openstack/deb-python-autobahn
Code Issues Proposed changes
c103777dc2
deb-python-autobahn/autobahn/util.py

776 lines
24 KiB
Python
Raw Blame History

###############################################################################
#
# The MIT License (MIT)
#
# Copyright (c) Crossbar.io Technologies GmbH
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", fWITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.
#
###############################################################################
from __future__ import absolute_import
import os
import time
import struct
import sys
import re
import base64
import math
import random
import binascii
from datetime import datetime, timedelta
from pprint import pformat
from array import array
import six
import txaio
__all__ = ("xor",
"utcnow",
"utcstr",
"id",
"rid",
"newid",
"rtime",
"Stopwatch",
"Tracker",
"EqualityMixin",
"ObservableMixin",
"IdGenerator",
"generate_token",
"generate_activation_code",
"generate_serial_number",
"generate_user_password")
def encode_truncate(text, limit, encoding='utf8', return_encoded=True):
"""
Given a string, return a truncated version of the string such that
the UTF8 encoding of the string is smaller than the given limit.
This function correctly truncates even in the presence of Unicode code
points that encode to multi-byte encodings which must not be truncated
in the middle.
:param text: The Unicode string to truncate.
:type text: unicode
:param limit: The number of bytes to limit the UTF8 encoding to.
:type limit: int
:returns: The truncated Unicode string.
:rtype: unicode
"""
assert(text is None or type(text) == six.text_type)
assert(type(limit) in six.integer_types)
assert(limit >= 0)
if text is None:
return
# encode the given string in the specified encoding
s = text.encode(encoding)
# when the resulting byte string is longer than the given limit ..
if len(s) > limit:
# .. truncate, and
s = s[:limit]
# decode back, ignoring errors that result from truncation
# in the middle of multi-byte encodings
text = s.decode(encoding, 'ignore')
if return_encoded:
s = text.encode(encoding)
if return_encoded:
return s
else:
return text
def xor(d1, d2):
"""
XOR two binary strings of arbitrary (equal) length.
:param d1: The first binary string.
:type d1: binary
:param d2: The second binary string.
:type d2: binary
:returns: XOR(d1, d2)
:rtype: binary
"""
if type(d1) != six.binary_type:
raise Exception("invalid type {} for d1 - must be binary".format(type(d1)))
if type(d2) != six.binary_type:
raise Exception("invalid type {} for d2 - must be binary".format(type(d2)))
if len(d1) != len(d2):
raise Exception("cannot XOR binary string of differing length ({} != {})".format(len(d1), len(d2)))
d1 = array('B', d1)
d2 = array('B', d2)
for i in range(len(d1)):
d1[i] ^= d2[i]
if six.PY3:
return d1.tobytes()
else:
return d1.tostring()
def utcstr(ts=None):
"""
Format UTC timestamp in ISO 8601 format.
Note: to parse an ISO 8601 formatted string, use the **iso8601**
module instead (e.g. ``iso8601.parse_date("2014-05-23T13:03:44.123Z")``).
:param ts: The timestamp to format.
:type ts: instance of :py:class:`datetime.datetime` or None
:returns: Timestamp formatted in ISO 8601 format.
:rtype: unicode
"""
assert(ts is None or isinstance(ts, datetime))
if ts is None:
ts = datetime.utcnow()
return u"{0}Z".format(ts.strftime(u"%Y-%m-%dT%H:%M:%S.%f")[:-3])
def utcnow():
"""
Get current time in UTC as ISO 8601 string.
:returns: Current time as string in ISO 8601 format.
:rtype: unicode
"""
return utcstr()
class IdGenerator(object):
"""
ID generator for WAMP request IDs.
WAMP request IDs are sequential per WAMP session, starting at 1 and
wrapping around at 2**53 (both value are inclusive [1, 2**53]).
The upper bound **2**53** is chosen since it is the maximum integer that can be
represented as a IEEE double such that all smaller integers are representable as well.
Hence, IDs can be safely used with languages that use IEEE double as their
main (or only) number type (JavaScript, Lua, etc).
See https://github.com/wamp-proto/wamp-proto/blob/master/spec/basic.md#ids
"""
def __init__(self):
self._next = 0 # starts at 1; next() pre-increments
def next(self):
"""
Returns next ID.
:returns: The next ID.
:rtype: int
"""
self._next += 1
if self._next > 9007199254740992:
self._next = 1
return self._next
# generator protocol
def __next__(self):
return self.next()
#
# Performance comparison of IdGenerator.next(), id() and rid().
#
# All tests were performed on:
#
# - Ubuntu 14.04 LTS x86-64
# - Intel Core i7 920 @ 3.3GHz
#
# The tests generated 100 mio. IDs and run-time was measured
# as wallclock from Unix "time" command. In each run, a single CPU
# core was essentially at 100% load all the time (though the sys/usr
# ratio was different).
#
# PyPy 2.6.1:
#
# IdGenerator.next() 0.5s
# id() 29.4s
# rid() 106.1s
#
# CPython 2.7.10:
#
# IdGenerator.next() 49.0s
# id() 370.5s
# rid() 196.4s
#
#
# Note on the ID range [0, 2**53]. We once reduced the range to [0, 2**31].
# This lead to extremely hard to track down issues due to ID collisions!
# Here: https://github.com/crossbario/autobahn-python/issues/419#issue-90483337
#
# 8 byte mask with 53 LSBs set (WAMP requires IDs from [0, 2**53]
_WAMP_ID_MASK = struct.unpack(">Q", b"\x00\x1f\xff\xff\xff\xff\xff\xff")[0]
def rid():
"""
Generate a new random integer ID from range **[0, 2**53]**.
The generated ID is uniformly distributed over the whole range, doesn't have
a period (no pseudo-random generator is used) and cryptographically strong.
The upper bound **2**53** is chosen since it is the maximum integer that can be
represented as a IEEE double such that all smaller integers are representable as well.
Hence, IDs can be safely used with languages that use IEEE double as their
main (or only) number type (JavaScript, Lua, etc).
:returns: A random integer ID.
:rtype: int
"""
return struct.unpack("@Q", os.urandom(8))[0] & _WAMP_ID_MASK
# noinspection PyShadowingBuiltins
def id():
"""
Generate a new random integer ID from range **[0, 2**53]**.
The generated ID is based on a pseudo-random number generator (Mersenne Twister,
which has a period of 2**19937-1). It is NOT cryptographically strong, and
hence NOT suitable to generate e.g. secret keys or access tokens.
The upper bound **2**53** is chosen since it is the maximum integer that can be
represented as a IEEE double such that all smaller integers are representable as well.
Hence, IDs can be safely used with languages that use IEEE double as their
main (or only) number type (JavaScript, Lua, etc).
:returns: A random integer ID.
:rtype: int
"""
return random.randint(0, 9007199254740992)
def newid(length=16):
"""
Generate a new random string ID.
The generated ID is uniformly distributed and cryptographically strong. It is
hence usable for things like secret keys and access tokens.
:param length: The length (in chars) of the ID to generate.
:type length: int
:returns: A random string ID.
:rtype: unicode
"""
l = int(math.ceil(float(length) * 6. / 8.))
return base64.b64encode(os.urandom(l))[:length].decode('ascii')
# a standard base36 character set
# DEFAULT_TOKEN_CHARS = string.digits + string.ascii_uppercase
# we take out the following 9 chars (leaving 27), because there
# is visual ambiguity: 0/O/D, 1/I, 8/B, 2/Z
DEFAULT_TOKEN_CHARS = u'345679ACEFGHJKLMNPQRSTUVWXY'
"""
Default set of characters to create rtokens from.
"""
DEFAULT_ZBASE32_CHARS = u'13456789abcdefghijkmnopqrstuwxyz'
"""
http://philzimmermann.com/docs/human-oriented-base-32-encoding.txt
Our choice of confusing characters to eliminate is: `0', `l', `v', and `2'. Our
reasoning is that `0' is potentially mistaken for `o', that `l' is potentially
mistaken for `1' or `i', that `v' is potentially mistaken for `u' or `r'
(especially in handwriting) and that `2' is potentially mistaken for `z'
(especially in handwriting).
Note that we choose to focus on typed and written transcription more than on
vocal, since humans already have a well-established system of disambiguating
spoken alphanumerics, such as the United States military's "Alpha Bravo Charlie
Delta" and telephone operators' "Is that 'd' as in 'dog'?".
"""
def generate_token(char_groups, chars_per_group, chars=None, sep=None, lower_case=False):
"""
Generate cryptographically strong tokens, which are strings like `M6X5-YO5W-T5IK`.
These can be used e.g. for used-only-once activation tokens or the like.
The returned token has an entropy of:
math.log(len(chars), 2.) * chars_per_group * char_groups
bits. With the default charset and 4 characters per group, rtoken() produces
tokens with the following entropy:
character groups entropy (at least) recommended use
2 38 bits
3 57 bits one-time activation or pairing code
4 76 bits secure user password
5 95 bits
6 114 bits globally unique serial / product code
7 133 bits
Here are 3 examples:
* token(3): 9QXT-UXJW-7R4H
* token(4): LPNN-JMET-KWEP-YK45
* token(6): NXW9-74LU-6NUH-VLPV-X6AG-QUE3
:param char_groups: Number of character groups (or characters if chars_per_group == 1).
:type char_groups: int
:param chars_per_group: Number of characters per character group (or 1 to return a token with no grouping).
:type chars_per_group: int
:param chars: Characters to choose from. Default is 27 character subset
of the ISO basic Latin alphabet (see: DEFAULT_TOKEN_CHARS).
:type chars: unicode or None
:param sep: When separating groups in the token, the separater string.
:type sep: unicode
:returns: The generated token.
:rtype: unicode
"""
assert(type(char_groups) in six.integer_types)
assert(type(chars_per_group) in six.integer_types)
assert(chars is None or type(chars) == six.text_type)
chars = chars or DEFAULT_TOKEN_CHARS
if lower_case:
chars = chars.lower()
sep = sep or u'-'
rng = random.SystemRandom()
token_value = u''.join(rng.choice(chars) for _ in range(char_groups * chars_per_group))
if chars_per_group > 1:
return sep.join(map(u''.join, zip(*[iter(token_value)] * chars_per_group)))
else:
return token_value
def generate_activation_code():
return generate_token(char_groups=3, chars_per_group=4, chars=DEFAULT_TOKEN_CHARS, sep=u'-', lower_case=False)
def generate_user_password():
return generate_token(char_groups=16, chars_per_group=1, chars=DEFAULT_ZBASE32_CHARS, sep=u'-', lower_case=True)
def generate_serial_number():
return generate_token(char_groups=6, chars_per_group=4, chars=DEFAULT_TOKEN_CHARS, sep=u'-', lower_case=False)
# Select the most precise walltime measurement function available
# on the platform
#
if sys.platform.startswith('win'):
# On Windows, this function returns wall-clock seconds elapsed since the
# first call to this function, as a floating point number, based on the
# Win32 function QueryPerformanceCounter(). The resolution is typically
# better than one microsecond
_rtime = time.clock
_ = _rtime() # this starts wallclock
else:
# On Unix-like platforms, this used the first available from this list:
# (1) gettimeofday() -- resolution in microseconds
# (2) ftime() -- resolution in milliseconds
# (3) time() -- resolution in seconds
_rtime = time.time
rtime = _rtime
"""
Precise wallclock time.
:returns: The current wallclock in seconds. Returned values are only guaranteed
to be meaningful relative to each other.
:rtype: float
"""
class Stopwatch(object):
"""
Stopwatch based on walltime.
This can be used to do code timing and uses the most precise walltime measurement
available on the platform. This is a very light-weight object,
so create/dispose is very cheap.
"""
def __init__(self, start=True):
"""
:param start: If ``True``, immediately start the stopwatch.
:type start: bool
"""
self._elapsed = 0
if start:
self._started = rtime()
self._running = True
else:
self._started = None
self._running = False
def elapsed(self):
"""
Return total time elapsed in seconds during which the stopwatch was running.
:returns: The elapsed time in seconds.
:rtype: float
"""
if self._running:
now = rtime()
return self._elapsed + (now - self._started)
else:
return self._elapsed
def pause(self):
"""
Pauses the stopwatch and returns total time elapsed in seconds during which
the stopwatch was running.
:returns: The elapsed time in seconds.
:rtype: float
"""
if self._running:
now = rtime()
self._elapsed += now - self._started
self._running = False
return self._elapsed
else:
return self._elapsed
def resume(self):
"""
Resumes a paused stopwatch and returns total elapsed time in seconds
during which the stopwatch was running.
:returns: The elapsed time in seconds.
:rtype: float
"""
if not self._running:
self._started = rtime()
self._running = True
return self._elapsed
else:
now = rtime()
return self._elapsed + (now - self._started)
def stop(self):
"""
Stops the stopwatch and returns total time elapsed in seconds during which
the stopwatch was (previously) running.
:returns: The elapsed time in seconds.
:rtype: float
"""
elapsed = self.pause()
self._elapsed = 0
self._started = None
self._running = False
return elapsed
class Tracker(object):
"""
A key-based statistics tracker.
"""
def __init__(self, tracker, tracked):
"""
"""
self.tracker = tracker
self.tracked = tracked
self._timings = {}
self._offset = rtime()
self._dt_offset = datetime.utcnow()
def track(self, key):
"""
Track elapsed for key.
:param key: Key under which to track the timing.
:type key: str
"""
self._timings[key] = rtime()
def diff(self, start_key, end_key, formatted=True):
"""
Get elapsed difference between two previously tracked keys.
:param start_key: First key for interval (older timestamp).
:type start_key: str
:param end_key: Second key for interval (younger timestamp).
:type end_key: str
:param formatted: If ``True``, format computed time period and return string.
:type formatted: bool
:returns: Computed time period in seconds (or formatted string).
:rtype: float or str
"""
if end_key in self._timings and start_key in self._timings:
d = self._timings[end_key] - self._timings[start_key]
if formatted:
if d < 0.00001: # 10us
s = "%d ns" % round(d * 1000000000.)
elif d < 0.01: # 10ms
s = "%d us" % round(d * 1000000.)
elif d < 10: # 10s
s = "%d ms" % round(d * 1000.)
else:
s = "%d s" % round(d)
return s.rjust(8)
else:
return d
else:
if formatted:
return "n.a.".rjust(8)
else:
return None
def absolute(self, key):
"""
Return the UTC wall-clock time at which a tracked event occurred.
:param key: The key
:type key: str
:returns: Timezone-naive datetime.
:rtype: instance of :py:class:`datetime.datetime`
"""
elapsed = self[key]
if elapsed is None:
raise KeyError("No such key \"%s\"." % elapsed)
return self._dt_offset + timedelta(seconds=elapsed)
def __getitem__(self, key):
if key in self._timings:
return self._timings[key] - self._offset
else:
return None
def __iter__(self):
return self._timings.__iter__()
def __str__(self):
return pformat(self._timings)
class EqualityMixin(object):
"""
Mixing to add equality comparison operators to a class.
Two objects are identical under this mixin, if and only if:
1. both object have the same class
2. all non-private object attributes are equal
"""
def __eq__(self, other):
"""
Compare this object to another object for equality.
:param other: The other object to compare with.
:type other: obj
:returns: ``True`` iff the objects are equal.
:rtype: bool
"""
if not isinstance(other, self.__class__):
return False
# we only want the actual message data attributes (not eg _serialize)
for k in self.__dict__:
if not k.startswith('_'):
if not self.__dict__[k] == other.__dict__[k]:
return False
return True
# return (isinstance(other, self.__class__) and self.__dict__ == other.__dict__)
def __ne__(self, other):
"""
Compare this object to another object for inequality.
:param other: The other object to compare with.
:type other: obj
:returns: ``True`` iff the objects are not equal.
:rtype: bool
"""
return not self.__eq__(other)
def wildcards2patterns(wildcards):
"""
Compute a list of regular expression patterns from a list of
wildcard strings. A wildcard string uses '*' as a wildcard character
matching anything.
:param wildcards: List of wildcard strings to compute regular expression patterns for.
:type wildcards: list of str
:returns: Computed regular expressions.
:rtype: list of obj
"""
# note that we add the ^ and $ so that the *entire* string must
# match. Without this, e.g. a prefix will match:
# re.match('.*good\\.com', 'good.com.evil.com') # match!
# re.match('.*good\\.com$', 'good.com.evil.com') # no match!
return [re.compile('^' + wc.replace('.', '\.').replace('*', '.*') + '$') for wc in wildcards]
class ObservableMixin(object):
"""
Internal utility for enabling event-listeners on particular objects
"""
# A "helper" style composable class (as opposed to a mix-in) might
# be a lot easier to deal with here. Having an __init__ method
# with a "mix in" style class can be fragile and error-prone,
# especially if it takes arguments. Since we don't use the
# "parent" beavior anywhere, I didn't add a .set_parent() (yet?)
# these are class-level globals; individual instances are
# initialized as-needed (e.g. the first .on() call adds a
# _listeners dict). Thus, subclasses don't have to call super()
# properly etc.
_parent = None
_valid_events = None
_listeners = None
def set_valid_events(self, valid_events=None):
"""
:param valid_events: if non-None, .on() or .fire() with an event
not listed in valid_events raises an exception.
"""
self._valid_events = list(valid_events)
def _check_event(self, event):
"""
Internal helper. Throws RuntimeError if we have a valid_events
list, and the given event isnt' in it. Does nothing otherwise.
"""
if self._valid_events and event not in self._valid_events:
raise RuntimeError(
"Invalid event '{event}'. Expected one of: {events}",
event=event,
events=', '.join(self._valid_events),
)
def on(self, event, handler):
"""
Add a handler for an event.
:param event: the name of the event
:param handler: a callable thats invoked when .fire() is
called for this events. Arguments will be whatever are given
to .fire()
"""
# print("adding '{}' to '{}': {}".format(event, hash(self), handler))
self._check_event(event)
if self._listeners is None:
self._listeners = dict()
if event not in self._listeners:
self._listeners[event] = []
self._listeners[event].append(handler)
def off(self, event=None, handler=None):
"""
Stop listening for a single event, or all events.
:param event: if None, remove all listeners. Otherwise, remove
listeners for the single named event.
:param handler: if None, remove all handlers for the named
event; otherwise remove just the given handler.
"""
if event is None:
if handler is not None:
# maybe this should mean "remove the given handler
# from any event at all that contains it"...?
raise RuntimeError(
"Can't specificy a specific handler without an event"
)
self._listeners = dict()
else:
if self._listeners is None:
return
self._check_event(event)
if event in self._listeners:
if handler is None:
del self._listeners[event]
else:
self._listeners[event].discard(handler)
def fire(self, event, *args, **kwargs):
"""
Fire a particular event.
:param event: the event to fire. All other args and kwargs are
passed on to the handler(s) for the event.
:return: a Deferred/Future gathering all async results from
all handlers and/or parent handlers.
"""
# print("firing '{}' from '{}'".format(event, hash(self)))
if self._listeners is None:
return txaio.create_future(result=[])
self._check_event(event)
res = []
for handler in self._listeners.get(event, []):
future = txaio.as_future(handler, *args, **kwargs)
res.append(future)
if self._parent is not None:
res.append(self._parent.fire(event, *args, **kwargs))
return txaio.gather(res, consume_exceptions=False)
class _LazyHexFormatter(object):
"""
This is used to avoid calling binascii.hexlify() on data given to
log.debug() calls unless debug is active (for example). Like::
self.log.debug(
"Some data: {octets}",
octets=_LazyHexFormatter(os.urandom(32)),
)
"""
__slots__ = ('obj',)
def __init__(self, obj):
self.obj = obj
def __str__(self):
return binascii.hexlify(self.obj).decode('ascii')
View Git Blame Copy Permalink