(目前)排名第一的答案(带有ActiveState代码)过于复杂。当仅仅一个函数就足够了时，我不认为有理由使用类。下面是两个实现，它们实现了相同的功能，但代码可读性更强。

这两个实现:

可以在python2或python3中正常工作 可以在Windows, OSX和Linux上工作 只读取一个字节(即，它们不等待换行符) 不要依赖任何外部库 是自包含的(没有函数定义之外的代码)

版本1:易读且简单

def getChar():
    try:
        # for Windows-based systems
        import msvcrt # If successful, we are on Windows
        return msvcrt.getch()

    except ImportError:
        # for POSIX-based systems (with termios & tty support)
        import tty, sys, termios  # raises ImportError if unsupported

        fd = sys.stdin.fileno()
        oldSettings = termios.tcgetattr(fd)

        try:
            tty.setcbreak(fd)
            answer = sys.stdin.read(1)
        finally:
            termios.tcsetattr(fd, termios.TCSADRAIN, oldSettings)

        return answer

版本2:避免重复导入和异常处理:

[编辑]我错过了ActiveState代码的一个优点。如果您计划多次读取字符，该代码可以避免在类unix系统上重复Windows导入和ImportError异常处理的成本(可以忽略不计)。虽然你可能应该更关心代码的可读性，而不是可以忽略的优化，这里有一个替代方案(它类似于Louis的答案，但getChar()是自包含的)，它的功能与ActiveState代码相同，更可读:

def getChar():
    # figure out which function to use once, and store it in _func
    if "_func" not in getChar.__dict__:
        try:
            # for Windows-based systems
            import msvcrt # If successful, we are on Windows
            getChar._func=msvcrt.getch

        except ImportError:
            # for POSIX-based systems (with termios & tty support)
            import tty, sys, termios # raises ImportError if unsupported

            def _ttyRead():
                fd = sys.stdin.fileno()
                oldSettings = termios.tcgetattr(fd)

                try:
                    tty.setcbreak(fd)
                    answer = sys.stdin.read(1)
                finally:
                    termios.tcsetattr(fd, termios.TCSADRAIN, oldSettings)

                return answer

            getChar._func=_ttyRead

    return getChar._func()

使用上述getChar()版本的示例代码:

from __future__ import print_function # put at top of file if using Python 2

# Example of a prompt for one character of input
promptStr   = "Please give me a character:"
responseStr = "Thank you for giving me a '{}'."
print(promptStr, end="\n> ")
answer = getChar()
print("\n")
print(responseStr.format(answer))

内置的raw_input应该会有所帮助。

for i in range(3):
    print ("So much work to do!")
k = raw_input("Press any key to continue...")
print ("Ok, back to work.")

另一个回答中的一个评论提到了cbreak模式，这对Unix实现很重要，因为您通常不希望^C (KeyboardError)被getchar使用(当您将终端设置为原始模式时，它会被使用，就像大多数其他回答所做的那样)。

另一个重要的细节是，如果您希望读取一个字符而不是一个字节，您应该从输入流中读取4个字节，因为这是UTF-8 (Python 3+)中单个字符所包含的最大字节数。对于键盘箭头等多字节字符，只读取一个字节将产生意想不到的结果。

下面是我修改后的Unix实现:

import contextlib
import os
import sys
import termios
import tty


_MAX_CHARACTER_BYTE_LENGTH = 4


@contextlib.contextmanager
def _tty_reset(file_descriptor):
    """
    A context manager that saves the tty flags of a file descriptor upon
    entering and restores them upon exiting.
    """
    old_settings = termios.tcgetattr(file_descriptor)
    try:
        yield
    finally:
        termios.tcsetattr(file_descriptor, termios.TCSADRAIN, old_settings)


def get_character(file=sys.stdin):
    """
    Read a single character from the given input stream (defaults to sys.stdin).
    """
    file_descriptor = file.fileno()
    with _tty_reset(file_descriptor):
        tty.setcbreak(file_descriptor)
        return os.read(file_descriptor, _MAX_CHARACTER_BYTE_LENGTH)

python中的curses包可以使用一些语句从终端进入“原始”模式进行字符输入。Curses的主要用途是接管屏幕进行输出，这可能不是你想要的。此代码片段使用print()语句代替，这是可用的，但必须注意curses如何更改附加到输出的行结束符。

#!/usr/bin/python3
# Demo of single char terminal input in raw mode with the curses package.
import sys, curses

def run_one_char(dummy):
    'Run until a carriage return is entered'
    char = ' '
    print('Welcome to curses', flush=True)
    while ord(char) != 13:
        char = one_char()

def one_char():
    'Read one character from the keyboard'
    print('\r? ', flush= True, end = '')

    ## A blocking single char read in raw mode. 
    char = sys.stdin.read(1)
    print('You entered %s\r' % char)
    return char

## Must init curses before calling any functions
curses.initscr()
## To make sure the terminal returns to its initial settings,
## and to set raw mode and guarantee cleanup on exit. 
curses.wrapper(run_one_char)
print('Curses be gone!')

sys.stdin.read(1)

基本上会从STDIN读取1个字节。

如果你必须使用不等待\n的方法，你可以按照之前的回答使用这段代码:

class _Getch:
    """Gets a single character from standard input.  Does not echo to the screen."""
    def __init__(self):
        try:
            self.impl = _GetchWindows()
        except ImportError:
            self.impl = _GetchUnix()

    def __call__(self): return self.impl()


class _GetchUnix:
    def __init__(self):
        import tty, sys

    def __call__(self):
        import sys, tty, termios
        fd = sys.stdin.fileno()
        old_settings = termios.tcgetattr(fd)
        try:
            tty.setraw(sys.stdin.fileno())
            ch = sys.stdin.read(1)
        finally:
            termios.tcsetattr(fd, termios.TCSADRAIN, old_settings)
        return ch


class _GetchWindows:
    def __init__(self):
        import msvcrt

    def __call__(self):
        import msvcrt
        return msvcrt.getch()


getch = _Getch()

(摘自http://code.activestate.com/recipes/134892/)

试试这个:http://home.wlu.edu/~levys/software/kbhit.py 它是非阻塞的(这意味着您可以有一个while循环并检测按键而无需停止它)和跨平台的。

import os

# Windows
if os.name == 'nt':
    import msvcrt

# Posix (Linux, OS X)
else:
    import sys
    import termios
    import atexit
    from select import select


class KBHit:

    def __init__(self):
        '''Creates a KBHit object that you can call to do various keyboard things.'''

        if os.name == 'nt':
            pass

        else:

            # Save the terminal settings
            self.fd = sys.stdin.fileno()
            self.new_term = termios.tcgetattr(self.fd)
            self.old_term = termios.tcgetattr(self.fd)

            # New terminal setting unbuffered
            self.new_term[3] = (self.new_term[3] & ~termios.ICANON & ~termios.ECHO)
            termios.tcsetattr(self.fd, termios.TCSAFLUSH, self.new_term)

            # Support normal-terminal reset at exit
            atexit.register(self.set_normal_term)


    def set_normal_term(self):
        ''' Resets to normal terminal.  On Windows this is a no-op.
        '''

        if os.name == 'nt':
            pass

        else:
            termios.tcsetattr(self.fd, termios.TCSAFLUSH, self.old_term)


    def getch(self):
        ''' Returns a keyboard character after kbhit() has been called.
            Should not be called in the same program as getarrow().
        '''

        s = ''

        if os.name == 'nt':
            return msvcrt.getch().decode('utf-8')

        else:
            return sys.stdin.read(1)


    def getarrow(self):
        ''' Returns an arrow-key code after kbhit() has been called. Codes are
        0 : up
        1 : right
        2 : down
        3 : left
        Should not be called in the same program as getch().
        '''

        if os.name == 'nt':
            msvcrt.getch() # skip 0xE0
            c = msvcrt.getch()
            vals = [72, 77, 80, 75]

        else:
            c = sys.stdin.read(3)[2]
            vals = [65, 67, 66, 68]

        return vals.index(ord(c.decode('utf-8')))


    def kbhit(self):
        ''' Returns True if keyboard character was hit, False otherwise.
        '''
        if os.name == 'nt':
            return msvcrt.kbhit()

        else:
            dr,dw,de = select([sys.stdin], [], [], 0)
            return dr != []

使用这个的例子:

import kbhit

kb = kbhit.KBHit()

while(True): 
    print("Key not pressed") #Do something
    if kb.kbhit(): #If a key is pressed:
        k_in = kb.getch() #Detect what key was pressed
        print("You pressed ", k_in, "!") #Do something
kb.set_normal_term()

或者您可以使用PyPi中的getch模块。但是这会阻塞while循环