Python嵌套生成器模拟无限数据流时输出不符合预期问题排查

• 子生成器未被完整迭代：你在_get_data的循环逻辑中，每次都会创建新的_data_yielder实例，仅调用next()获取它第一次yield的返回值后就直接丢弃了该实例，没有继续迭代子生成器的后续内容，所以每个字母只会输出索引为0的结果。
• send值未向下传递：你对外层_get_data生成器调用send()方法时，传入的值是发送给外层生成器的，而原代码中_get_data没有任何接收send值并传递给子生成器的逻辑，所以_data_yielder里的bla变量永远拿不到你传入的值，自然不会有对应的打印输出。

使用Python的yield from语法处理子生成器即可，yield from会自动完成两层生成器的透传逻辑：一方面把子生成器yield的所有值直接返回给最外层的调用方，另一方面也会把外层调用方的send、throw等操作自动转发给子生成器，完全匹配你的使用需求。

仅需要修改_get_data函数的最后一行即可，修改后的完整代码如下：

def _data_yielder(v, iter_to):
    for i in range(iter_to):
        bla = yield (v, i)
        print(f"Result received from send: {bla}")

def _get_data():
    """
    generator that iterates some expensive-to-open files
    """
    for i, v in enumerate(["A", "B", "C"]):
        offset = i + 2
        print(f"{v} should go up to {(v, offset-1)}")
        # 替换原有的yield next逻辑为yield from
        yield from _data_yielder(v, offset)
        
def _reset():
    resetted_generator = _get_data()
    initial_value = next(resetted_generator)
    return initial_value, resetted_generator

count = 0
iter_data, data_generator_process = _reset()
print(f"\toutput value: {iter_data}")   
while True:
    try:
        iter_data = data_generator_process.send(f"test_{count}")
        count += 1
        print(f"\toutput value: {iter_data}")
    except StopIteration as e:
        print(e)
        break
    except Exception as e:
        print(e)

修改后运行即可得到你预期的输出，同时你send的test_{count}值也会被_data_yielder正确接收并打印，输出内容如下：

A should go up to ('A', 1)
	output value: ('A', 0)
Result received from send: test_0
	output value: ('A', 1)
B should go up to ('B', 2)
Result received from send: test_1
	output value: ('B', 0)
Result received from send: test_2
	output value: ('B', 1)
C should go up to ('C', 3)
Result received from send: test_3
	output value: ('C', 0)
Result received from send: test_4
	output value: ('C', 1)
Result received from send: test_5

内容的提问来源于stack exchange，提问作者IanQ