mirror of
https://github.com/morpheus65535/bazarr
synced 2024-12-26 09:37:25 +00:00
1360 lines
49 KiB
Python
1360 lines
49 KiB
Python
# -*- coding:iso-8859-1 -*-
|
|
"""
|
|
This module offers a generic date/time string parser which is able to parse
|
|
most known formats to represent a date and/or time.
|
|
|
|
This module attempts to be forgiving with regards to unlikely input formats,
|
|
returning a datetime object even for dates which are ambiguous. If an element
|
|
of a date/time stamp is omitted, the following rules are applied:
|
|
- If AM or PM is left unspecified, a 24-hour clock is assumed, however, an hour
|
|
on a 12-hour clock (``0 <= hour <= 12``) *must* be specified if AM or PM is
|
|
specified.
|
|
- If a time zone is omitted, a timezone-naive datetime is returned.
|
|
|
|
If any other elements are missing, they are taken from the
|
|
:class:`datetime.datetime` object passed to the parameter ``default``. If this
|
|
results in a day number exceeding the valid number of days per month, the
|
|
value falls back to the end of the month.
|
|
|
|
Additional resources about date/time string formats can be found below:
|
|
|
|
- `A summary of the international standard date and time notation
|
|
<http://www.cl.cam.ac.uk/~mgk25/iso-time.html>`_
|
|
- `W3C Date and Time Formats <http://www.w3.org/TR/NOTE-datetime>`_
|
|
- `Time Formats (Planetary Rings Node) <http://pds-rings.seti.org/tools/time_formats.html>`_
|
|
- `CPAN ParseDate module
|
|
<http://search.cpan.org/~muir/Time-modules-2013.0912/lib/Time/ParseDate.pm>`_
|
|
- `Java SimpleDateFormat Class
|
|
<https://docs.oracle.com/javase/6/docs/api/java/text/SimpleDateFormat.html>`_
|
|
"""
|
|
from __future__ import unicode_literals
|
|
|
|
import datetime
|
|
import string
|
|
import time
|
|
import collections
|
|
import re
|
|
from io import StringIO
|
|
from calendar import monthrange, isleap
|
|
|
|
from six import text_type, binary_type, integer_types
|
|
|
|
from . import relativedelta
|
|
from . import tz
|
|
|
|
__all__ = ["parse", "parserinfo"]
|
|
|
|
|
|
class _timelex(object):
|
|
# Fractional seconds are sometimes split by a comma
|
|
_split_decimal = re.compile("([\.,])")
|
|
|
|
def __init__(self, instream):
|
|
if isinstance(instream, binary_type):
|
|
instream = instream.decode()
|
|
|
|
if isinstance(instream, text_type):
|
|
instream = StringIO(instream)
|
|
|
|
if getattr(instream, 'read', None) is None:
|
|
raise TypeError('Parser must be a string or character stream, not '
|
|
'{itype}'.format(itype=instream.__class__.__name__))
|
|
|
|
self.instream = instream
|
|
self.charstack = []
|
|
self.tokenstack = []
|
|
self.eof = False
|
|
|
|
def get_token(self):
|
|
"""
|
|
This function breaks the time string into lexical units (tokens), which
|
|
can be parsed by the parser. Lexical units are demarcated by changes in
|
|
the character set, so any continuous string of letters is considered
|
|
one unit, any continuous string of numbers is considered one unit.
|
|
|
|
The main complication arises from the fact that dots ('.') can be used
|
|
both as separators (e.g. "Sep.20.2009") or decimal points (e.g.
|
|
"4:30:21.447"). As such, it is necessary to read the full context of
|
|
any dot-separated strings before breaking it into tokens; as such, this
|
|
function maintains a "token stack", for when the ambiguous context
|
|
demands that multiple tokens be parsed at once.
|
|
"""
|
|
if self.tokenstack:
|
|
return self.tokenstack.pop(0)
|
|
|
|
seenletters = False
|
|
token = None
|
|
state = None
|
|
|
|
while not self.eof:
|
|
# We only realize that we've reached the end of a token when we
|
|
# find a character that's not part of the current token - since
|
|
# that character may be part of the next token, it's stored in the
|
|
# charstack.
|
|
if self.charstack:
|
|
nextchar = self.charstack.pop(0)
|
|
else:
|
|
nextchar = self.instream.read(1)
|
|
while nextchar == '\x00':
|
|
nextchar = self.instream.read(1)
|
|
|
|
if not nextchar:
|
|
self.eof = True
|
|
break
|
|
elif not state:
|
|
# First character of the token - determines if we're starting
|
|
# to parse a word, a number or something else.
|
|
token = nextchar
|
|
if self.isword(nextchar):
|
|
state = 'a'
|
|
elif self.isnum(nextchar):
|
|
state = '0'
|
|
elif self.isspace(nextchar):
|
|
token = ' '
|
|
break # emit token
|
|
else:
|
|
break # emit token
|
|
elif state == 'a':
|
|
# If we've already started reading a word, we keep reading
|
|
# letters until we find something that's not part of a word.
|
|
seenletters = True
|
|
if self.isword(nextchar):
|
|
token += nextchar
|
|
elif nextchar == '.':
|
|
token += nextchar
|
|
state = 'a.'
|
|
else:
|
|
self.charstack.append(nextchar)
|
|
break # emit token
|
|
elif state == '0':
|
|
# If we've already started reading a number, we keep reading
|
|
# numbers until we find something that doesn't fit.
|
|
if self.isnum(nextchar):
|
|
token += nextchar
|
|
elif nextchar == '.' or (nextchar == ',' and len(token) >= 2):
|
|
token += nextchar
|
|
state = '0.'
|
|
else:
|
|
self.charstack.append(nextchar)
|
|
break # emit token
|
|
elif state == 'a.':
|
|
# If we've seen some letters and a dot separator, continue
|
|
# parsing, and the tokens will be broken up later.
|
|
seenletters = True
|
|
if nextchar == '.' or self.isword(nextchar):
|
|
token += nextchar
|
|
elif self.isnum(nextchar) and token[-1] == '.':
|
|
token += nextchar
|
|
state = '0.'
|
|
else:
|
|
self.charstack.append(nextchar)
|
|
break # emit token
|
|
elif state == '0.':
|
|
# If we've seen at least one dot separator, keep going, we'll
|
|
# break up the tokens later.
|
|
if nextchar == '.' or self.isnum(nextchar):
|
|
token += nextchar
|
|
elif self.isword(nextchar) and token[-1] == '.':
|
|
token += nextchar
|
|
state = 'a.'
|
|
else:
|
|
self.charstack.append(nextchar)
|
|
break # emit token
|
|
|
|
if (state in ('a.', '0.') and (seenletters or token.count('.') > 1 or
|
|
token[-1] in '.,')):
|
|
l = self._split_decimal.split(token)
|
|
token = l[0]
|
|
for tok in l[1:]:
|
|
if tok:
|
|
self.tokenstack.append(tok)
|
|
|
|
if state == '0.' and token.count('.') == 0:
|
|
token = token.replace(',', '.')
|
|
|
|
return token
|
|
|
|
def __iter__(self):
|
|
return self
|
|
|
|
def __next__(self):
|
|
token = self.get_token()
|
|
if token is None:
|
|
raise StopIteration
|
|
|
|
return token
|
|
|
|
def next(self):
|
|
return self.__next__() # Python 2.x support
|
|
|
|
@classmethod
|
|
def split(cls, s):
|
|
return list(cls(s))
|
|
|
|
@classmethod
|
|
def isword(cls, nextchar):
|
|
""" Whether or not the next character is part of a word """
|
|
return nextchar.isalpha()
|
|
|
|
@classmethod
|
|
def isnum(cls, nextchar):
|
|
""" Whether the next character is part of a number """
|
|
return nextchar.isdigit()
|
|
|
|
@classmethod
|
|
def isspace(cls, nextchar):
|
|
""" Whether the next character is whitespace """
|
|
return nextchar.isspace()
|
|
|
|
|
|
class _resultbase(object):
|
|
|
|
def __init__(self):
|
|
for attr in self.__slots__:
|
|
setattr(self, attr, None)
|
|
|
|
def _repr(self, classname):
|
|
l = []
|
|
for attr in self.__slots__:
|
|
value = getattr(self, attr)
|
|
if value is not None:
|
|
l.append("%s=%s" % (attr, repr(value)))
|
|
return "%s(%s)" % (classname, ", ".join(l))
|
|
|
|
def __len__(self):
|
|
return (sum(getattr(self, attr) is not None
|
|
for attr in self.__slots__))
|
|
|
|
def __repr__(self):
|
|
return self._repr(self.__class__.__name__)
|
|
|
|
|
|
class parserinfo(object):
|
|
"""
|
|
Class which handles what inputs are accepted. Subclass this to customize
|
|
the language and acceptable values for each parameter.
|
|
|
|
:param dayfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the day (``True``) or month (``False``). If
|
|
``yearfirst`` is set to ``True``, this distinguishes between YDM
|
|
and YMD. Default is ``False``.
|
|
|
|
:param yearfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the year. If ``True``, the first number is taken
|
|
to be the year, otherwise the last number is taken to be the year.
|
|
Default is ``False``.
|
|
"""
|
|
|
|
# m from a.m/p.m, t from ISO T separator
|
|
JUMP = [" ", ".", ",", ";", "-", "/", "'",
|
|
"at", "on", "and", "ad", "m", "t", "of",
|
|
"st", "nd", "rd", "th"]
|
|
|
|
WEEKDAYS = [("Mon", "Monday"),
|
|
("Tue", "Tuesday"),
|
|
("Wed", "Wednesday"),
|
|
("Thu", "Thursday"),
|
|
("Fri", "Friday"),
|
|
("Sat", "Saturday"),
|
|
("Sun", "Sunday")]
|
|
MONTHS = [("Jan", "January"),
|
|
("Feb", "February"),
|
|
("Mar", "March"),
|
|
("Apr", "April"),
|
|
("May", "May"),
|
|
("Jun", "June"),
|
|
("Jul", "July"),
|
|
("Aug", "August"),
|
|
("Sep", "Sept", "September"),
|
|
("Oct", "October"),
|
|
("Nov", "November"),
|
|
("Dec", "December")]
|
|
HMS = [("h", "hour", "hours"),
|
|
("m", "minute", "minutes"),
|
|
("s", "second", "seconds")]
|
|
AMPM = [("am", "a"),
|
|
("pm", "p")]
|
|
UTCZONE = ["UTC", "GMT", "Z"]
|
|
PERTAIN = ["of"]
|
|
TZOFFSET = {}
|
|
|
|
def __init__(self, dayfirst=False, yearfirst=False):
|
|
self._jump = self._convert(self.JUMP)
|
|
self._weekdays = self._convert(self.WEEKDAYS)
|
|
self._months = self._convert(self.MONTHS)
|
|
self._hms = self._convert(self.HMS)
|
|
self._ampm = self._convert(self.AMPM)
|
|
self._utczone = self._convert(self.UTCZONE)
|
|
self._pertain = self._convert(self.PERTAIN)
|
|
|
|
self.dayfirst = dayfirst
|
|
self.yearfirst = yearfirst
|
|
|
|
self._year = time.localtime().tm_year
|
|
self._century = self._year // 100 * 100
|
|
|
|
def _convert(self, lst):
|
|
dct = {}
|
|
for i, v in enumerate(lst):
|
|
if isinstance(v, tuple):
|
|
for v in v:
|
|
dct[v.lower()] = i
|
|
else:
|
|
dct[v.lower()] = i
|
|
return dct
|
|
|
|
def jump(self, name):
|
|
return name.lower() in self._jump
|
|
|
|
def weekday(self, name):
|
|
if len(name) >= 3:
|
|
try:
|
|
return self._weekdays[name.lower()]
|
|
except KeyError:
|
|
pass
|
|
return None
|
|
|
|
def month(self, name):
|
|
if len(name) >= 3:
|
|
try:
|
|
return self._months[name.lower()] + 1
|
|
except KeyError:
|
|
pass
|
|
return None
|
|
|
|
def hms(self, name):
|
|
try:
|
|
return self._hms[name.lower()]
|
|
except KeyError:
|
|
return None
|
|
|
|
def ampm(self, name):
|
|
try:
|
|
return self._ampm[name.lower()]
|
|
except KeyError:
|
|
return None
|
|
|
|
def pertain(self, name):
|
|
return name.lower() in self._pertain
|
|
|
|
def utczone(self, name):
|
|
return name.lower() in self._utczone
|
|
|
|
def tzoffset(self, name):
|
|
if name in self._utczone:
|
|
return 0
|
|
|
|
return self.TZOFFSET.get(name)
|
|
|
|
def convertyear(self, year, century_specified=False):
|
|
if year < 100 and not century_specified:
|
|
year += self._century
|
|
if abs(year - self._year) >= 50:
|
|
if year < self._year:
|
|
year += 100
|
|
else:
|
|
year -= 100
|
|
return year
|
|
|
|
def validate(self, res):
|
|
# move to info
|
|
if res.year is not None:
|
|
res.year = self.convertyear(res.year, res.century_specified)
|
|
|
|
if res.tzoffset == 0 and not res.tzname or res.tzname == 'Z':
|
|
res.tzname = "UTC"
|
|
res.tzoffset = 0
|
|
elif res.tzoffset != 0 and res.tzname and self.utczone(res.tzname):
|
|
res.tzoffset = 0
|
|
return True
|
|
|
|
|
|
class _ymd(list):
|
|
def __init__(self, tzstr, *args, **kwargs):
|
|
super(self.__class__, self).__init__(*args, **kwargs)
|
|
self.century_specified = False
|
|
self.tzstr = tzstr
|
|
|
|
@staticmethod
|
|
def token_could_be_year(token, year):
|
|
try:
|
|
return int(token) == year
|
|
except ValueError:
|
|
return False
|
|
|
|
@staticmethod
|
|
def find_potential_year_tokens(year, tokens):
|
|
return [token for token in tokens if _ymd.token_could_be_year(token, year)]
|
|
|
|
def find_probable_year_index(self, tokens):
|
|
"""
|
|
attempt to deduce if a pre 100 year was lost
|
|
due to padded zeros being taken off
|
|
"""
|
|
for index, token in enumerate(self):
|
|
potential_year_tokens = _ymd.find_potential_year_tokens(token, tokens)
|
|
if len(potential_year_tokens) == 1 and len(potential_year_tokens[0]) > 2:
|
|
return index
|
|
|
|
def append(self, val):
|
|
if hasattr(val, '__len__'):
|
|
if val.isdigit() and len(val) > 2:
|
|
self.century_specified = True
|
|
elif val > 100:
|
|
self.century_specified = True
|
|
|
|
super(self.__class__, self).append(int(val))
|
|
|
|
def resolve_ymd(self, mstridx, yearfirst, dayfirst):
|
|
len_ymd = len(self)
|
|
year, month, day = (None, None, None)
|
|
|
|
if len_ymd > 3:
|
|
raise ValueError("More than three YMD values")
|
|
elif len_ymd == 1 or (mstridx != -1 and len_ymd == 2):
|
|
# One member, or two members with a month string
|
|
if mstridx != -1:
|
|
month = self[mstridx]
|
|
del self[mstridx]
|
|
|
|
if len_ymd > 1 or mstridx == -1:
|
|
if self[0] > 31:
|
|
year = self[0]
|
|
else:
|
|
day = self[0]
|
|
|
|
elif len_ymd == 2:
|
|
# Two members with numbers
|
|
if self[0] > 31:
|
|
# 99-01
|
|
year, month = self
|
|
elif self[1] > 31:
|
|
# 01-99
|
|
month, year = self
|
|
elif dayfirst and self[1] <= 12:
|
|
# 13-01
|
|
day, month = self
|
|
else:
|
|
# 01-13
|
|
month, day = self
|
|
|
|
elif len_ymd == 3:
|
|
# Three members
|
|
if mstridx == 0:
|
|
month, day, year = self
|
|
elif mstridx == 1:
|
|
if self[0] > 31 or (yearfirst and self[2] <= 31):
|
|
# 99-Jan-01
|
|
year, month, day = self
|
|
else:
|
|
# 01-Jan-01
|
|
# Give precendence to day-first, since
|
|
# two-digit years is usually hand-written.
|
|
day, month, year = self
|
|
|
|
elif mstridx == 2:
|
|
# WTF!?
|
|
if self[1] > 31:
|
|
# 01-99-Jan
|
|
day, year, month = self
|
|
else:
|
|
# 99-01-Jan
|
|
year, day, month = self
|
|
|
|
else:
|
|
if self[0] > 31 or \
|
|
self.find_probable_year_index(_timelex.split(self.tzstr)) == 0 or \
|
|
(yearfirst and self[1] <= 12 and self[2] <= 31):
|
|
# 99-01-01
|
|
if dayfirst and self[2] <= 12:
|
|
year, day, month = self
|
|
else:
|
|
year, month, day = self
|
|
elif self[0] > 12 or (dayfirst and self[1] <= 12):
|
|
# 13-01-01
|
|
day, month, year = self
|
|
else:
|
|
# 01-13-01
|
|
month, day, year = self
|
|
|
|
return year, month, day
|
|
|
|
|
|
class parser(object):
|
|
def __init__(self, info=None):
|
|
self.info = info or parserinfo()
|
|
|
|
def parse(self, timestr, default=None, ignoretz=False, tzinfos=None, **kwargs):
|
|
"""
|
|
Parse the date/time string into a :class:`datetime.datetime` object.
|
|
|
|
:param timestr:
|
|
Any date/time string using the supported formats.
|
|
|
|
:param default:
|
|
The default datetime object, if this is a datetime object and not
|
|
``None``, elements specified in ``timestr`` replace elements in the
|
|
default object.
|
|
|
|
:param ignoretz:
|
|
If set ``True``, time zones in parsed strings are ignored and a
|
|
naive :class:`datetime.datetime` object is returned.
|
|
|
|
:param tzinfos:
|
|
Additional time zone names / aliases which may be present in the
|
|
string. This argument maps time zone names (and optionally offsets
|
|
from those time zones) to time zones. This parameter can be a
|
|
dictionary with timezone aliases mapping time zone names to time
|
|
zones or a function taking two parameters (``tzname`` and
|
|
``tzoffset``) and returning a time zone.
|
|
|
|
The timezones to which the names are mapped can be an integer
|
|
offset from UTC in minutes or a :class:`tzinfo` object.
|
|
|
|
.. doctest::
|
|
:options: +NORMALIZE_WHITESPACE
|
|
|
|
>>> from dateutil.parser import parse
|
|
>>> from dateutil.tz import gettz
|
|
>>> tzinfos = {"BRST": -10800, "CST": gettz("America/Chicago")}
|
|
>>> parse("2012-01-19 17:21:00 BRST", tzinfos=tzinfos)
|
|
datetime.datetime(2012, 1, 19, 17, 21, tzinfo=tzoffset(u'BRST', -10800))
|
|
>>> parse("2012-01-19 17:21:00 CST", tzinfos=tzinfos)
|
|
datetime.datetime(2012, 1, 19, 17, 21,
|
|
tzinfo=tzfile('/usr/share/zoneinfo/America/Chicago'))
|
|
|
|
This parameter is ignored if ``ignoretz`` is set.
|
|
|
|
:param **kwargs:
|
|
Keyword arguments as passed to ``_parse()``.
|
|
|
|
:return:
|
|
Returns a :class:`datetime.datetime` object or, if the
|
|
``fuzzy_with_tokens`` option is ``True``, returns a tuple, the
|
|
first element being a :class:`datetime.datetime` object, the second
|
|
a tuple containing the fuzzy tokens.
|
|
|
|
:raises ValueError:
|
|
Raised for invalid or unknown string format, if the provided
|
|
:class:`tzinfo` is not in a valid format, or if an invalid date
|
|
would be created.
|
|
|
|
:raises OverflowError:
|
|
Raised if the parsed date exceeds the largest valid C integer on
|
|
your system.
|
|
"""
|
|
|
|
if default is None:
|
|
effective_dt = datetime.datetime.now()
|
|
default = datetime.datetime.now().replace(hour=0, minute=0,
|
|
second=0, microsecond=0)
|
|
else:
|
|
effective_dt = default
|
|
|
|
res, skipped_tokens = self._parse(timestr, **kwargs)
|
|
|
|
if res is None:
|
|
raise ValueError("Unknown string format")
|
|
|
|
if len(res) == 0:
|
|
raise ValueError("String does not contain a date.")
|
|
|
|
repl = {}
|
|
for attr in ("year", "month", "day", "hour",
|
|
"minute", "second", "microsecond"):
|
|
value = getattr(res, attr)
|
|
if value is not None:
|
|
repl[attr] = value
|
|
|
|
if 'day' not in repl:
|
|
# If the default day exceeds the last day of the month, fall back to
|
|
# the end of the month.
|
|
cyear = default.year if res.year is None else res.year
|
|
cmonth = default.month if res.month is None else res.month
|
|
cday = default.day if res.day is None else res.day
|
|
|
|
if cday > monthrange(cyear, cmonth)[1]:
|
|
repl['day'] = monthrange(cyear, cmonth)[1]
|
|
|
|
ret = default.replace(**repl)
|
|
|
|
if res.weekday is not None and not res.day:
|
|
ret = ret+relativedelta.relativedelta(weekday=res.weekday)
|
|
|
|
if not ignoretz:
|
|
if (isinstance(tzinfos, collections.Callable) or
|
|
tzinfos and res.tzname in tzinfos):
|
|
|
|
if isinstance(tzinfos, collections.Callable):
|
|
tzdata = tzinfos(res.tzname, res.tzoffset)
|
|
else:
|
|
tzdata = tzinfos.get(res.tzname)
|
|
|
|
if isinstance(tzdata, datetime.tzinfo):
|
|
tzinfo = tzdata
|
|
elif isinstance(tzdata, text_type):
|
|
tzinfo = tz.tzstr(tzdata)
|
|
elif isinstance(tzdata, integer_types):
|
|
tzinfo = tz.tzoffset(res.tzname, tzdata)
|
|
else:
|
|
raise ValueError("Offset must be tzinfo subclass, "
|
|
"tz string, or int offset.")
|
|
ret = ret.replace(tzinfo=tzinfo)
|
|
elif res.tzname and res.tzname in time.tzname:
|
|
ret = ret.replace(tzinfo=tz.tzlocal())
|
|
elif res.tzoffset == 0:
|
|
ret = ret.replace(tzinfo=tz.tzutc())
|
|
elif res.tzoffset:
|
|
ret = ret.replace(tzinfo=tz.tzoffset(res.tzname, res.tzoffset))
|
|
|
|
if kwargs.get('fuzzy_with_tokens', False):
|
|
return ret, skipped_tokens
|
|
else:
|
|
return ret
|
|
|
|
class _result(_resultbase):
|
|
__slots__ = ["year", "month", "day", "weekday",
|
|
"hour", "minute", "second", "microsecond",
|
|
"tzname", "tzoffset", "ampm"]
|
|
|
|
def _parse(self, timestr, dayfirst=None, yearfirst=None, fuzzy=False,
|
|
fuzzy_with_tokens=False):
|
|
"""
|
|
Private method which performs the heavy lifting of parsing, called from
|
|
``parse()``, which passes on its ``kwargs`` to this function.
|
|
|
|
:param timestr:
|
|
The string to parse.
|
|
|
|
:param dayfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the day (``True``) or month (``False``). If
|
|
``yearfirst`` is set to ``True``, this distinguishes between YDM
|
|
and YMD. If set to ``None``, this value is retrieved from the
|
|
current :class:`parserinfo` object (which itself defaults to
|
|
``False``).
|
|
|
|
:param yearfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the year. If ``True``, the first number is taken
|
|
to be the year, otherwise the last number is taken to be the year.
|
|
If this is set to ``None``, the value is retrieved from the current
|
|
:class:`parserinfo` object (which itself defaults to ``False``).
|
|
|
|
:param fuzzy:
|
|
Whether to allow fuzzy parsing, allowing for string like "Today is
|
|
January 1, 2047 at 8:21:00AM".
|
|
|
|
:param fuzzy_with_tokens:
|
|
If ``True``, ``fuzzy`` is automatically set to True, and the parser
|
|
will return a tuple where the first element is the parsed
|
|
:class:`datetime.datetime` datetimestamp and the second element is
|
|
a tuple containing the portions of the string which were ignored:
|
|
|
|
.. doctest::
|
|
|
|
>>> from dateutil.parser import parse
|
|
>>> parse("Today is January 1, 2047 at 8:21:00AM", fuzzy_with_tokens=True)
|
|
(datetime.datetime(2047, 1, 1, 8, 21), (u'Today is ', u' ', u'at '))
|
|
|
|
"""
|
|
if fuzzy_with_tokens:
|
|
fuzzy = True
|
|
|
|
info = self.info
|
|
|
|
if dayfirst is None:
|
|
dayfirst = info.dayfirst
|
|
|
|
if yearfirst is None:
|
|
yearfirst = info.yearfirst
|
|
|
|
res = self._result()
|
|
l = _timelex.split(timestr) # Splits the timestr into tokens
|
|
|
|
# keep up with the last token skipped so we can recombine
|
|
# consecutively skipped tokens (-2 for when i begins at 0).
|
|
last_skipped_token_i = -2
|
|
skipped_tokens = list()
|
|
|
|
try:
|
|
# year/month/day list
|
|
ymd = _ymd(timestr)
|
|
|
|
# Index of the month string in ymd
|
|
mstridx = -1
|
|
|
|
len_l = len(l)
|
|
i = 0
|
|
while i < len_l:
|
|
|
|
# Check if it's a number
|
|
try:
|
|
value_repr = l[i]
|
|
value = float(value_repr)
|
|
except ValueError:
|
|
value = None
|
|
|
|
if value is not None:
|
|
# Token is a number
|
|
len_li = len(l[i])
|
|
i += 1
|
|
|
|
if (len(ymd) == 3 and len_li in (2, 4)
|
|
and res.hour is None and (i >= len_l or (l[i] != ':' and
|
|
info.hms(l[i]) is None))):
|
|
# 19990101T23[59]
|
|
s = l[i-1]
|
|
res.hour = int(s[:2])
|
|
|
|
if len_li == 4:
|
|
res.minute = int(s[2:])
|
|
|
|
elif len_li == 6 or (len_li > 6 and l[i-1].find('.') == 6):
|
|
# YYMMDD or HHMMSS[.ss]
|
|
s = l[i-1]
|
|
|
|
if not ymd and l[i-1].find('.') == -1:
|
|
#ymd.append(info.convertyear(int(s[:2])))
|
|
|
|
ymd.append(s[:2])
|
|
ymd.append(s[2:4])
|
|
ymd.append(s[4:])
|
|
else:
|
|
# 19990101T235959[.59]
|
|
res.hour = int(s[:2])
|
|
res.minute = int(s[2:4])
|
|
res.second, res.microsecond = _parsems(s[4:])
|
|
|
|
elif len_li in (8, 12, 14):
|
|
# YYYYMMDD
|
|
s = l[i-1]
|
|
ymd.append(s[:4])
|
|
ymd.append(s[4:6])
|
|
ymd.append(s[6:8])
|
|
|
|
if len_li > 8:
|
|
res.hour = int(s[8:10])
|
|
res.minute = int(s[10:12])
|
|
|
|
if len_li > 12:
|
|
res.second = int(s[12:])
|
|
|
|
elif ((i < len_l and info.hms(l[i]) is not None) or
|
|
(i+1 < len_l and l[i] == ' ' and
|
|
info.hms(l[i+1]) is not None)):
|
|
|
|
# HH[ ]h or MM[ ]m or SS[.ss][ ]s
|
|
if l[i] == ' ':
|
|
i += 1
|
|
|
|
idx = info.hms(l[i])
|
|
|
|
while True:
|
|
if idx == 0:
|
|
res.hour = int(value)
|
|
|
|
if value % 1:
|
|
res.minute = int(60*(value % 1))
|
|
|
|
elif idx == 1:
|
|
res.minute = int(value)
|
|
|
|
if value % 1:
|
|
res.second = int(60*(value % 1))
|
|
|
|
elif idx == 2:
|
|
res.second, res.microsecond = \
|
|
_parsems(value_repr)
|
|
|
|
i += 1
|
|
|
|
if i >= len_l or idx == 2:
|
|
break
|
|
|
|
# 12h00
|
|
try:
|
|
value_repr = l[i]
|
|
value = float(value_repr)
|
|
except ValueError:
|
|
break
|
|
else:
|
|
i += 1
|
|
idx += 1
|
|
|
|
if i < len_l:
|
|
newidx = info.hms(l[i])
|
|
|
|
if newidx is not None:
|
|
idx = newidx
|
|
|
|
elif (i == len_l and l[i-2] == ' ' and
|
|
info.hms(l[i-3]) is not None):
|
|
# X h MM or X m SS
|
|
idx = info.hms(l[i-3]) + 1
|
|
|
|
if idx == 1:
|
|
res.minute = int(value)
|
|
|
|
if value % 1:
|
|
res.second = int(60*(value % 1))
|
|
elif idx == 2:
|
|
res.second, res.microsecond = \
|
|
_parsems(value_repr)
|
|
i += 1
|
|
|
|
elif i+1 < len_l and l[i] == ':':
|
|
# HH:MM[:SS[.ss]]
|
|
res.hour = int(value)
|
|
i += 1
|
|
value = float(l[i])
|
|
res.minute = int(value)
|
|
|
|
if value % 1:
|
|
res.second = int(60*(value % 1))
|
|
|
|
i += 1
|
|
|
|
if i < len_l and l[i] == ':':
|
|
res.second, res.microsecond = _parsems(l[i+1])
|
|
i += 2
|
|
|
|
elif i < len_l and l[i] in ('-', '/', '.'):
|
|
sep = l[i]
|
|
ymd.append(value_repr)
|
|
i += 1
|
|
|
|
if i < len_l and not info.jump(l[i]):
|
|
try:
|
|
# 01-01[-01]
|
|
ymd.append(l[i])
|
|
except ValueError:
|
|
# 01-Jan[-01]
|
|
value = info.month(l[i])
|
|
|
|
if value is not None:
|
|
ymd.append(value)
|
|
assert mstridx == -1
|
|
mstridx = len(ymd)-1
|
|
else:
|
|
return None, None
|
|
|
|
i += 1
|
|
|
|
if i < len_l and l[i] == sep:
|
|
# We have three members
|
|
i += 1
|
|
value = info.month(l[i])
|
|
|
|
if value is not None:
|
|
ymd.append(value)
|
|
mstridx = len(ymd)-1
|
|
assert mstridx == -1
|
|
else:
|
|
ymd.append(l[i])
|
|
|
|
i += 1
|
|
elif i >= len_l or info.jump(l[i]):
|
|
if i+1 < len_l and info.ampm(l[i+1]) is not None:
|
|
# 12 am
|
|
res.hour = int(value)
|
|
|
|
if res.hour < 12 and info.ampm(l[i+1]) == 1:
|
|
res.hour += 12
|
|
elif res.hour == 12 and info.ampm(l[i+1]) == 0:
|
|
res.hour = 0
|
|
|
|
i += 1
|
|
else:
|
|
# Year, month or day
|
|
ymd.append(value)
|
|
i += 1
|
|
elif info.ampm(l[i]) is not None:
|
|
|
|
# 12am
|
|
res.hour = int(value)
|
|
|
|
if res.hour < 12 and info.ampm(l[i]) == 1:
|
|
res.hour += 12
|
|
elif res.hour == 12 and info.ampm(l[i]) == 0:
|
|
res.hour = 0
|
|
i += 1
|
|
|
|
elif not fuzzy:
|
|
return None, None
|
|
else:
|
|
i += 1
|
|
continue
|
|
|
|
# Check weekday
|
|
value = info.weekday(l[i])
|
|
if value is not None:
|
|
res.weekday = value
|
|
i += 1
|
|
continue
|
|
|
|
# Check month name
|
|
value = info.month(l[i])
|
|
if value is not None:
|
|
ymd.append(value)
|
|
assert mstridx == -1
|
|
mstridx = len(ymd)-1
|
|
|
|
i += 1
|
|
if i < len_l:
|
|
if l[i] in ('-', '/'):
|
|
# Jan-01[-99]
|
|
sep = l[i]
|
|
i += 1
|
|
ymd.append(l[i])
|
|
i += 1
|
|
|
|
if i < len_l and l[i] == sep:
|
|
# Jan-01-99
|
|
i += 1
|
|
ymd.append(l[i])
|
|
i += 1
|
|
|
|
elif (i+3 < len_l and l[i] == l[i+2] == ' '
|
|
and info.pertain(l[i+1])):
|
|
# Jan of 01
|
|
# In this case, 01 is clearly year
|
|
try:
|
|
value = int(l[i+3])
|
|
except ValueError:
|
|
# Wrong guess
|
|
pass
|
|
else:
|
|
# Convert it here to become unambiguous
|
|
ymd.append(str(info.convertyear(value)))
|
|
i += 4
|
|
continue
|
|
|
|
# Check am/pm
|
|
value = info.ampm(l[i])
|
|
if value is not None:
|
|
# For fuzzy parsing, 'a' or 'am' (both valid English words)
|
|
# may erroneously trigger the AM/PM flag. Deal with that
|
|
# here.
|
|
val_is_ampm = True
|
|
|
|
# If there's already an AM/PM flag, this one isn't one.
|
|
if fuzzy and res.ampm is not None:
|
|
val_is_ampm = False
|
|
|
|
# If AM/PM is found and hour is not, raise a ValueError
|
|
if res.hour is None:
|
|
if fuzzy:
|
|
val_is_ampm = False
|
|
else:
|
|
raise ValueError('No hour specified with ' +
|
|
'AM or PM flag.')
|
|
elif not 0 <= res.hour <= 12:
|
|
# If AM/PM is found, it's a 12 hour clock, so raise
|
|
# an error for invalid range
|
|
if fuzzy:
|
|
val_is_ampm = False
|
|
else:
|
|
raise ValueError('Invalid hour specified for ' +
|
|
'12-hour clock.')
|
|
|
|
if val_is_ampm:
|
|
if value == 1 and res.hour < 12:
|
|
res.hour += 12
|
|
elif value == 0 and res.hour == 12:
|
|
res.hour = 0
|
|
|
|
res.ampm = value
|
|
|
|
i += 1
|
|
continue
|
|
|
|
# Check for a timezone name
|
|
if (res.hour is not None and len(l[i]) <= 5 and
|
|
res.tzname is None and res.tzoffset is None and
|
|
not [x for x in l[i] if x not in
|
|
string.ascii_uppercase]):
|
|
res.tzname = l[i]
|
|
res.tzoffset = info.tzoffset(res.tzname)
|
|
i += 1
|
|
|
|
# Check for something like GMT+3, or BRST+3. Notice
|
|
# that it doesn't mean "I am 3 hours after GMT", but
|
|
# "my time +3 is GMT". If found, we reverse the
|
|
# logic so that timezone parsing code will get it
|
|
# right.
|
|
if i < len_l and l[i] in ('+', '-'):
|
|
l[i] = ('+', '-')[l[i] == '+']
|
|
res.tzoffset = None
|
|
if info.utczone(res.tzname):
|
|
# With something like GMT+3, the timezone
|
|
# is *not* GMT.
|
|
res.tzname = None
|
|
|
|
continue
|
|
|
|
# Check for a numbered timezone
|
|
if res.hour is not None and l[i] in ('+', '-'):
|
|
signal = (-1, 1)[l[i] == '+']
|
|
i += 1
|
|
len_li = len(l[i])
|
|
|
|
if len_li == 4:
|
|
# -0300
|
|
res.tzoffset = int(l[i][:2])*3600+int(l[i][2:])*60
|
|
elif i+1 < len_l and l[i+1] == ':':
|
|
# -03:00
|
|
res.tzoffset = int(l[i])*3600+int(l[i+2])*60
|
|
i += 2
|
|
elif len_li <= 2:
|
|
# -[0]3
|
|
res.tzoffset = int(l[i][:2])*3600
|
|
else:
|
|
return None, None
|
|
i += 1
|
|
|
|
res.tzoffset *= signal
|
|
|
|
# Look for a timezone name between parenthesis
|
|
if (i+3 < len_l and
|
|
info.jump(l[i]) and l[i+1] == '(' and l[i+3] == ')' and
|
|
3 <= len(l[i+2]) <= 5 and
|
|
not [x for x in l[i+2]
|
|
if x not in string.ascii_uppercase]):
|
|
# -0300 (BRST)
|
|
res.tzname = l[i+2]
|
|
i += 4
|
|
continue
|
|
|
|
# Check jumps
|
|
if not (info.jump(l[i]) or fuzzy):
|
|
return None, None
|
|
|
|
if last_skipped_token_i == i - 1:
|
|
# recombine the tokens
|
|
skipped_tokens[-1] += l[i]
|
|
else:
|
|
# just append
|
|
skipped_tokens.append(l[i])
|
|
last_skipped_token_i = i
|
|
i += 1
|
|
|
|
# Process year/month/day
|
|
year, month, day = ymd.resolve_ymd(mstridx, yearfirst, dayfirst)
|
|
if year is not None:
|
|
res.year = year
|
|
res.century_specified = ymd.century_specified
|
|
|
|
if month is not None:
|
|
res.month = month
|
|
|
|
if day is not None:
|
|
res.day = day
|
|
|
|
except (IndexError, ValueError, AssertionError):
|
|
return None, None
|
|
|
|
if not info.validate(res):
|
|
return None, None
|
|
|
|
if fuzzy_with_tokens:
|
|
return res, tuple(skipped_tokens)
|
|
else:
|
|
return res, None
|
|
|
|
DEFAULTPARSER = parser()
|
|
|
|
|
|
def parse(timestr, parserinfo=None, **kwargs):
|
|
"""
|
|
|
|
Parse a string in one of the supported formats, using the
|
|
``parserinfo`` parameters.
|
|
|
|
:param timestr:
|
|
A string containing a date/time stamp.
|
|
|
|
:param parserinfo:
|
|
A :class:`parserinfo` object containing parameters for the parser.
|
|
If ``None``, the default arguments to the :class:`parserinfo`
|
|
constructor are used.
|
|
|
|
The ``**kwargs`` parameter takes the following keyword arguments:
|
|
|
|
:param default:
|
|
The default datetime object, if this is a datetime object and not
|
|
``None``, elements specified in ``timestr`` replace elements in the
|
|
default object.
|
|
|
|
:param ignoretz:
|
|
If set ``True``, time zones in parsed strings are ignored and a naive
|
|
:class:`datetime` object is returned.
|
|
|
|
:param tzinfos:
|
|
Additional time zone names / aliases which may be present in the
|
|
string. This argument maps time zone names (and optionally offsets
|
|
from those time zones) to time zones. This parameter can be a
|
|
dictionary with timezone aliases mapping time zone names to time
|
|
zones or a function taking two parameters (``tzname`` and
|
|
``tzoffset``) and returning a time zone.
|
|
|
|
The timezones to which the names are mapped can be an integer
|
|
offset from UTC in minutes or a :class:`tzinfo` object.
|
|
|
|
.. doctest::
|
|
:options: +NORMALIZE_WHITESPACE
|
|
|
|
>>> from dateutil.parser import parse
|
|
>>> from dateutil.tz import gettz
|
|
>>> tzinfos = {"BRST": -10800, "CST": gettz("America/Chicago")}
|
|
>>> parse("2012-01-19 17:21:00 BRST", tzinfos=tzinfos)
|
|
datetime.datetime(2012, 1, 19, 17, 21, tzinfo=tzoffset(u'BRST', -10800))
|
|
>>> parse("2012-01-19 17:21:00 CST", tzinfos=tzinfos)
|
|
datetime.datetime(2012, 1, 19, 17, 21,
|
|
tzinfo=tzfile('/usr/share/zoneinfo/America/Chicago'))
|
|
|
|
This parameter is ignored if ``ignoretz`` is set.
|
|
|
|
:param dayfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the day (``True``) or month (``False``). If
|
|
``yearfirst`` is set to ``True``, this distinguishes between YDM and
|
|
YMD. If set to ``None``, this value is retrieved from the current
|
|
:class:`parserinfo` object (which itself defaults to ``False``).
|
|
|
|
:param yearfirst:
|
|
Whether to interpret the first value in an ambiguous 3-integer date
|
|
(e.g. 01/05/09) as the year. If ``True``, the first number is taken to
|
|
be the year, otherwise the last number is taken to be the year. If
|
|
this is set to ``None``, the value is retrieved from the current
|
|
:class:`parserinfo` object (which itself defaults to ``False``).
|
|
|
|
:param fuzzy:
|
|
Whether to allow fuzzy parsing, allowing for string like "Today is
|
|
January 1, 2047 at 8:21:00AM".
|
|
|
|
:param fuzzy_with_tokens:
|
|
If ``True``, ``fuzzy`` is automatically set to True, and the parser
|
|
will return a tuple where the first element is the parsed
|
|
:class:`datetime.datetime` datetimestamp and the second element is
|
|
a tuple containing the portions of the string which were ignored:
|
|
|
|
.. doctest::
|
|
|
|
>>> from dateutil.parser import parse
|
|
>>> parse("Today is January 1, 2047 at 8:21:00AM", fuzzy_with_tokens=True)
|
|
(datetime.datetime(2011, 1, 1, 8, 21), (u'Today is ', u' ', u'at '))
|
|
|
|
:return:
|
|
Returns a :class:`datetime.datetime` object or, if the
|
|
``fuzzy_with_tokens`` option is ``True``, returns a tuple, the
|
|
first element being a :class:`datetime.datetime` object, the second
|
|
a tuple containing the fuzzy tokens.
|
|
|
|
:raises ValueError:
|
|
Raised for invalid or unknown string format, if the provided
|
|
:class:`tzinfo` is not in a valid format, or if an invalid date
|
|
would be created.
|
|
|
|
:raises OverflowError:
|
|
Raised if the parsed date exceeds the largest valid C integer on
|
|
your system.
|
|
"""
|
|
if parserinfo:
|
|
return parser(parserinfo).parse(timestr, **kwargs)
|
|
else:
|
|
return DEFAULTPARSER.parse(timestr, **kwargs)
|
|
|
|
|
|
class _tzparser(object):
|
|
|
|
class _result(_resultbase):
|
|
|
|
__slots__ = ["stdabbr", "stdoffset", "dstabbr", "dstoffset",
|
|
"start", "end"]
|
|
|
|
class _attr(_resultbase):
|
|
__slots__ = ["month", "week", "weekday",
|
|
"yday", "jyday", "day", "time"]
|
|
|
|
def __repr__(self):
|
|
return self._repr("")
|
|
|
|
def __init__(self):
|
|
_resultbase.__init__(self)
|
|
self.start = self._attr()
|
|
self.end = self._attr()
|
|
|
|
def parse(self, tzstr):
|
|
res = self._result()
|
|
l = _timelex.split(tzstr)
|
|
try:
|
|
|
|
len_l = len(l)
|
|
|
|
i = 0
|
|
while i < len_l:
|
|
# BRST+3[BRDT[+2]]
|
|
j = i
|
|
while j < len_l and not [x for x in l[j]
|
|
if x in "0123456789:,-+"]:
|
|
j += 1
|
|
if j != i:
|
|
if not res.stdabbr:
|
|
offattr = "stdoffset"
|
|
res.stdabbr = "".join(l[i:j])
|
|
else:
|
|
offattr = "dstoffset"
|
|
res.dstabbr = "".join(l[i:j])
|
|
i = j
|
|
if (i < len_l and (l[i] in ('+', '-') or l[i][0] in
|
|
"0123456789")):
|
|
if l[i] in ('+', '-'):
|
|
# Yes, that's right. See the TZ variable
|
|
# documentation.
|
|
signal = (1, -1)[l[i] == '+']
|
|
i += 1
|
|
else:
|
|
signal = -1
|
|
len_li = len(l[i])
|
|
if len_li == 4:
|
|
# -0300
|
|
setattr(res, offattr, (int(l[i][:2])*3600 +
|
|
int(l[i][2:])*60)*signal)
|
|
elif i+1 < len_l and l[i+1] == ':':
|
|
# -03:00
|
|
setattr(res, offattr,
|
|
(int(l[i])*3600+int(l[i+2])*60)*signal)
|
|
i += 2
|
|
elif len_li <= 2:
|
|
# -[0]3
|
|
setattr(res, offattr,
|
|
int(l[i][:2])*3600*signal)
|
|
else:
|
|
return None
|
|
i += 1
|
|
if res.dstabbr:
|
|
break
|
|
else:
|
|
break
|
|
|
|
if i < len_l:
|
|
for j in range(i, len_l):
|
|
if l[j] == ';':
|
|
l[j] = ','
|
|
|
|
assert l[i] == ','
|
|
|
|
i += 1
|
|
|
|
if i >= len_l:
|
|
pass
|
|
elif (8 <= l.count(',') <= 9 and
|
|
not [y for x in l[i:] if x != ','
|
|
for y in x if y not in "0123456789"]):
|
|
# GMT0BST,3,0,30,3600,10,0,26,7200[,3600]
|
|
for x in (res.start, res.end):
|
|
x.month = int(l[i])
|
|
i += 2
|
|
if l[i] == '-':
|
|
value = int(l[i+1])*-1
|
|
i += 1
|
|
else:
|
|
value = int(l[i])
|
|
i += 2
|
|
if value:
|
|
x.week = value
|
|
x.weekday = (int(l[i])-1) % 7
|
|
else:
|
|
x.day = int(l[i])
|
|
i += 2
|
|
x.time = int(l[i])
|
|
i += 2
|
|
if i < len_l:
|
|
if l[i] in ('-', '+'):
|
|
signal = (-1, 1)[l[i] == "+"]
|
|
i += 1
|
|
else:
|
|
signal = 1
|
|
res.dstoffset = (res.stdoffset+int(l[i]))*signal
|
|
elif (l.count(',') == 2 and l[i:].count('/') <= 2 and
|
|
not [y for x in l[i:] if x not in (',', '/', 'J', 'M',
|
|
'.', '-', ':')
|
|
for y in x if y not in "0123456789"]):
|
|
for x in (res.start, res.end):
|
|
if l[i] == 'J':
|
|
# non-leap year day (1 based)
|
|
i += 1
|
|
x.jyday = int(l[i])
|
|
elif l[i] == 'M':
|
|
# month[-.]week[-.]weekday
|
|
i += 1
|
|
x.month = int(l[i])
|
|
i += 1
|
|
assert l[i] in ('-', '.')
|
|
i += 1
|
|
x.week = int(l[i])
|
|
if x.week == 5:
|
|
x.week = -1
|
|
i += 1
|
|
assert l[i] in ('-', '.')
|
|
i += 1
|
|
x.weekday = (int(l[i])-1) % 7
|
|
else:
|
|
# year day (zero based)
|
|
x.yday = int(l[i])+1
|
|
|
|
i += 1
|
|
|
|
if i < len_l and l[i] == '/':
|
|
i += 1
|
|
# start time
|
|
len_li = len(l[i])
|
|
if len_li == 4:
|
|
# -0300
|
|
x.time = (int(l[i][:2])*3600+int(l[i][2:])*60)
|
|
elif i+1 < len_l and l[i+1] == ':':
|
|
# -03:00
|
|
x.time = int(l[i])*3600+int(l[i+2])*60
|
|
i += 2
|
|
if i+1 < len_l and l[i+1] == ':':
|
|
i += 2
|
|
x.time += int(l[i])
|
|
elif len_li <= 2:
|
|
# -[0]3
|
|
x.time = (int(l[i][:2])*3600)
|
|
else:
|
|
return None
|
|
i += 1
|
|
|
|
assert i == len_l or l[i] == ','
|
|
|
|
i += 1
|
|
|
|
assert i >= len_l
|
|
|
|
except (IndexError, ValueError, AssertionError):
|
|
return None
|
|
|
|
return res
|
|
|
|
|
|
DEFAULTTZPARSER = _tzparser()
|
|
|
|
|
|
def _parsetz(tzstr):
|
|
return DEFAULTTZPARSER.parse(tzstr)
|
|
|
|
|
|
def _parsems(value):
|
|
"""Parse a I[.F] seconds value into (seconds, microseconds)."""
|
|
if "." not in value:
|
|
return int(value), 0
|
|
else:
|
|
i, f = value.split(".")
|
|
return int(i), int(f.ljust(6, "0")[:6])
|
|
|
|
|
|
# vim:ts=4:sw=4:et
|