16.3.thread — Multiple threads of control¶

This module provides low-level primitives for working with multiple threads(also calledlight-weight processes ortasks) — multiple threads ofcontrol sharing their global data space. For synchronization, simple locks(also calledmutexes orbinary semaphores) are provided.Thethreading module provides an easier to use and higher-levelthreading API built on top of this module.

The module is optional. It is supported on Windows, Linux, SGI IRIX, Solaris2.x, as well as on systems that have a POSIX thread (a.k.a. “pthread”)implementation. For systems lacking thethread module, thedummy_thread module is available. It duplicates this module’s interfaceand can be used as a drop-in replacement.

It defines the following constant and functions:

exceptionthread.error¶

Raised on thread-specific errors.

thread.LockType¶

This is the type of lock objects.

thread.start_new_thread(function,args[,kwargs])¶

Start a new thread and return its identifier. The thread executes the functionfunction with the argument listargs (which must be a tuple). The optionalkwargs argument specifies a dictionary of keyword arguments. When the functionreturns, the thread silently exits. When the function terminates with anunhandled exception, a stack trace is printed and then the thread exits (butother threads continue to run).

thread.interrupt_main()¶

Raise aKeyboardInterrupt exception in the main thread. A subthread canuse this function to interrupt the main thread.

New in version 2.3.

thread.exit()¶

Raise theSystemExit exception. When not caught, this will cause thethread to exit silently.

thread.allocate_lock()¶

Return a new lock object. Methods of locks are described below. The lock isinitially unlocked.

thread.get_ident()¶

Return the ‘thread identifier’ of the current thread. This is a nonzerointeger. Its value has no direct meaning; it is intended as a magic cookie tobe used e.g. to index a dictionary of thread-specific data. Thread identifiersmay be recycled when a thread exits and another thread is created.

thread.stack_size([size])¶

Return the thread stack size used when creating new threads. The optionalsize argument specifies the stack size to be used for subsequently createdthreads, and must be 0 (use platform or configured default) or a positiveinteger value of at least 32,768 (32kB). Ifsize is not specified,0 is used. If changing the thread stack size isunsupported, theerror exception is raised. If the specified stack size isinvalid, aValueError is raised and the stack size is unmodified. 32kBis currently the minimum supported stack size value to guarantee sufficientstack space for the interpreter itself. Note that some platforms may haveparticular restrictions on values for the stack size, such as requiring aminimum stack size > 32kB or requiring allocation in multiples of the systemmemory page size - platform documentation should be referred to for moreinformation (4kB pages are common; using multiples of 4096 for the stack size isthe suggested approach in the absence of more specific information).Availability: Windows, systems with POSIX threads.

New in version 2.5.

Lock objects have the following methods:

lock.acquire([waitflag])¶

Without the optional argument, this method acquires the lock unconditionally, ifnecessary waiting until it is released by another thread (only one thread at atime can acquire a lock — that’s their reason for existence). If the integerwaitflag argument is present, the action depends on its value: if it is zero,the lock is only acquired if it can be acquired immediately without waiting,while if it is nonzero, the lock is acquired unconditionally as before. Thereturn value isTrue if the lock is acquired successfully,False if not.

lock.release()¶

Releases the lock. The lock must have been acquired earlier, but notnecessarily by the same thread.

lock.locked()¶

Return the status of the lock:True if it has been acquired by some thread,False if not.

In addition to these methods, lock objects can also be used via thewith statement, e.g.:

importthreada_lock=thread.allocate_lock()witha_lock:print"a_lock is locked while this executes"

Caveats:

• Threads interact strangely with interrupts: theKeyboardInterruptexception will be received by an arbitrary thread. (When thesignalmodule is available, interrupts always go to the main thread.)

• Callingsys.exit() or raising theSystemExit exception isequivalent to callingthread.exit().

• It is not possible to interrupt theacquire() method on a lock — theKeyboardInterrupt exception will happen after the lock has been acquired.

• When the main thread exits, it is system defined whether the other threadssurvive. On SGI IRIX using the native thread implementation, they survive. Onmost other systems, they are killed without executingtry …finally clauses or executing object destructors.

• When the main thread exits, it does not do any of its usual cleanup (exceptthattry …finally clauses are honored), and thestandard I/O files are not flushed.