Nonblocking Stream Queue
This is a simple python package for nonblocking reading from streams.
It could be expanded for general nonblocking usage if desired.
A simple nonblocking queue reader is provided, with implementation from https://stackoverflow.com/questions/375427/a-non-blocking-read-on-a-subprocess-pipe-in-python/4896288#4896288 .
When constructed, the reader spawns a thread and begins reading everything from the stream.
Installation
$ python3 -m pip install nonblocking-stream-queue
Usage
import sys
import nonblocking_stream_queue as nonblocking
reader = nonblocking.Reader(
sys.stdin, # object to read from
max_size=4096, # max size of each read
lines=False, # whether or not to break reads into lines
max_count=None, # max queued reads
drop_timeout=None, # time to wait for queue to drain before dropping
drop_older=False, # which end of the queue to drop from
pre_cb=None, # if set, data = (pre_cb(), read())
post_cb=None, # if set, data = post_cb(data)
drop_cb=None, # if set, call with dropped data
verbose=False, # if set, display progress filling max_count
)
print(reader.read_one()) # outputs up to 4096 characters, or None if nothing queued
print(reader.read_many()) # outputs all 4096 character chunks queued
print(''.join(reader.read_many())) # outputs all queued text
reader.block() # waits for data to be present or end, returns number of reads queued
reader.is_pumping() # False if data has ended
Timestamping
import sys, time
import nonblocking_stream_queue as nonblocking
reader = nonblocking.Reader(
sys.stdin.buffer,
pre_cb=lambda: time.time(),
post_cb=lambda time_data_tuple: (time_data_tuple[0], time_data_tuple[1], time.time())
)
while reader.block():
start_time, data, end_time = reader.read_one()
print(start_time, data, end_time)
Lines
import sys
import nonblocking_stream_queue as nonblocking
reader = nonblocking.Reader(
sys.stdin,
lines=True
)
while reader.block():
print(reader.read_one().rstrip()) # outputs each line of text that is input
Other solutions
There are likely many other existing solutions to this common task.
Here's one:
