README.md

function mySecondFunction(arg1, arg2, callback) {
  // arg1 now equals 'one' and arg2 now equals 'two'
  callback(null, 'three');
}
function myLastFunction(arg1, callback) {
  // arg1 now equals 'three'
  callback(null, 'done');
}
```

Or, if you need to pass any argument to the first function:

```js
async.waterfall([
    async.apply(myFirstFunction, 'zero'),
    mySecondFunction,
    myLastFunction,
], function (err, result) {
    // result now equals 'done'
});
function myFirstFunction(arg1, callback) {
  // arg1 now equals 'zero'
  callback(null, 'one', 'two');
}
function mySecondFunction(arg1, arg2, callback) {
  // arg1 now equals 'one' and arg2 now equals 'two'
  callback(null, 'three');
}
function myLastFunction(arg1, callback) {
  // arg1 now equals 'three'
  callback(null, 'done');
}
```

---------------------------------------
<a name="compose" />
### compose(fn1, fn2...)

Creates a function which is a composition of the passed asynchronous
functions. Each function consumes the return value of the function that
follows. Composing functions `f()`, `g()`, and `h()` would produce the result of
`f(g(h()))`, only this version uses callbacks to obtain the return values.

Each function is executed with the `this` binding of the composed function.

__Arguments__

* `functions...` - the asynchronous functions to compose


__Example__

```js
function add1(n, callback) {
    setTimeout(function () {
        callback(null, n + 1);
    }, 10);
}

function mul3(n, callback) {
    setTimeout(function () {
        callback(null, n * 3);
    }, 10);
}

var add1mul3 = async.compose(mul3, add1);

add1mul3(4, function (err, result) {
   // result now equals 15
});
```

---------------------------------------
<a name="seq" />
### seq(fn1, fn2...)

Version of the compose function that is more natural to read.
Each function consumes the return value of the previous function.
It is the equivalent of [`compose`](#compose) with the arguments reversed.

Each function is executed with the `this` binding of the composed function.

__Arguments__

* `functions...` - the asynchronous functions to compose


__Example__

```js
// Requires lodash (or underscore), express3 and dresende's orm2.
// Part of an app, that fetches cats of the logged user.
// This example uses `seq` function to avoid overnesting and error
// handling clutter.
app.get('/cats', function(request, response) {
  var User = request.models.User;
  async.seq(
    _.bind(User.get, User),  // 'User.get' has signature (id, callback(err, data))
    function(user, fn) {
      user.getCats(fn);      // 'getCats' has signature (callback(err, data))
    }
  )(req.session.user_id, function (err, cats) {
    if (err) {
      console.error(err);
      response.json({ status: 'error', message: err.message });
    } else {
      response.json({ status: 'ok', message: 'Cats found', data: cats });
    }
  });
});
```

---------------------------------------
<a name="applyEach" />
### applyEach(fns, args..., callback)

Applies the provided arguments to each function in the array, calling
`callback` after all functions have completed. If you only provide the first
argument, then it will return a function which lets you pass in the
arguments as if it were a single function call.

__Arguments__

* `fns` - the asynchronous functions to all call with the same arguments
* `args...` - any number of separate arguments to pass to the function
* `callback` - the final argument should be the callback, called when all
  functions have completed processing


__Example__

```js
async.applyEach([enableSearch, updateSchema], 'bucket', callback);

// partial application example:
async.each(
    buckets,
    async.applyEach([enableSearch, updateSchema]),
    callback
);
```

__Related__

* applyEachSeries(tasks, args..., [callback])

---------------------------------------

<a name="queue" />
### queue(worker, [concurrency])

Creates a `queue` object with the specified `concurrency`. Tasks added to the
`queue` are processed in parallel (up to the `concurrency` limit). If all
`worker`s are in progress, the task is queued until one becomes available.
Once a `worker` completes a `task`, that `task`'s callback is called.

__Arguments__

* `worker(task, callback)` - An asynchronous function for processing a queued
  task, which must call its `callback(err)` argument when finished, with an
  optional `error` as an argument.  If you want to handle errors from an individual task, pass a callback to `q.push()`.
* `concurrency` - An `integer` for determining how many `worker` functions should be
  run in parallel.  If omitted, the concurrency defaults to `1`.  If the concurrency is `0`, an error is thrown.

__Queue objects__

The `queue` object returned by this function has the following properties and
methods:

* `length()` - a function returning the number of items waiting to be processed.
* `started` - a function returning whether or not any items have been pushed and processed by the queue
* `running()` - a function returning the number of items currently being processed.
* `workersList()` - a function returning the array of items currently being processed.
* `idle()` - a function returning false if there are items waiting or being processed, or true if not.
* `concurrency` - an integer for determining how many `worker` functions should be
  run in parallel. This property can be changed after a `queue` is created to
  alter the concurrency on-the-fly.
* `push(task, [callback])` - add a new task to the `queue`. Calls `callback` once
  the `worker` has finished processing the task. Instead of a single task, a `tasks` array
  can be submitted. The respective callback is used for every task in the list.
* `unshift(task, [callback])` - add a new task to the front of the `queue`.
* `saturated` - a callback that is called when the `queue` length hits the `concurrency` limit,
   and further tasks will be queued.
* `empty` - a callback that is called when the last item from the `queue` is given to a `worker`.
* `drain` - a callback that is called when the last item from the `queue` has returned from the `worker`.
* `paused` - a boolean for determining whether the queue is in a paused state
* `pause()` - a function that pauses the processing of tasks until `resume()` is called.
* `resume()` - a function that resumes the processing of queued tasks when the queue is paused.
* `kill()` - a function that removes the `drain` callback and empties remaining tasks from the queue forcing it to go idle.

__Example__

```js
// create a queue object with concurrency 2

var q = async.queue(function (task, callback) {
    console.log('hello ' + task.name);
    callback();
}, 2);


// assign a callback
q.drain = function() {
    console.log('all items have been processed');
}

// add some items to the queue

q.push({name: 'foo'}, function (err) {
    console.log('finished processing foo');
});
q.push({name: 'bar'}, function (err) {
    console.log('finished processing bar');
});

// add some items to the queue (batch-wise)

q.push([{name: 'baz'},{name: 'bay'},{name: 'bax'}], function (err) {
    console.log('finished processing item');
});

// add some items to the front of the queue

q.unshift({name: 'bar'}, function (err) {
    console.log('finished processing bar');
});
```


---------------------------------------

<a name="priorityQueue" />
### priorityQueue(worker, concurrency)

The same as [`queue`](#queue) only tasks are assigned a priority and completed in ascending priority order. There are two differences between `queue` and `priorityQueue` objects:

* `push(task, priority, [callback])` - `priority` should be a number. If an array of
  `tasks` is given, all tasks will be assigned the same priority.
* The `unshift` method was removed.

---------------------------------------

<a name="cargo" />
### cargo(worker, [payload])

Creates a `cargo` object with the specified payload. Tasks added to the
cargo will be processed altogether (up to the `payload` limit). If the
`worker` is in progress, the task is queued until it becomes available. Once
the `worker` has completed some tasks, each callback of those tasks is called.
Check out [these](https://camo.githubusercontent.com/6bbd36f4cf5b35a0f11a96dcd2e97711ffc2fb37/68747470733a2f2f662e636c6f75642e6769746875622e636f6d2f6173736574732f313637363837312f36383130382f62626330636662302d356632392d313165322d393734662d3333393763363464633835382e676966) [animations](https://camo.githubusercontent.com/f4810e00e1c5f5f8addbe3e9f49064fd5d102699/68747470733a2f2f662e636c6f75642e6769746875622e636f6d2f6173736574732f313637363837312f36383130312f38346339323036362d356632392d313165322d383134662d3964336430323431336266642e676966) for how `cargo` and `queue` work.

While [queue](#queue) passes only one task to one of a group of workers
at a time, cargo passes an array of tasks to a single worker, repeating
when the worker is finished.

__Arguments__

* `worker(tasks, callback)` - An asynchronous function for processing an array of
  queued tasks, which must call its `callback(err)` argument when finished, with
  an optional `err` argument.
* `payload` - An optional `integer` for determining how many tasks should be
  processed per round; if omitted, the default is unlimited.

__Cargo objects__

The `cargo` object returned by this function has the following properties and
methods:

* `length()` - A function returning the number of items waiting to be processed.
* `payload` - An `integer` for determining how many tasks should be
  process per round. This property can be changed after a `cargo` is created to
  alter the payload on-the-fly.
* `push(task, [callback])` - Adds `task` to the `queue`. The callback is called
  once the `worker` has finished processing the task. Instead of a single task, an array of `tasks`
  can be submitted. The respective callback is used for every task in the list.
* `saturated` - A callback that is called when the `queue.length()` hits the concurrency and further tasks will be queued.
* `empty` - A callback that is called when the last item from the `queue` is given to a `worker`.
* `drain` - A callback that is called when the last item from the `queue` has returned from the `worker`.
* `idle()`, `pause()`, `resume()`, `kill()` - cargo inherits all of the same methods and event calbacks as [`queue`](#queue)

__Example__

```js
// create a cargo object with payload 2

var cargo = async.cargo(function (tasks, callback) {
    for(var i=0; i<tasks.length; i++){
      console.log('hello ' + tasks[i].name);
    }
    callback();
}, 2);


// add some items

cargo.push({name: 'foo'}, function (err) {
    console.log('finished processing foo');
});
cargo.push({name: 'bar'}, function (err) {
    console.log('finished processing bar');
});
cargo.push({name: 'baz'}, function (err) {
    console.log('finished processing baz');
});
```

---------------------------------------

<a name="auto" />
### auto(tasks, [concurrency], [callback])

Determines the best order for running the functions in `tasks`, based on their requirements. Each function can optionally depend on other functions being completed first, and each function is run as soon as its requirements are satisfied.

If any of the functions pass an error to their callback, the `auto` sequence will stop. Further tasks will not execute (so any other functions depending on it will not run), and the main `callback` is immediately called with the error.  Functions also receive an object containing the results of functions which have completed so far.

Note, all functions are called with a `results` object as a second argument,
so it is unsafe to pass functions in the `tasks` object which cannot handle the
extra argument.

For example, this snippet of code:

```js
async.auto({
  readData: async.apply(fs.readFile, 'data.txt', 'utf-8')
}, callback);
```

will have the effect of calling `readFile` with the results object as the last
argument, which will fail:

```js
fs.readFile('data.txt', 'utf-8', cb, {});
```

Instead, wrap the call to `readFile` in a function which does not forward the
`results` object:

```js
async.auto({
  readData: function(cb, results){
    fs.readFile('data.txt', 'utf-8', cb);
  }
}, callback);
```

__Arguments__

* `tasks` - An object. Each of its properties is either a function or an array of
  requirements, with the function itself the last item in the array. The object's key
  of a property serves as the name of the task defined by that property,
  i.e. can be used when specifying requirements for other tasks.
  The function receives two arguments: (1) a `callback(err, result)` which must be
  called when finished, passing an `error` (which can be `null`) and the result of
  the function's execution, and (2) a `results` object, containing the results of
  the previously executed functions.
* `concurrency` - An optional `integer` for determining the maximum number of tasks that can be run in parallel. By default, as many as possible.
* `callback(err, results)` - An optional callback which is called when all the
  tasks have been completed. It receives the `err` argument if any `tasks`
  pass an error to their callback. Results are always returned; however, if
  an error occurs, no further `tasks` will be performed, and the results
  object will only contain partial results.


__Example__

```js
async.auto({
    get_data: function(callback){
        console.log('in get_data');
        // async code to get some data
        callback(null, 'data', 'converted to array');
    },
    make_folder: function(callback){
        console.log('in make_folder');
        // async code to create a directory to store a file in
        // this is run at the same time as getting the data
        callback(null, 'folder');
    },
    write_file: ['get_data', 'make_folder', function(callback, results){
        console.log('in write_file', JSON.stringify(results));
        // once there is some data and the directory exists,
        // write the data to a file in the directory
        callback(null, 'filename');
    }],
    email_link: ['write_file', function(callback, results){
        console.log('in email_link', JSON.stringify(results));
        // once the file is written let's email a link to it...
        // results.write_file contains the filename returned by write_file.
        callback(null, {'file':results.write_file, 'email':'user@example.com'});
    }]
}, function(err, results) {
    console.log('err = ', err);
    console.log('results = ', results);
});
```

This is a fairly trivial example, but to do this using the basic parallel and
series functions would look like this:

```js
async.parallel([
    function(callback){
        console.log('in get_data');
        // async code to get some data
        callback(null, 'data', 'converted to array');
    },
    function(callback){
        console.log('in make_folder');
        // async code to create a directory to store a file in
        // this is run at the same time as getting the data
        callback(null, 'folder');
    }
],
function(err, results){
    async.series([
        function(callback){
            console.log('in write_file', JSON.stringify(results));
            // once there is some data and the directory exists,
            // write the data to a file in the directory
            results.push('filename');
            callback(null);
        },
        function(callback){
            console.log('in email_link', JSON.stringify(results));
            // once the file is written let's email a link to it...
            callback(null, {'file':results.pop(), 'email':'user@example.com'});
        }
    ]);
});
```

For a complicated series of `async` tasks, using the [`auto`](#auto) function makes adding
new tasks much easier (and the code more readable).


---------------------------------------

<a name="retry" />
### retry([opts = {times: 5, interval: 0}| 5], task, [callback])

Attempts to get a successful response from `task` no more than `times` times before
returning an error. If the task is successful, the `callback` will be passed the result
of the successful task. If all attempts fail, the callback will be passed the error and
result (if any) of the final attempt.

__Arguments__

* `opts` - Can be either an object with `times` and `interval` or a number.
  * `times` - The number of attempts to make before giving up.  The default is `5`.
  * `interval` - The time to wait between retries, in milliseconds.  The default is `0`.
  * If `opts` is a number, the number specifies the number of times to retry, with the default interval of `0`. 
* `task(callback, results)` - A function which receives two arguments: (1) a `callback(err, result)`
  which must be called when finished, passing `err` (which can be `null`) and the `result` of
  the function's execution, and (2) a `results` object, containing the results of
  the previously executed functions (if nested inside another control flow).
* `callback(err, results)` - An optional callback which is called when the
  task has succeeded, or after the final failed attempt. It receives the `err` and `result` arguments of the last attempt at completing the `task`.

The [`retry`](#retry) function can be used as a stand-alone control flow by passing a callback, as shown below:

```js
// try calling apiMethod 3 times
async.retry(3, apiMethod, function(err, result) {
    // do something with the result
});
```

```js
// try calling apiMethod 3 times, waiting 200 ms between each retry 
async.retry({times: 3, interval: 200}, apiMethod, function(err, result) {
    // do something with the result
});
```

```js
// try calling apiMethod the default 5 times no delay between each retry 
async.retry(apiMethod, function(err, result) {
    // do something with the result
});
```

It can also be embedded within other control flow functions to retry individual methods
that are not as reliable, like this:

```js
async.auto({
    users: api.getUsers.bind(api),
    payments: async.retry(3, api.getPayments.bind(api))
}, function(err, results) {
  // do something with the results
});
```


---------------------------------------

<a name="iterator" />
### iterator(tasks)

Creates an iterator function which calls the next function in the `tasks` array,
returning a continuation to call the next one after that. It's also possible to
“peek” at the next iterator with `iterator.next()`.

This function is used internally by the `async` module, but can be useful when
you want to manually control the flow of functions in series.

__Arguments__

* `tasks` - An array of functions to run.

__Example__

```js
var iterator = async.iterator([
    function(){ sys.p('one'); },
    function(){ sys.p('two'); },
    function(){ sys.p('three'); }
]);

node> var iterator2 = iterator();
'one'
node> var iterator3 = iterator2();
'two'
node> iterator3();
'three'
node> var nextfn = iterator2.next();
node> nextfn();
'three'
```

---------------------------------------

<a name="apply" />
### apply(function, arguments..)

Creates a continuation function with some arguments already applied.

Useful as a shorthand when combined with other control flow functions. Any arguments
passed to the returned function are added to the arguments originally passed
to apply.

__Arguments__

* `function` - The function you want to eventually apply all arguments to.
* `arguments...` - Any number of arguments to automatically apply when the
  continuation is called.

__Example__

```js
// using apply

async.parallel([
    async.apply(fs.writeFile, 'testfile1', 'test1'),
    async.apply(fs.writeFile, 'testfile2', 'test2'),
]);


// the same process without using apply

async.parallel([
    function(callback){
        fs.writeFile('testfile1', 'test1', callback);
    },
    function(callback){
        fs.writeFile('testfile2', 'test2', callback);
    }
]);
```

It's possible to pass any number of additional arguments when calling the
continuation:

```js
node> var fn = async.apply(sys.puts, 'one');
node> fn('two', 'three');
one
two
three
```

---------------------------------------

<a name="nextTick" />
### nextTick(callback), setImmediate(callback)

Calls `callback` on a later loop around the event loop. In Node.js this just
calls `process.nextTick`; in the browser it falls back to `setImmediate(callback)`
if available, otherwise `setTimeout(callback, 0)`, which means other higher priority
events may precede the execution of `callback`.

This is used internally for browser-compatibility purposes.

__Arguments__

* `callback` - The function to call on a later loop around the event loop.

__Example__

```js
var call_order = [];
async.nextTick(function(){
    call_order.push('two');
    // call_order now equals ['one','two']
});
call_order.push('one')
```

<a name="times" />
### times(n, iterator, [callback])

Calls the `iterator` function `n` times, and accumulates results in the same manner
you would use with [`map`](#map).

__Arguments__

* `n` - The number of times to run the function.
* `iterator` - The function to call `n` times.
* `callback` - see [`map`](#map)

__Example__

```js
// Pretend this is some complicated async factory
var createUser = function(id, callback) {
  callback(null, {
    id: 'user' + id
  })
}
// generate 5 users
async.times(5, function(n, next){
    createUser(n, function(err, user) {
      next(err, user)
    })
}, function(err, users) {
  // we should now have 5 users
});
```

__Related__

* timesSeries(n, iterator, [callback])
* timesLimit(n, limit, iterator, [callback])


## Utils

<a name="memoize" />
### memoize(fn, [hasher])

Caches the results of an `async` function. When creating a hash to store function
results against, the callback is omitted from the hash and an optional hash
function can be used.

If no hash function is specified, the first argument is used as a hash key, which may work reasonably if it is a string or a data type that converts to a distinct string. Note that objects and arrays will not behave reasonably. Neither will cases where the other arguments are significant. In such cases, specify your own hash function.

The cache of results is exposed as the `memo` property of the function returned
by `memoize`.

__Arguments__

* `fn` - The function to proxy and cache results from.
* `hasher` - An optional function for generating a custom hash for storing
  results. It has all the arguments applied to it apart from the callback, and
  must be synchronous.

__Example__

```js
var slow_fn = function (name, callback) {
    // do something
    callback(null, result);
};
var fn = async.memoize(slow_fn);

// fn can now be used as if it were slow_fn
fn('some name', function () {
    // callback
});
```

<a name="unmemoize" />
### unmemoize(fn)

Undoes a [`memoize`](#memoize)d function, reverting it to the original, unmemoized
form. Handy for testing.

__Arguments__

* `fn` - the memoized function

---------------------------------------

<a name="ensureAsync" />
### ensureAsync(fn)

Wrap an async function and ensure it calls its callback on a later tick of the event loop.  If the function already calls its callback on a next tick, no extra deferral is added. This is useful for preventing stack overflows (`RangeError: Maximum call stack size exceeded`) and generally keeping [Zalgo](http://blog.izs.me/post/59142742143/designing-apis-for-asynchrony) contained.

__Arguments__

* `fn` - an async function, one that expects a node-style callback as its last argument

Returns a wrapped function with the exact same call signature as the function passed in.

__Example__

```js
function sometimesAsync(arg, callback) {
  if (cache[arg]) {
    return callback(null, cache[arg]); // this would be synchronous!!
  } else {
    doSomeIO(arg, callback); // this IO would be asynchronous
  }
}

// this has a risk of stack overflows if many results are cached in a row
async.mapSeries(args, sometimesAsync, done);

// this will defer sometimesAsync's callback if necessary,
// preventing stack overflows
async.mapSeries(args, async.ensureAsync(sometimesAsync), done);

```

---------------------------------------

<a name="constant">
### constant(values...)

Returns a function that when called, calls-back with the values provided.  Useful as the first function in a `waterfall`, or for plugging values in to `auto`.

__Example__

```js
async.waterfall([
  async.constant(42),
  function (value, next) {
    // value === 42
  },
  //...
], callback);

async.waterfall([
  async.constant(filename, "utf8"),
  fs.readFile,
  function (fileData, next) {
    //...
  }
  //...
], callback);

async.auto({
  hostname: async.constant("https://server.net/"),
  port: findFreePort,
  launchServer: ["hostname", "port", function (cb, options) {
    startServer(options, cb);
  }],
  //...
}, callback);

```

---------------------------------------

<a name="asyncify">
<a name="wrapSync">
### asyncify(func)

__Alias:__ `wrapSync`

Take a sync function and make it async, passing its return value to a callback. This is useful for plugging sync functions into a waterfall, series, or other async functions. Any arguments passed to the generated function will be passed to the wrapped function (except for the final callback argument). Errors thrown will be passed to the callback.

__Example__

```js
async.waterfall([
  async.apply(fs.readFile, filename, "utf8"),
  async.asyncify(JSON.parse),
  function (data, next) {
    // data is the result of parsing the text.
    // If there was a parsing error, it would have been caught.
  }
], callback)
```

If the function passed to `asyncify` returns a Promise, that promises's resolved/rejected state will be used to call the callback, rather than simply the synchronous return value.  Example:

```js
async.waterfall([
  async.apply(fs.readFile, filename, "utf8"),
  async.asyncify(function (contents) {
    return db.model.create(contents);
  }),
  function (model, next) {
    // `model` is the instantiated model object. 
    // If there was an error, this function would be skipped.
  }
], callback)
```

This also means you can asyncify ES2016 `async` functions.

```js
var q = async.queue(async.asyncify(async function (file) {
  var intermediateStep = await processFile(file);
  return await somePromise(intermediateStep)
}));

q.push(files);
```

---------------------------------------

<a name="log" />
### log(function, arguments)

Logs the result of an `async` function to the `console`. Only works in Node.js or
in browsers that support `console.log` and `console.error` (such as FF and Chrome).
If multiple arguments are returned from the async function, `console.log` is
called on each argument in order.

__Arguments__

* `function` - The function you want to eventually apply all arguments to.
* `arguments...` - Any number of arguments to apply to the function.

__Example__

```js
var hello = function(name, callback){
    setTimeout(function(){
        callback(null, 'hello ' + name);
    }, 1000);
};
```
```js
node> async.log(hello, 'world');
'hello world'
```

---------------------------------------

<a name="dir" />
### dir(function, arguments)

Logs the result of an `async` function to the `console` using `console.dir` to
display the properties of the resulting object. Only works in Node.js or
in browsers that support `console.dir` and `console.error` (such as FF and Chrome).
If multiple arguments are returned from the async function, `console.dir` is
called on each argument in order.

__Arguments__

* `function` - The function you want to eventually apply all arguments to.
* `arguments...` - Any number of arguments to apply to the function.

__Example__

```js
var hello = function(name, callback){
    setTimeout(function(){
        callback(null, {hello: name});
    }, 1000);
};
```
```js
node> async.dir(hello, 'world');
{hello: 'world'}
```

---------------------------------------

<a name="noConflict" />
### noConflict()

Changes the value of `async` back to its original value, returning a reference to the
`async` object.