
    events.send("my event content","myevent",millis());
  }
}

Setup Event Source in the browser

if (!!window.EventSource) {
  var source = new EventSource("/events");

  source.addEventListener(
    "open",
    function (e) {
      console.log("Events Connected");
    },
    false
  );

  source.addEventListener(
    "error",
    function (e) {
      if (e.target.readyState != EventSource.OPEN) {
        console.log("Events Disconnected");
      }
    },
    false
  );

  source.addEventListener(
    "message",
    function (e) {
      console.log("message", e.data);
    },
    false
  );

  source.addEventListener(
    "myevent",
    function (e) {
      console.log("myevent", e.data);
    },
    false
  );
}

Scanning for available WiFi Networks

//First request will return 0 results unless you start scan from somewhere else (loop/setup)
//Do not request more often than 3-5 seconds
server.on("/scan", HTTP_GET, [](AsyncWebServerRequest *request){
  String json = "[";
  int n = WiFi.scanComplete();
  if(n == -2){
    WiFi.scanNetworks(true);
  } else if(n){
    for (int i = 0; i < n; ++i){
      if(i) json += ",";
      json += "{";
      json += "\"rssi\":"+String(WiFi.RSSI(i));
      json += ",\"ssid\":\""+WiFi.SSID(i)+"\"";
      json += ",\"bssid\":\""+WiFi.BSSIDstr(i)+"\"";
      json += ",\"channel\":"+String(WiFi.channel(i));
      json += ",\"secure\":"+String(WiFi.encryptionType(i));
      json += ",\"hidden\":"+String(WiFi.isHidden(i)?"true":"false");
      json += "}";
    }
    WiFi.scanDelete();
    if(WiFi.scanComplete() == -2){
      WiFi.scanNetworks(true);
    }
  }
  json += "]";
  request->send(200, "application/json", json);
  json = String();
});

Remove handlers and rewrites

Server goes through handlers in same order as they were added. You can't simple add handler with same path to override them. To remove handler:

// save callback for particular URL path
auto handler = server.on("/some/path", [](AsyncWebServerRequest *request){
  //do something useful
});
// when you don't need handler anymore remove it
server.removeHandler(&handler);

// same with rewrites
server.removeRewrite(&someRewrite);

server.onNotFound([](AsyncWebServerRequest *request){
  request->send(404);
});

// remove server.onNotFound handler
server.onNotFound(NULL);

// remove all rewrites, handlers and onNotFound/onFileUpload/onRequestBody callbacks
server.reset();

Setting up the server

#include "ESPAsyncTCP.h"
#include "ESPAsyncWebServer.h"

AsyncWebServer server(80);
AsyncWebSocket ws("/ws"); // access at ws://[esp ip]/ws
AsyncEventSource events("/events"); // event source (Server-Sent events)

const char* ssid = "your-ssid";
const char* password = "your-pass";
const char* http_username = "admin";
const char* http_password = "admin";

//flag to use from web update to reboot the ESP
bool shouldReboot = false;

void onRequest(AsyncWebServerRequest *request){
  //Handle Unknown Request
  request->send(404);
}

void onBody(AsyncWebServerRequest *request, uint8_t *data, size_t len, size_t index, size_t total){
  //Handle body
}

void onUpload(AsyncWebServerRequest *request, String filename, size_t index, uint8_t *data, size_t len, bool final){
  //Handle upload
}

void onEvent(AsyncWebSocket * server, AsyncWebSocketClient * client, AwsEventType type, void * arg, uint8_t *data, size_t len){
  //Handle WebSocket event
}

void setup(){
  Serial.begin(115200);
  WiFi.mode(WIFI_STA);
  WiFi.begin(ssid, password);
  if (WiFi.waitForConnectResult() != WL_CONNECTED) {
    Serial.printf("WiFi Failed!\n");
    return;
  }

  // attach AsyncWebSocket
  ws.onEvent(onEvent);
  server.addHandler(&ws);

  // attach AsyncEventSource
  server.addHandler(&events);

  // respond to GET requests on URL /heap
  server.on("/heap", HTTP_GET, [](AsyncWebServerRequest *request){
    request->send(200, "text/plain", String(ESP.getFreeHeap()));
  });

  // upload a file to /upload
  server.on("/upload", HTTP_POST, [](AsyncWebServerRequest *request){
    request->send(200);
  }, onUpload);

  // send a file when /index is requested
  server.on("/index", HTTP_ANY, [](AsyncWebServerRequest *request){
    request->send(SPIFFS, "/index.htm");
  });

  // HTTP basic authentication
  server.on("/login", HTTP_GET, [](AsyncWebServerRequest *request){
    if(!request->authenticate(http_username, http_password))
        return request->requestAuthentication();
    request->send(200, "text/plain", "Login Success!");
  });

  // Simple Firmware Update Form
  server.on("/update", HTTP_GET, [](AsyncWebServerRequest *request){
    request->send(200, "text/html", "<form method='POST' action='/update' enctype='multipart/form-data'><input type='file' name='update'><input type='submit' value='Update'></form>");
  });
  server.on("/update", HTTP_POST, [](AsyncWebServerRequest *request){
    shouldReboot = !Update.hasError();
    AsyncWebServerResponse *response = request->beginResponse(200, "text/plain", shouldReboot?"OK":"FAIL");
    response->addHeader("Connection", "close");
    request->send(response);
  },[](AsyncWebServerRequest *request, String filename, size_t index, uint8_t *data, size_t len, bool final){
    if(!index){
      Serial.printf("Update Start: %s\n", filename.c_str());
      Update.runAsync(true);
      if(!Update.begin((ESP.getFreeSketchSpace() - 0x1000) & 0xFFFFF000)){
        Update.printError(Serial);
      }
    }
    if(!Update.hasError()){
      if(Update.write(data, len) != len){
        Update.printError(Serial);
      }
    }
    if(final){
      if(Update.end(true)){
        Serial.printf("Update Success: %uB\n", index+len);
      } else {
        Update.printError(Serial);
      }
    }
  });

  // attach filesystem root at URL /fs
  server.serveStatic("/fs", SPIFFS, "/");

  // Catch-All Handlers
  // Any request that can not find a Handler that canHandle it
  // ends in the callbacks below.
  server.onNotFound(onRequest);
  server.onFileUpload(onUpload);
  server.onRequestBody(onBody);

  server.begin();
}

void loop(){
  if(shouldReboot){
    Serial.println("Rebooting...");
    delay(100);
    ESP.restart();
  }
  static char temp[128];
  sprintf(temp, "Seconds since boot: %u", millis()/1000);
  events.send(temp, "time"); //send event "time"
}

Setup global and class functions as request handlers

#include <Arduino.h>
#include <ESPAsyncWebserver.h>
#include <Hash.h>
#include <functional>

void handleRequest(AsyncWebServerRequest *request){}

class WebClass {
public :
  AsyncWebServer classWebServer = AsyncWebServer(81);

  WebClass(){};

  void classRequest (AsyncWebServerRequest *request){}

  void begin(){
    // attach global request handler
    classWebServer.on("/example", HTTP_ANY, handleRequest);

    // attach class request handler
    classWebServer.on("/example", HTTP_ANY, std::bind(&WebClass::classRequest, this, std::placeholders::_1));
  }
};

AsyncWebServer globalWebServer(80);
WebClass webClassInstance;

void setup() {
  // attach global request handler
  globalWebServer.on("/example", HTTP_ANY, handleRequest);

  // attach class request handler
  globalWebServer.on("/example", HTTP_ANY, std::bind(&WebClass::classRequest, webClassInstance, std::placeholders::_1));
}

void loop() {

}

Methods for controlling websocket connections

  // Disable client connections if it was activated
  if ( ws.enabled() )
    ws.enable(false);

  // enable client connections if it was disabled
  if ( !ws.enabled() )
    ws.enable(true);

Example of OTA code

  // OTA callbacks
  ArduinoOTA.onStart([]() {
    // Clean SPIFFS
    SPIFFS.end();

    // Disable client connections
    ws.enable(false);

    // Advertise connected clients what's going on
    ws.textAll("OTA Update Started");

    // Close them
    ws.closeAll();

  });

Adding Default Headers

In some cases, such as when working with CORS, or with some sort of custom authentication system, you might need to define a header that should get added to all responses (including static, websocket and EventSource). The DefaultHeaders singleton allows you to do this.

Example:

DefaultHeaders::Instance().addHeader("Access-Control-Allow-Origin", "*");
webServer.begin();

NOTE: You will still need to respond to the OPTIONS method for CORS pre-flight in most cases. (unless you are only using GET)

This is one option:

webServer.onNotFound([](AsyncWebServerRequest *request) {
  if (request->method() == HTTP_OPTIONS) {
    request->send(200);
  } else {
    request->send(404);
  }
});

Path variable

With path variable you can create a custom regex rule for a specific parameter in a route. For example we want a sensorId parameter in a route rule to match only a integer.

  server.on("^\\/sensor\\/([0-9]+)$", HTTP_GET, [] (AsyncWebServerRequest *request) {
      String sensorId = request->pathArg(0);
  });

NOTE: All regex patterns starts with ^ and ends with $

To enable the Path variable support, you have to define the buildflag -DASYNCWEBSERVER_REGEX.

For Arduino IDE create/update platform.local.txt:

Windows: C:\Users(username)\AppData\Local\Arduino15\packages\{espxxxx}\hardware\espxxxx\{version}\platform.local.txt

Linux: ~/.arduino15/packages/{espxxxx}/hardware/{espxxxx}/{version}/platform.local.txt

Add/Update the following line:

  compiler.cpp.extra_flags=-DDASYNCWEBSERVER_REGEX

For platformio modify platformio.ini:

[env:myboard]
build_flags =
  -DASYNCWEBSERVER_REGEX

NOTE: By enabling ASYNCWEBSERVER_REGEX, <regex> will be included. This will add an 100k to your binary.

About

Asynchronous HTTP and WebSocket Server Library for ESP32, ESP8266 and RP2040

mathieu.carbou.me/ESPAsyncWebServer/

Resources

Readme

License

LGPL-3.0 license

Code of conduct

Code of conduct
Activity

Stars

90 stars

Watchers

9 watching

Forks

17 forks
Report repository

Releases 65

v3.3.23 Latest
Nov 25, 2024
+ 64 releases

Sponsor this project

  •  
Learn more about GitHub Sponsors

Packages

No packages published

Languages

  • C++ 99.3%
  • Other 0.7%