Repository navigation
Expand file tree
/
Copy pathstreams.py
More file actions
746 lines (610 loc) · 26 KB
/
Copy pathstreams.py
File metadata and controls
746 lines (610 loc) · 26 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
__all__ = (
'StreamReader', 'StreamWriter', 'StreamReaderProtocol',
'open_connection', 'start_server')
import socket
import sys
import weakref
if hasattr(socket, 'AF_UNIX'):
__all__ += ('open_unix_connection', 'start_unix_server')
from . import coroutines
from . import events
from . import exceptions
from . import format_helpers
from . import protocols
from .log import logger
from .tasks import sleep
_DEFAULT_LIMIT = 2 ** 16 # 64 KiB
async def open_connection(host=None, port=None, *,
limit=_DEFAULT_LIMIT, **kwds):
"""A wrapper for create_connection() returning a (reader, writer) pair.
The reader returned is a StreamReader instance; the writer is a
StreamWriter instance.
The arguments are all the usual arguments to create_connection()
except protocol_factory; most common are positional host and port,
with various optional keyword arguments following.
Additional optional keyword arguments are loop (to set the event loop
instance to use) and limit (to set the buffer limit passed to the
StreamReader).
(If you want to customize the StreamReader and/or
StreamReaderProtocol classes, just copy the code -- there's
really nothing special here except some convenience.)
"""
loop = events.get_running_loop()
reader = StreamReader(limit=limit, loop=loop)
protocol = StreamReaderProtocol(reader, loop=loop)
transport, _ = await loop.create_connection(
lambda: protocol, host, port, **kwds)
writer = StreamWriter(transport, protocol, reader, loop)
return reader, writer
async def start_server(client_connected_cb, host=None, port=None, *,
limit=_DEFAULT_LIMIT, **kwds):
"""Start a socket server, call back for each client connected.
The first parameter, `client_connected_cb`, takes two parameters:
client_reader, client_writer. client_reader is a StreamReader
object, while client_writer is a StreamWriter object. This
parameter can either be a plain callback function or a coroutine;
if it is a coroutine, it will be automatically converted into a
Task.
The rest of the arguments are all the usual arguments to
loop.create_server() except protocol_factory; most common are
positional host and port, with various optional keyword arguments
following. The return value is the same as loop.create_server().
Additional optional keyword arguments are loop (to set the event loop
instance to use) and limit (to set the buffer limit passed to the
StreamReader).
The return value is the same as loop.create_server(), i.e. a
Server object which can be used to stop the service.
"""
loop = events.get_running_loop()
def factory():
reader = StreamReader(limit=limit, loop=loop)
protocol = StreamReaderProtocol(reader, client_connected_cb,
loop=loop)
return protocol
return await loop.create_server(factory, host, port, **kwds)
if hasattr(socket, 'AF_UNIX'):
# UNIX Domain Sockets are supported on this platform
async def open_unix_connection(path=None, *,
limit=_DEFAULT_LIMIT, **kwds):
"""Similar to `open_connection` but works with UNIX Domain Sockets."""
loop = events.get_running_loop()
reader = StreamReader(limit=limit, loop=loop)
protocol = StreamReaderProtocol(reader, loop=loop)
transport, _ = await loop.create_unix_connection(
lambda: protocol, path, **kwds)
writer = StreamWriter(transport, protocol, reader, loop)
return reader, writer
async def start_unix_server(client_connected_cb, path=None, *,
limit=_DEFAULT_LIMIT, **kwds):
"""Similar to `start_server` but works with UNIX Domain Sockets."""
loop = events.get_running_loop()
def factory():
reader = StreamReader(limit=limit, loop=loop)
protocol = StreamReaderProtocol(reader, client_connected_cb,
loop=loop)
return protocol
return await loop.create_unix_server(factory, path, **kwds)
class FlowControlMixin(protocols.Protocol):
"""Reusable flow control logic for StreamWriter.drain().
This implements the protocol methods pause_writing(),
resume_writing() and connection_lost(). If the subclass overrides
these it must call the super methods.
StreamWriter.drain() must wait for _drain_helper() coroutine.
"""
def __init__(self, loop=None):
if loop is None:
self._loop = events._get_event_loop(stacklevel=4)
else:
self._loop = loop
self._paused = False
self._drain_waiter = None
self._connection_lost = False
def pause_writing(self):
assert not self._paused
self._paused = True
if self._loop.get_debug():
logger.debug("%r pauses writing", self)
def resume_writing(self):
assert self._paused
self._paused = False
if self._loop.get_debug():
logger.debug("%r resumes writing", self)
waiter = self._drain_waiter
if waiter is not None:
self._drain_waiter = None
if not waiter.done():
waiter.set_result(None)
def connection_lost(self, exc):
self._connection_lost = True
# Wake up the writer if currently paused.
if not self._paused:
return
waiter = self._drain_waiter
if waiter is None:
return
self._drain_waiter = None
if waiter.done():
return
if exc is None:
waiter.set_result(None)
else:
waiter.set_exception(exc)
async def _drain_helper(self):
if self._connection_lost:
raise ConnectionResetError('Connection lost')
if not self._paused:
return
waiter = self._drain_waiter
assert waiter is None or waiter.cancelled()
waiter = self._loop.create_future()
self._drain_waiter = waiter
await waiter
def _get_close_waiter(self, stream):
raise NotImplementedError
class StreamReaderProtocol(FlowControlMixin, protocols.Protocol):
"""Helper class to adapt between Protocol and StreamReader.
(This is a helper class instead of making StreamReader itself a
Protocol subclass, because the StreamReader has other potential
uses, and to prevent the user of the StreamReader to accidentally
call inappropriate methods of the protocol.)
"""
_source_traceback = None
def __init__(self, stream_reader, client_connected_cb=None, loop=None):
super().__init__(loop=loop)
if stream_reader is not None:
self._stream_reader_wr = weakref.ref(stream_reader)
self._source_traceback = stream_reader._source_traceback
else:
self._stream_reader_wr = None
if client_connected_cb is not None:
# This is a stream created by the `create_server()` function.
# Keep a strong reference to the reader until a connection
# is established.
self._strong_reader = stream_reader
self._reject_connection = False
self._stream_writer = None
self._transport = None
self._client_connected_cb = client_connected_cb
self._over_ssl = False
self._closed = self._loop.create_future()
@property
def _stream_reader(self):
if self._stream_reader_wr is None:
return None
return self._stream_reader_wr()
def _replace_writer(self, writer):
loop = self._loop
transport = writer.transport
self._stream_writer = writer
self._transport = transport
self._over_ssl = transport.get_extra_info('sslcontext') is not None
def connection_made(self, transport):
if self._reject_connection:
context = {
'message': ('An open stream was garbage collected prior to '
'establishing network connection; '
'call "stream.close()" explicitly.')
}
if self._source_traceback:
context['source_traceback'] = self._source_traceback
self._loop.call_exception_handler(context)
transport.abort()
return
self._transport = transport
reader = self._stream_reader
if reader is not None:
reader.set_transport(transport)
self._over_ssl = transport.get_extra_info('sslcontext') is not None
if self._client_connected_cb is not None:
self._stream_writer = StreamWriter(transport, self,
reader,
self._loop)
res = self._client_connected_cb(reader,
self._stream_writer)
if coroutines.iscoroutine(res):
self._loop.create_task(res)
self._strong_reader = None
def connection_lost(self, exc):
reader = self._stream_reader
if reader is not None:
if exc is None:
reader.feed_eof()
else:
reader.set_exception(exc)
if not self._closed.done():
if exc is None:
self._closed.set_result(None)
else:
self._closed.set_exception(exc)
super().connection_lost(exc)
self._stream_reader_wr = None
self._stream_writer = None
self._transport = None
def data_received(self, data):
reader = self._stream_reader
if reader is not None:
reader.feed_data(data)
def eof_received(self):
reader = self._stream_reader
if reader is not None:
reader.feed_eof()
if self._over_ssl:
# Prevent a warning in SSLProtocol.eof_received:
# "returning true from eof_received()
# has no effect when using ssl"
return False
return True
def _get_close_waiter(self, stream):
return self._closed
def __del__(self):
# Prevent reports about unhandled exceptions.
# Better than self._closed._log_traceback = False hack
try:
closed = self._closed
except AttributeError:
pass # failed constructor
else:
if closed.done() and not closed.cancelled():
closed.exception()
class StreamWriter:
"""Wraps a Transport.
This exposes write(), writelines(), [can_]write_eof(),
get_extra_info() and close(). It adds drain() which returns an
optional Future on which you can wait for flow control. It also
adds a transport property which references the Transport
directly.
"""
def __init__(self, transport, protocol, reader, loop):
self._transport = transport
self._protocol = protocol
# drain() expects that the reader has an exception() method
assert reader is None or isinstance(reader, StreamReader)
self._reader = reader
self._loop = loop
self._complete_fut = self._loop.create_future()
self._complete_fut.set_result(None)
def __repr__(self):
info = [self.__class__.__name__, f'transport={self._transport!r}']
if self._reader is not None:
info.append(f'reader={self._reader!r}')
return '<{}>'.format(' '.join(info))
@property
def transport(self):
return self._transport
def write(self, data):
self._transport.write(data)
def writelines(self, data):
self._transport.writelines(data)
def write_eof(self):
return self._transport.write_eof()
def can_write_eof(self):
return self._transport.can_write_eof()
def close(self):
return self._transport.close()
def is_closing(self):
return self._transport.is_closing()
async def wait_closed(self):
await self._protocol._get_close_waiter(self)
def get_extra_info(self, name, default=None):
return self._transport.get_extra_info(name, default)
async def drain(self):
"""Flush the write buffer.
The intended use is to write
w.write(data)
await w.drain()
"""
if self._reader is not None:
exc = self._reader.exception()
if exc is not None:
raise exc
if self._transport.is_closing():
# Wait for protocol.connection_lost() call
# Raise connection closing error if any,
# ConnectionResetError otherwise
# Yield to the event loop so connection_lost() may be
# called. Without this, _drain_helper() would return
# immediately, and code that calls
# write(...); await drain()
# in a loop would never call connection_lost(), so it
# would not see an error when the socket is closed.
await sleep(0)
await self._protocol._drain_helper()
async def start_tls(self, sslcontext, *,
server_hostname=None,
ssl_handshake_timeout=None):
"""Upgrade an existing stream-based connection to TLS."""
server_side = self._protocol._client_connected_cb is not None
protocol = self._protocol
await self.drain()
new_transport = await self._loop.start_tls( # type: ignore
self._transport, protocol, sslcontext,
server_side=server_side, server_hostname=server_hostname,
ssl_handshake_timeout=ssl_handshake_timeout)
self._transport = new_transport
protocol._replace_writer(self)
class StreamReader:
_source_traceback = None
def __init__(self, limit=_DEFAULT_LIMIT, loop=None):
# The line length limit is a security feature;
# it also doubles as half the buffer limit.
if limit <= 0:
raise ValueError('Limit cannot be <= 0')
self._limit = limit
if loop is None:
self._loop = events._get_event_loop()
else:
self._loop = loop
self._buffer = bytearray()
self._eof = False # Whether we're done.
self._waiter = None # A future used by _wait_for_data()
self._exception = None
self._transport = None
self._paused = False
if self._loop.get_debug():
self._source_traceback = format_helpers.extract_stack(
sys._getframe(1))
def __repr__(self):
info = ['StreamReader']
if self._buffer:
info.append(f'{len(self._buffer)} bytes')
if self._eof:
info.append('eof')
if self._limit != _DEFAULT_LIMIT:
info.append(f'limit={self._limit}')
if self._waiter:
info.append(f'waiter={self._waiter!r}')
if self._exception:
info.append(f'exception={self._exception!r}')
if self._transport:
info.append(f'transport={self._transport!r}')
if self._paused:
info.append('paused')
return '<{}>'.format(' '.join(info))
def exception(self):
return self._exception
def set_exception(self, exc):
self._exception = exc
waiter = self._waiter
if waiter is not None:
self._waiter = None
if not waiter.cancelled():
waiter.set_exception(exc)
def _wakeup_waiter(self):
"""Wakeup read*() functions waiting for data or EOF."""
waiter = self._waiter
if waiter is not None:
self._waiter = None
if not waiter.cancelled():
waiter.set_result(None)
def set_transport(self, transport):
assert self._transport is None, 'Transport already set'
self._transport = transport
def _maybe_resume_transport(self):
if self._paused and len(self._buffer) <= self._limit:
self._paused = False
self._transport.resume_reading()
def feed_eof(self):
self._eof = True
self._wakeup_waiter()
def at_eof(self):
"""Return True if the buffer is empty and 'feed_eof' was called."""
return self._eof and not self._buffer
def feed_data(self, data):
assert not self._eof, 'feed_data after feed_eof'
if not data:
return
self._buffer.extend(data)
self._wakeup_waiter()
if (self._transport is not None and
not self._paused and
len(self._buffer) > 2 * self._limit):
try:
self._transport.pause_reading()
except NotImplementedError:
# The transport can't be paused.
# We'll just have to buffer all data.
# Forget the transport so we don't keep trying.
self._transport = None
else:
self._paused = True
async def _wait_for_data(self, func_name):
"""Wait until feed_data() or feed_eof() is called.
If stream was paused, automatically resume it.
"""
# StreamReader uses a future to link the protocol feed_data() method
# to a read coroutine. Running two read coroutines at the same time
# would have an unexpected behaviour. It would not possible to know
# which coroutine would get the next data.
if self._waiter is not None:
raise RuntimeError(
f'{func_name}() called while another coroutine is '
f'already waiting for incoming data')
assert not self._eof, '_wait_for_data after EOF'
# Waiting for data while paused will make deadlock, so prevent it.
# This is essential for readexactly(n) for case when n > self._limit.
if self._paused:
self._paused = False
self._transport.resume_reading()
self._waiter = self._loop.create_future()
try:
await self._waiter
finally:
self._waiter = None
async def readline(self):
"""Read chunk of data from the stream until newline (b'\n') is found.
On success, return chunk that ends with newline. If only partial
line can be read due to EOF, return incomplete line without
terminating newline. When EOF was reached while no bytes read, empty
bytes object is returned.
If limit is reached, ValueError will be raised. In that case, if
newline was found, complete line including newline will be removed
from internal buffer. Else, internal buffer will be cleared. Limit is
compared against part of the line without newline.
If stream was paused, this function will automatically resume it if
needed.
"""
sep = b'\n'
seplen = len(sep)
try:
line = await self.readuntil(sep)
except exceptions.IncompleteReadError as e:
return e.partial
except exceptions.LimitOverrunError as e:
if self._buffer.startswith(sep, e.consumed):
del self._buffer[:e.consumed + seplen]
else:
self._buffer.clear()
self._maybe_resume_transport()
raise ValueError(e.args[0])
return line
async def readuntil(self, separator=b'\n'):
"""Read data from the stream until ``separator`` is found.
On success, the data and separator will be removed from the
internal buffer (consumed). Returned data will include the
separator at the end.
Configured stream limit is used to check result. Limit sets the
maximal length of data that can be returned, not counting the
separator.
If an EOF occurs and the complete separator is still not found,
an IncompleteReadError exception will be raised, and the internal
buffer will be reset. The IncompleteReadError.partial attribute
may contain the separator partially.
If the data cannot be read because of over limit, a
LimitOverrunError exception will be raised, and the data
will be left in the internal buffer, so it can be read again.
"""
seplen = len(separator)
if seplen == 0:
raise ValueError('Separator should be at least one-byte string')
if self._exception is not None:
raise self._exception
# Consume whole buffer except last bytes, which length is
# one less than seplen. Let's check corner cases with
# separator='SEPARATOR':
# * we have received almost complete separator (without last
# byte). i.e buffer='some textSEPARATO'. In this case we
# can safely consume len(separator) - 1 bytes.
# * last byte of buffer is first byte of separator, i.e.
# buffer='abcdefghijklmnopqrS'. We may safely consume
# everything except that last byte, but this require to
# analyze bytes of buffer that match partial separator.
# This is slow and/or require FSM. For this case our
# implementation is not optimal, since require rescanning
# of data that is known to not belong to separator. In
# real world, separator will not be so long to notice
# performance problems. Even when reading MIME-encoded
# messages :)
# `offset` is the number of bytes from the beginning of the buffer
# where there is no occurrence of `separator`.
offset = 0
# Loop until we find `separator` in the buffer, exceed the buffer size,
# or an EOF has happened.
while True:
buflen = len(self._buffer)
# Check if we now have enough data in the buffer for `separator` to
# fit.
if buflen - offset >= seplen:
isep = self._buffer.find(separator, offset)
if isep != -1:
# `separator` is in the buffer. `isep` will be used later
# to retrieve the data.
break
# see upper comment for explanation.
offset = buflen + 1 - seplen
if offset > self._limit:
raise exceptions.LimitOverrunError(
'Separator is not found, and chunk exceed the limit',
offset)
# Complete message (with full separator) may be present in buffer
# even when EOF flag is set. This may happen when the last chunk
# adds data which makes separator be found. That's why we check for
# EOF *ater* inspecting the buffer.
if self._eof:
chunk = bytes(self._buffer)
self._buffer.clear()
raise exceptions.IncompleteReadError(chunk, None)
# _wait_for_data() will resume reading if stream was paused.
await self._wait_for_data('readuntil')
if isep > self._limit:
raise exceptions.LimitOverrunError(
'Separator is found, but chunk is longer than limit', isep)
chunk = self._buffer[:isep + seplen]
del self._buffer[:isep + seplen]
self._maybe_resume_transport()
return bytes(chunk)
async def read(self, n=-1):
"""Read up to `n` bytes from the stream.
If n is not provided, or set to -1, read until EOF and return all read
bytes. If the EOF was received and the internal buffer is empty, return
an empty bytes object.
If n is zero, return empty bytes object immediately.
If n is positive, this function try to read `n` bytes, and may return
less or equal bytes than requested, but at least one byte. If EOF was
received before any byte is read, this function returns empty byte
object.
Returned value is not limited with limit, configured at stream
creation.
If stream was paused, this function will automatically resume it if
needed.
"""
if self._exception is not None:
raise self._exception
if n == 0:
return b''
if n < 0:
# This used to just loop creating a new waiter hoping to
# collect everything in self._buffer, but that would
# deadlock if the subprocess sends more than self.limit
# bytes. So just call self.read(self._limit) until EOF.
blocks = []
while True:
block = await self.read(self._limit)
if not block:
break
blocks.append(block)
return b''.join(blocks)
if not self._buffer and not self._eof:
await self._wait_for_data('read')
# This will work right even if buffer is less than n bytes
data = bytes(self._buffer[:n])
del self._buffer[:n]
self._maybe_resume_transport()
return data
async def readexactly(self, n):
"""Read exactly `n` bytes.
Raise an IncompleteReadError if EOF is reached before `n` bytes can be
read. The IncompleteReadError.partial attribute of the exception will
contain the partial read bytes.
if n is zero, return empty bytes object.
Returned value is not limited with limit, configured at stream
creation.
If stream was paused, this function will automatically resume it if
needed.
"""
if n < 0:
raise ValueError('readexactly size can not be less than zero')
if self._exception is not None:
raise self._exception
if n == 0:
return b''
while len(self._buffer) < n:
if self._eof:
incomplete = bytes(self._buffer)
self._buffer.clear()
raise exceptions.IncompleteReadError(incomplete, n)
await self._wait_for_data('readexactly')
if len(self._buffer) == n:
data = bytes(self._buffer)
self._buffer.clear()
else:
data = bytes(self._buffer[:n])
del self._buffer[:n]
self._maybe_resume_transport()
return data
def __aiter__(self):
return self
async def __anext__(self):
val = await self.readline()
if val == b'':
raise StopAsyncIteration
return val