Question

我希望能够在恰好是BehaviorSubject的流上使用重播运算符。本质上，我希望印刷版订阅接收预先发送的排放：0、1、2、3、4、5。

我试图用.subscribe()使可观察的热点。另外，我确实知道将来源设为ReplaySubject可以解决问题，但这对我来说不是一个选择。

与大多数语言不同，Python的rx v3.0使用管道命令来链接操作（例如replay()或ref_count()或publish()），而不是通常的'。链。这是管道命令的链接：https://rxpy.readthedocs.io/en/latest/migration.html#pipe-based-operator-chaining

我非常确定问题与我给replay（）的第一个参数有关：lambda x: x。

import rx.subject
from rx import operators as op

stream = rx.subject.BehaviorSubject(0)
replayable_observable = stream.pipe(op.replay(lambda x: x, buffer_size=100))
replayable_observable.subscribe()
stream.subscribe()
for x in [1, 2, 3, 4, 5]:
    stream.on_next(x)
replayable_observable.subscribe(lambda value: print("Received {0}".format(value)))
for x in [6, 7, 8, 9, 10]:
    stream.on_next(x)

我希望收到0-10；或1-10。但是相反，我收到了5-10。

Received 5
Received 6
Received 7
Received 8
Received 9
Received 10

Answer 1

似乎replay的第一个参数（mapper函数）旨在允许您在多播后将更多运算符链接到可观察的源上。例如，如果将其从lambda x: x更改为lambda x: x.pipe(op.map(lambda y: y * 2))，您将获得加倍的值。

replay的{{3}}似乎已过时，因为他们给出的示例mapper函数仍然使用旧方法链而不是管道方法。此外，在replay参数中实际上使用非默认值的所有mapper的{{3}}在该项目的GitHub存储库中似乎已被注释掉，因此没有明确的示例如何正确使用此参数。

通过查看源代码，我可以说的是，当您使用replay而不指定mapper时，您将得到一个ConnectableObservable而不是Observable 。在将值正确推入主题之前连接此ConnectableObservable可以缓冲结果。

import rx.subject
from rx import operators as op

stream = rx.subject.BehaviorSubject(0)
replayable_observable = stream.pipe(op.replay(buffer_size=100))
replayable_observable.connect()
for x in [1, 2, 3, 4, 5]:
    stream.on_next(x)
replayable_observable.subscribe(lambda value: print("Received {0}".format(value)))
for x in [6, 7, 8, 9, 10]:
    stream.on_next(x)

# Received 0
# Received 1
# Received 2
# Received 3
# Received 4
# Received 5
# Received 6
# Received 7
# Received 8
# Received 9
# Received 10

如何使用python rx的重播？

1 个答案: