The term "mutex" usually refers to a data structure used to synchronize concurrent processes running on different threads. For example, before accessing a non-threadsafe resource, a thread will lock the mutex. This is guaranteed to block the thread until no other thread holds a lock on the mutex and thus enforces exclusive access to the resource. Once the operation is complete, the thread releases the lock, allowing other threads to acquire a lock and access the resource.
While Javascript is strictly single-threaded, the asynchronous nature of its execution model allows for race conditions that require similar synchronization primitives. Consider for example a library communicating with a web worker that needs to exchange several subsequent messages with the worker in order to achieve a task. As these messages are exchanged in an asynchronous manner, it is perfectly possible that the library is called again during this process. Depending on the way state is handled during the async process, this will lead to race conditions that are hard to fix and even harder to track down.
This library solves the problem by applying the concept of mutexes to Javascript. Locking the mutex will return a promise that resolves once the mutex becomes available. Once the async process is complete (usually taking multiple spins of the event loop), a callback supplied to the caller is called in order to release the mutex, allowing the next scheduled worker to execute.
Imagine a situation where you need to control access to several instances of a shared resource. For example, you might want to distribute images between several worker processes that perform transformations, or you might want to create a web crawler that performs a defined number of requests in parallel.
A semaphore is a data structure that is initialized to a positive integer value and that can be locked multiple times. As long as the semaphore value is positive, locking it will return the current value and the locking process will continue execution immediately; the semaphore will be decremented upon locking. Releasing the lock will increment the semaphore again.
Once the semaphore has reached zero, the next process that attempts to acquire a lock will be suspended until another process releases its lock and this increments the semaphore again.
This library provides a semaphore implementation for Javascript that is similar to the mutex implementation described above.