diff options
Diffstat (limited to '.venv/lib/python3.12/site-packages/dateutil/parser')
3 files changed, 2090 insertions, 0 deletions
diff --git a/.venv/lib/python3.12/site-packages/dateutil/parser/__init__.py b/.venv/lib/python3.12/site-packages/dateutil/parser/__init__.py new file mode 100644 index 00000000..d174b0e4 --- /dev/null +++ b/.venv/lib/python3.12/site-packages/dateutil/parser/__init__.py @@ -0,0 +1,61 @@ +# -*- coding: utf-8 -*- +from ._parser import parse, parser, parserinfo, ParserError +from ._parser import DEFAULTPARSER, DEFAULTTZPARSER +from ._parser import UnknownTimezoneWarning + +from ._parser import __doc__ + +from .isoparser import isoparser, isoparse + +__all__ = ['parse', 'parser', 'parserinfo', + 'isoparse', 'isoparser', + 'ParserError', + 'UnknownTimezoneWarning'] + + +### +# Deprecate portions of the private interface so that downstream code that +# is improperly relying on it is given *some* notice. + + +def __deprecated_private_func(f): + from functools import wraps + import warnings + + msg = ('{name} is a private function and may break without warning, ' + 'it will be moved and or renamed in future versions.') + msg = msg.format(name=f.__name__) + + @wraps(f) + def deprecated_func(*args, **kwargs): + warnings.warn(msg, DeprecationWarning) + return f(*args, **kwargs) + + return deprecated_func + +def __deprecate_private_class(c): + import warnings + + msg = ('{name} is a private class and may break without warning, ' + 'it will be moved and or renamed in future versions.') + msg = msg.format(name=c.__name__) + + class private_class(c): + __doc__ = c.__doc__ + + def __init__(self, *args, **kwargs): + warnings.warn(msg, DeprecationWarning) + super(private_class, self).__init__(*args, **kwargs) + + private_class.__name__ = c.__name__ + + return private_class + + +from ._parser import _timelex, _resultbase +from ._parser import _tzparser, _parsetz + +_timelex = __deprecate_private_class(_timelex) +_tzparser = __deprecate_private_class(_tzparser) +_resultbase = __deprecate_private_class(_resultbase) +_parsetz = __deprecated_private_func(_parsetz) diff --git a/.venv/lib/python3.12/site-packages/dateutil/parser/_parser.py b/.venv/lib/python3.12/site-packages/dateutil/parser/_parser.py new file mode 100644 index 00000000..37d1663b --- /dev/null +++ b/.venv/lib/python3.12/site-packages/dateutil/parser/_parser.py @@ -0,0 +1,1613 @@ +# -*- coding: utf-8 -*- +""" +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 + <https://www.cl.cam.ac.uk/~mgk25/iso-time.html>`_ +- `W3C Date and Time Formats <https://www.w3.org/TR/NOTE-datetime>`_ +- `Time Formats (Planetary Rings Node) <https://pds-rings.seti.org:443/tools/time_formats.html>`_ +- `CPAN ParseDate module + <https://metacpan.org/pod/release/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 re +import string +import time +import warnings + +from calendar import monthrange +from io import StringIO + +import six +from six import integer_types, text_type + +from decimal import Decimal + +from warnings import warn + +from .. import relativedelta +from .. import tz + +__all__ = ["parse", "parserinfo", "ParserError"] + + +# TODO: pandas.core.tools.datetimes imports this explicitly. Might be worth +# making public and/or figuring out if there is something we can +# take off their plate. +class _timelex(object): + # Fractional seconds are sometimes split by a comma + _split_decimal = re.compile("([.,])") + + def __init__(self, instream): + if isinstance(instream, (bytes, bytearray)): + instream = instream.decode() + + if isinstance(instream, text_type): + instream = StringIO(instream) + elif 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"), # TODO: "Tues" + ("Wed", "Wednesday"), + ("Thu", "Thursday"), # TODO: "Thurs" + ("Fri", "Friday"), + ("Sat", "Saturday"), + ("Sun", "Sunday")] + MONTHS = [("Jan", "January"), + ("Feb", "February"), # TODO: "Febr" + ("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", "z"] + PERTAIN = ["of"] + TZOFFSET = {} + # TODO: ERA = ["AD", "BC", "CE", "BCE", "Stardate", + # "Anno Domini", "Year of Our Lord"] + + 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): + try: + return self._weekdays[name.lower()] + except KeyError: + pass + return None + + def month(self, name): + 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): + """ + Converts two-digit years to year within [-50, 49] + range of self._year (current local time) + """ + + # Function contract is that the year is always positive + assert year >= 0 + + if year < 100 and not century_specified: + # assume current century to start + year += self._century + + if year >= self._year + 50: # if too far in future + year -= 100 + elif year < self._year - 50: # if too far in past + 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' 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, *args, **kwargs): + super(self.__class__, self).__init__(*args, **kwargs) + self.century_specified = False + self.dstridx = None + self.mstridx = None + self.ystridx = None + + @property + def has_year(self): + return self.ystridx is not None + + @property + def has_month(self): + return self.mstridx is not None + + @property + def has_day(self): + return self.dstridx is not None + + def could_be_day(self, value): + if self.has_day: + return False + elif not self.has_month: + return 1 <= value <= 31 + elif not self.has_year: + # Be permissive, assume leap year + month = self[self.mstridx] + return 1 <= value <= monthrange(2000, month)[1] + else: + month = self[self.mstridx] + year = self[self.ystridx] + return 1 <= value <= monthrange(year, month)[1] + + def append(self, val, label=None): + if hasattr(val, '__len__'): + if val.isdigit() and len(val) > 2: + self.century_specified = True + if label not in [None, 'Y']: # pragma: no cover + raise ValueError(label) + label = 'Y' + elif val > 100: + self.century_specified = True + if label not in [None, 'Y']: # pragma: no cover + raise ValueError(label) + label = 'Y' + + super(self.__class__, self).append(int(val)) + + if label == 'M': + if self.has_month: + raise ValueError('Month is already set') + self.mstridx = len(self) - 1 + elif label == 'D': + if self.has_day: + raise ValueError('Day is already set') + self.dstridx = len(self) - 1 + elif label == 'Y': + if self.has_year: + raise ValueError('Year is already set') + self.ystridx = len(self) - 1 + + def _resolve_from_stridxs(self, strids): + """ + Try to resolve the identities of year/month/day elements using + ystridx, mstridx, and dstridx, if enough of these are specified. + """ + if len(self) == 3 and len(strids) == 2: + # we can back out the remaining stridx value + missing = [x for x in range(3) if x not in strids.values()] + key = [x for x in ['y', 'm', 'd'] if x not in strids] + assert len(missing) == len(key) == 1 + key = key[0] + val = missing[0] + strids[key] = val + + assert len(self) == len(strids) # otherwise this should not be called + out = {key: self[strids[key]] for key in strids} + return (out.get('y'), out.get('m'), out.get('d')) + + def resolve_ymd(self, yearfirst, dayfirst): + len_ymd = len(self) + year, month, day = (None, None, None) + + strids = (('y', self.ystridx), + ('m', self.mstridx), + ('d', self.dstridx)) + + strids = {key: val for key, val in strids if val is not None} + if (len(self) == len(strids) > 0 or + (len(self) == 3 and len(strids) == 2)): + return self._resolve_from_stridxs(strids) + + mstridx = self.mstridx + + if len_ymd > 3: + raise ValueError("More than three YMD values") + elif len_ymd == 1 or (mstridx is not None and len_ymd == 2): + # One member, or two members with a month string + if mstridx is not None: + month = self[mstridx] + # since mstridx is 0 or 1, self[mstridx-1] always + # looks up the other element + other = self[mstridx - 1] + else: + other = self[0] + + if len_ymd > 1 or mstridx is None: + if other > 31: + year = other + else: + day = other + + 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: + if self[1] > 31: + # Apr-2003-25 + month, year, day = self + else: + 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 precedence 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.ystridx == 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 seconds or a :class:`tzinfo` object. + + .. doctest:: + :options: +NORMALIZE_WHITESPACE + + >>> from dateutil.parser import parse + >>> from dateutil.tz import gettz + >>> tzinfos = {"BRST": -7200, "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', -7200)) + >>> 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 ParserError: + 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 TypeError: + Raised for non-string or character stream input. + + :raises OverflowError: + Raised if the parsed date exceeds the largest valid C integer on + your system. + """ + + if default is None: + default = datetime.datetime.now().replace(hour=0, minute=0, + second=0, microsecond=0) + + res, skipped_tokens = self._parse(timestr, **kwargs) + + if res is None: + raise ParserError("Unknown string format: %s", timestr) + + if len(res) == 0: + raise ParserError("String does not contain a date: %s", timestr) + + try: + ret = self._build_naive(res, default) + except ValueError as e: + six.raise_from(ParserError(str(e) + ": %s", timestr), e) + + if not ignoretz: + ret = self._build_tzaware(ret, res, tzinfos) + + 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","any_unused_tokens"] + + 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 + + skipped_idxs = [] + + # year/month/day list + ymd = _ymd() + + len_l = len(l) + i = 0 + try: + while i < len_l: + + # Check if it's a number + value_repr = l[i] + try: + value = float(value_repr) + except ValueError: + value = None + + if value is not None: + # Numeric token + i = self._parse_numeric_token(l, i, info, ymd, res, fuzzy) + + # Check weekday + elif info.weekday(l[i]) is not None: + value = info.weekday(l[i]) + res.weekday = value + + # Check month name + elif info.month(l[i]) is not None: + value = info.month(l[i]) + ymd.append(value, 'M') + + if i + 1 < len_l: + if l[i + 1] in ('-', '/'): + # Jan-01[-99] + sep = l[i + 1] + ymd.append(l[i + 2]) + + if i + 3 < len_l and l[i + 3] == sep: + # Jan-01-99 + ymd.append(l[i + 4]) + i += 2 + + i += 2 + + elif (i + 4 < len_l and l[i + 1] == l[i + 3] == ' ' and + info.pertain(l[i + 2])): + # Jan of 01 + # In this case, 01 is clearly year + if l[i + 4].isdigit(): + # Convert it here to become unambiguous + value = int(l[i + 4]) + year = str(info.convertyear(value)) + ymd.append(year, 'Y') + else: + # Wrong guess + pass + # TODO: not hit in tests + i += 4 + + # Check am/pm + elif info.ampm(l[i]) is not None: + value = info.ampm(l[i]) + val_is_ampm = self._ampm_valid(res.hour, res.ampm, fuzzy) + + if val_is_ampm: + res.hour = self._adjust_ampm(res.hour, value) + res.ampm = value + + elif fuzzy: + skipped_idxs.append(i) + + # Check for a timezone name + elif self._could_be_tzname(res.hour, res.tzname, res.tzoffset, l[i]): + res.tzname = l[i] + res.tzoffset = info.tzoffset(res.tzname) + + # 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 + 1 < len_l and l[i + 1] in ('+', '-'): + l[i + 1] = ('+', '-')[l[i + 1] == '+'] + res.tzoffset = None + if info.utczone(res.tzname): + # With something like GMT+3, the timezone + # is *not* GMT. + res.tzname = None + + # Check for a numbered timezone + elif res.hour is not None and l[i] in ('+', '-'): + signal = (-1, 1)[l[i] == '+'] + len_li = len(l[i + 1]) + + # TODO: check that l[i + 1] is integer? + if len_li == 4: + # -0300 + hour_offset = int(l[i + 1][:2]) + min_offset = int(l[i + 1][2:]) + elif i + 2 < len_l and l[i + 2] == ':': + # -03:00 + hour_offset = int(l[i + 1]) + min_offset = int(l[i + 3]) # TODO: Check that l[i+3] is minute-like? + i += 2 + elif len_li <= 2: + # -[0]3 + hour_offset = int(l[i + 1][:2]) + min_offset = 0 + else: + raise ValueError(timestr) + + res.tzoffset = signal * (hour_offset * 3600 + min_offset * 60) + + # Look for a timezone name between parenthesis + if (i + 5 < len_l and + info.jump(l[i + 2]) and l[i + 3] == '(' and + l[i + 5] == ')' and + 3 <= len(l[i + 4]) and + self._could_be_tzname(res.hour, res.tzname, + None, l[i + 4])): + # -0300 (BRST) + res.tzname = l[i + 4] + i += 4 + + i += 1 + + # Check jumps + elif not (info.jump(l[i]) or fuzzy): + raise ValueError(timestr) + + else: + skipped_idxs.append(i) + i += 1 + + # Process year/month/day + year, month, day = ymd.resolve_ymd(yearfirst, dayfirst) + + res.century_specified = ymd.century_specified + res.year = year + res.month = month + res.day = day + + except (IndexError, ValueError): + return None, None + + if not info.validate(res): + return None, None + + if fuzzy_with_tokens: + skipped_tokens = self._recombine_skipped(l, skipped_idxs) + return res, tuple(skipped_tokens) + else: + return res, None + + def _parse_numeric_token(self, tokens, idx, info, ymd, res, fuzzy): + # Token is a number + value_repr = tokens[idx] + try: + value = self._to_decimal(value_repr) + except Exception as e: + six.raise_from(ValueError('Unknown numeric token'), e) + + len_li = len(value_repr) + + len_l = len(tokens) + + if (len(ymd) == 3 and len_li in (2, 4) and + res.hour is None and + (idx + 1 >= len_l or + (tokens[idx + 1] != ':' and + info.hms(tokens[idx + 1]) is None))): + # 19990101T23[59] + s = tokens[idx] + res.hour = int(s[:2]) + + if len_li == 4: + res.minute = int(s[2:]) + + elif len_li == 6 or (len_li > 6 and tokens[idx].find('.') == 6): + # YYMMDD or HHMMSS[.ss] + s = tokens[idx] + + if not ymd and '.' not in tokens[idx]: + ymd.append(s[:2]) + ymd.append(s[2:4]) + ymd.append(s[4:]) + else: + # 19990101T235959[.59] + + # TODO: Check if res attributes already set. + res.hour = int(s[:2]) + res.minute = int(s[2:4]) + res.second, res.microsecond = self._parsems(s[4:]) + + elif len_li in (8, 12, 14): + # YYYYMMDD + s = tokens[idx] + ymd.append(s[:4], 'Y') + 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 self._find_hms_idx(idx, tokens, info, allow_jump=True) is not None: + # HH[ ]h or MM[ ]m or SS[.ss][ ]s + hms_idx = self._find_hms_idx(idx, tokens, info, allow_jump=True) + (idx, hms) = self._parse_hms(idx, tokens, info, hms_idx) + if hms is not None: + # TODO: checking that hour/minute/second are not + # already set? + self._assign_hms(res, value_repr, hms) + + elif idx + 2 < len_l and tokens[idx + 1] == ':': + # HH:MM[:SS[.ss]] + res.hour = int(value) + value = self._to_decimal(tokens[idx + 2]) # TODO: try/except for this? + (res.minute, res.second) = self._parse_min_sec(value) + + if idx + 4 < len_l and tokens[idx + 3] == ':': + res.second, res.microsecond = self._parsems(tokens[idx + 4]) + + idx += 2 + + idx += 2 + + elif idx + 1 < len_l and tokens[idx + 1] in ('-', '/', '.'): + sep = tokens[idx + 1] + ymd.append(value_repr) + + if idx + 2 < len_l and not info.jump(tokens[idx + 2]): + if tokens[idx + 2].isdigit(): + # 01-01[-01] + ymd.append(tokens[idx + 2]) + else: + # 01-Jan[-01] + value = info.month(tokens[idx + 2]) + + if value is not None: + ymd.append(value, 'M') + else: + raise ValueError() + + if idx + 3 < len_l and tokens[idx + 3] == sep: + # We have three members + value = info.month(tokens[idx + 4]) + + if value is not None: + ymd.append(value, 'M') + else: + ymd.append(tokens[idx + 4]) + idx += 2 + + idx += 1 + idx += 1 + + elif idx + 1 >= len_l or info.jump(tokens[idx + 1]): + if idx + 2 < len_l and info.ampm(tokens[idx + 2]) is not None: + # 12 am + hour = int(value) + res.hour = self._adjust_ampm(hour, info.ampm(tokens[idx + 2])) + idx += 1 + else: + # Year, month or day + ymd.append(value) + idx += 1 + + elif info.ampm(tokens[idx + 1]) is not None and (0 <= value < 24): + # 12am + hour = int(value) + res.hour = self._adjust_ampm(hour, info.ampm(tokens[idx + 1])) + idx += 1 + + elif ymd.could_be_day(value): + ymd.append(value) + + elif not fuzzy: + raise ValueError() + + return idx + + def _find_hms_idx(self, idx, tokens, info, allow_jump): + len_l = len(tokens) + + if idx+1 < len_l and info.hms(tokens[idx+1]) is not None: + # There is an "h", "m", or "s" label following this token. We take + # assign the upcoming label to the current token. + # e.g. the "12" in 12h" + hms_idx = idx + 1 + + elif (allow_jump and idx+2 < len_l and tokens[idx+1] == ' ' and + info.hms(tokens[idx+2]) is not None): + # There is a space and then an "h", "m", or "s" label. + # e.g. the "12" in "12 h" + hms_idx = idx + 2 + + elif idx > 0 and info.hms(tokens[idx-1]) is not None: + # There is a "h", "m", or "s" preceding this token. Since neither + # of the previous cases was hit, there is no label following this + # token, so we use the previous label. + # e.g. the "04" in "12h04" + hms_idx = idx-1 + + elif (1 < idx == len_l-1 and tokens[idx-1] == ' ' and + info.hms(tokens[idx-2]) is not None): + # If we are looking at the final token, we allow for a + # backward-looking check to skip over a space. + # TODO: Are we sure this is the right condition here? + hms_idx = idx - 2 + + else: + hms_idx = None + + return hms_idx + + def _assign_hms(self, res, value_repr, hms): + # See GH issue #427, fixing float rounding + value = self._to_decimal(value_repr) + + if hms == 0: + # Hour + res.hour = int(value) + if value % 1: + res.minute = int(60*(value % 1)) + + elif hms == 1: + (res.minute, res.second) = self._parse_min_sec(value) + + elif hms == 2: + (res.second, res.microsecond) = self._parsems(value_repr) + + def _could_be_tzname(self, hour, tzname, tzoffset, token): + return (hour is not None and + tzname is None and + tzoffset is None and + len(token) <= 5 and + (all(x in string.ascii_uppercase for x in token) + or token in self.info.UTCZONE)) + + def _ampm_valid(self, hour, ampm, fuzzy): + """ + 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 ampm is not None: + val_is_ampm = False + + # If AM/PM is found and hour is not, raise a ValueError + if hour is None: + if fuzzy: + val_is_ampm = False + else: + raise ValueError('No hour specified with AM or PM flag.') + elif not 0 <= 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.') + + return val_is_ampm + + def _adjust_ampm(self, hour, ampm): + if hour < 12 and ampm == 1: + hour += 12 + elif hour == 12 and ampm == 0: + hour = 0 + return hour + + def _parse_min_sec(self, value): + # TODO: Every usage of this function sets res.second to the return + # value. Are there any cases where second will be returned as None and + # we *don't* want to set res.second = None? + minute = int(value) + second = None + + sec_remainder = value % 1 + if sec_remainder: + second = int(60 * sec_remainder) + return (minute, second) + + def _parse_hms(self, idx, tokens, info, hms_idx): + # TODO: Is this going to admit a lot of false-positives for when we + # just happen to have digits and "h", "m" or "s" characters in non-date + # text? I guess hex hashes won't have that problem, but there's plenty + # of random junk out there. + if hms_idx is None: + hms = None + new_idx = idx + elif hms_idx > idx: + hms = info.hms(tokens[hms_idx]) + new_idx = hms_idx + else: + # Looking backwards, increment one. + hms = info.hms(tokens[hms_idx]) + 1 + new_idx = idx + + return (new_idx, hms) + + # ------------------------------------------------------------------ + # Handling for individual tokens. These are kept as methods instead + # of functions for the sake of customizability via subclassing. + + def _parsems(self, 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]) + + def _to_decimal(self, val): + try: + decimal_value = Decimal(val) + # See GH 662, edge case, infinite value should not be converted + # via `_to_decimal` + if not decimal_value.is_finite(): + raise ValueError("Converted decimal value is infinite or NaN") + except Exception as e: + msg = "Could not convert %s to decimal" % val + six.raise_from(ValueError(msg), e) + else: + return decimal_value + + # ------------------------------------------------------------------ + # Post-Parsing construction of datetime output. These are kept as + # methods instead of functions for the sake of customizability via + # subclassing. + + def _build_tzinfo(self, tzinfos, tzname, tzoffset): + if callable(tzinfos): + tzdata = tzinfos(tzname, tzoffset) + else: + tzdata = tzinfos.get(tzname) + # handle case where tzinfo is paased an options that returns None + # eg tzinfos = {'BRST' : None} + if isinstance(tzdata, datetime.tzinfo) or tzdata is None: + tzinfo = tzdata + elif isinstance(tzdata, text_type): + tzinfo = tz.tzstr(tzdata) + elif isinstance(tzdata, integer_types): + tzinfo = tz.tzoffset(tzname, tzdata) + else: + raise TypeError("Offset must be tzinfo subclass, tz string, " + "or int offset.") + return tzinfo + + def _build_tzaware(self, naive, res, tzinfos): + if (callable(tzinfos) or (tzinfos and res.tzname in tzinfos)): + tzinfo = self._build_tzinfo(tzinfos, res.tzname, res.tzoffset) + aware = naive.replace(tzinfo=tzinfo) + aware = self._assign_tzname(aware, res.tzname) + + elif res.tzname and res.tzname in time.tzname: + aware = naive.replace(tzinfo=tz.tzlocal()) + + # Handle ambiguous local datetime + aware = self._assign_tzname(aware, res.tzname) + + # This is mostly relevant for winter GMT zones parsed in the UK + if (aware.tzname() != res.tzname and + res.tzname in self.info.UTCZONE): + aware = aware.replace(tzinfo=tz.UTC) + + elif res.tzoffset == 0: + aware = naive.replace(tzinfo=tz.UTC) + + elif res.tzoffset: + aware = naive.replace(tzinfo=tz.tzoffset(res.tzname, res.tzoffset)) + + elif not res.tzname and not res.tzoffset: + # i.e. no timezone information was found. + aware = naive + + elif res.tzname: + # tz-like string was parsed but we don't know what to do + # with it + warnings.warn("tzname {tzname} identified but not understood. " + "Pass `tzinfos` argument in order to correctly " + "return a timezone-aware datetime. In a future " + "version, this will raise an " + "exception.".format(tzname=res.tzname), + category=UnknownTimezoneWarning) + aware = naive + + return aware + + def _build_naive(self, res, default): + 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] + + naive = default.replace(**repl) + + if res.weekday is not None and not res.day: + naive = naive + relativedelta.relativedelta(weekday=res.weekday) + + return naive + + def _assign_tzname(self, dt, tzname): + if dt.tzname() != tzname: + new_dt = tz.enfold(dt, fold=1) + if new_dt.tzname() == tzname: + return new_dt + + return dt + + def _recombine_skipped(self, tokens, skipped_idxs): + """ + >>> tokens = ["foo", " ", "bar", " ", "19June2000", "baz"] + >>> skipped_idxs = [0, 1, 2, 5] + >>> _recombine_skipped(tokens, skipped_idxs) + ["foo bar", "baz"] + """ + skipped_tokens = [] + for i, idx in enumerate(sorted(skipped_idxs)): + if i > 0 and idx - 1 == skipped_idxs[i - 1]: + skipped_tokens[-1] = skipped_tokens[-1] + tokens[idx] + else: + skipped_tokens.append(tokens[idx]) + + return skipped_tokens + + +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 seconds or a :class:`tzinfo` object. + + .. doctest:: + :options: +NORMALIZE_WHITESPACE + + >>> from dateutil.parser import parse + >>> from dateutil.tz import gettz + >>> tzinfos = {"BRST": -7200, "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', -7200)) + >>> 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(2047, 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 ParserError: + Raised for invalid or unknown string formats, 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 = [x for x in re.split(r'([,:.]|[a-zA-Z]+|[0-9]+)',tzstr) if x] + used_idxs = list() + 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]) + + for ii in range(j): + used_idxs.append(ii) + 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] == '+'] + used_idxs.append(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) + used_idxs.append(i) + i += 2 + elif len_li <= 2: + # -[0]3 + setattr(res, offattr, + int(l[i][:2]) * 3600 * signal) + else: + return None + used_idxs.append(i) + 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]) + used_idxs.append(i) + i += 2 + if l[i] == '-': + value = int(l[i + 1]) * -1 + used_idxs.append(i) + i += 1 + else: + value = int(l[i]) + used_idxs.append(i) + i += 2 + if value: + x.week = value + x.weekday = (int(l[i]) - 1) % 7 + else: + x.day = int(l[i]) + used_idxs.append(i) + i += 2 + x.time = int(l[i]) + used_idxs.append(i) + i += 2 + if i < len_l: + if l[i] in ('-', '+'): + signal = (-1, 1)[l[i] == "+"] + used_idxs.append(i) + i += 1 + else: + signal = 1 + used_idxs.append(i) + res.dstoffset = (res.stdoffset + int(l[i]) * signal) + + # This was a made-up format that is not in normal use + warn(('Parsed time zone "%s"' % tzstr) + + 'is in a non-standard dateutil-specific format, which ' + + 'is now deprecated; support for parsing this format ' + + 'will be removed in future versions. It is recommended ' + + 'that you switch to a standard format like the GNU ' + + 'TZ variable format.', tz.DeprecatedTzFormatWarning) + 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) + used_idxs.append(i) + i += 1 + x.jyday = int(l[i]) + elif l[i] == 'M': + # month[-.]week[-.]weekday + used_idxs.append(i) + i += 1 + x.month = int(l[i]) + used_idxs.append(i) + i += 1 + assert l[i] in ('-', '.') + used_idxs.append(i) + i += 1 + x.week = int(l[i]) + if x.week == 5: + x.week = -1 + used_idxs.append(i) + i += 1 + assert l[i] in ('-', '.') + used_idxs.append(i) + i += 1 + x.weekday = (int(l[i]) - 1) % 7 + else: + # year day (zero based) + x.yday = int(l[i]) + 1 + + used_idxs.append(i) + i += 1 + + if i < len_l and l[i] == '/': + used_idxs.append(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 + used_idxs.append(i) + i += 2 + if i + 1 < len_l and l[i + 1] == ':': + used_idxs.append(i) + i += 2 + x.time += int(l[i]) + elif len_li <= 2: + # -[0]3 + x.time = (int(l[i][:2]) * 3600) + else: + return None + used_idxs.append(i) + i += 1 + + assert i == len_l or l[i] == ',' + + i += 1 + + assert i >= len_l + + except (IndexError, ValueError, AssertionError): + return None + + unused_idxs = set(range(len_l)).difference(used_idxs) + res.any_unused_tokens = not {l[n] for n in unused_idxs}.issubset({",",":"}) + return res + + +DEFAULTTZPARSER = _tzparser() + + +def _parsetz(tzstr): + return DEFAULTTZPARSER.parse(tzstr) + + +class ParserError(ValueError): + """Exception subclass used for any failure to parse a datetime string. + + This is a subclass of :py:exc:`ValueError`, and should be raised any time + earlier versions of ``dateutil`` would have raised ``ValueError``. + + .. versionadded:: 2.8.1 + """ + def __str__(self): + try: + return self.args[0] % self.args[1:] + except (TypeError, IndexError): + return super(ParserError, self).__str__() + + def __repr__(self): + args = ", ".join("'%s'" % arg for arg in self.args) + return "%s(%s)" % (self.__class__.__name__, args) + + +class UnknownTimezoneWarning(RuntimeWarning): + """Raised when the parser finds a timezone it cannot parse into a tzinfo. + + .. versionadded:: 2.7.0 + """ +# vim:ts=4:sw=4:et diff --git a/.venv/lib/python3.12/site-packages/dateutil/parser/isoparser.py b/.venv/lib/python3.12/site-packages/dateutil/parser/isoparser.py new file mode 100644 index 00000000..7060087d --- /dev/null +++ b/.venv/lib/python3.12/site-packages/dateutil/parser/isoparser.py @@ -0,0 +1,416 @@ +# -*- coding: utf-8 -*- +""" +This module offers a parser for ISO-8601 strings + +It is intended to support all valid date, time and datetime formats per the +ISO-8601 specification. + +..versionadded:: 2.7.0 +""" +from datetime import datetime, timedelta, time, date +import calendar +from dateutil import tz + +from functools import wraps + +import re +import six + +__all__ = ["isoparse", "isoparser"] + + +def _takes_ascii(f): + @wraps(f) + def func(self, str_in, *args, **kwargs): + # If it's a stream, read the whole thing + str_in = getattr(str_in, 'read', lambda: str_in)() + + # If it's unicode, turn it into bytes, since ISO-8601 only covers ASCII + if isinstance(str_in, six.text_type): + # ASCII is the same in UTF-8 + try: + str_in = str_in.encode('ascii') + except UnicodeEncodeError as e: + msg = 'ISO-8601 strings should contain only ASCII characters' + six.raise_from(ValueError(msg), e) + + return f(self, str_in, *args, **kwargs) + + return func + + +class isoparser(object): + def __init__(self, sep=None): + """ + :param sep: + A single character that separates date and time portions. If + ``None``, the parser will accept any single character. + For strict ISO-8601 adherence, pass ``'T'``. + """ + if sep is not None: + if (len(sep) != 1 or ord(sep) >= 128 or sep in '0123456789'): + raise ValueError('Separator must be a single, non-numeric ' + + 'ASCII character') + + sep = sep.encode('ascii') + + self._sep = sep + + @_takes_ascii + def isoparse(self, dt_str): + """ + Parse an ISO-8601 datetime string into a :class:`datetime.datetime`. + + An ISO-8601 datetime string consists of a date portion, followed + optionally by a time portion - the date and time portions are separated + by a single character separator, which is ``T`` in the official + standard. Incomplete date formats (such as ``YYYY-MM``) may *not* be + combined with a time portion. + + Supported date formats are: + + Common: + + - ``YYYY`` + - ``YYYY-MM`` + - ``YYYY-MM-DD`` or ``YYYYMMDD`` + + Uncommon: + + - ``YYYY-Www`` or ``YYYYWww`` - ISO week (day defaults to 0) + - ``YYYY-Www-D`` or ``YYYYWwwD`` - ISO week and day + + The ISO week and day numbering follows the same logic as + :func:`datetime.date.isocalendar`. + + Supported time formats are: + + - ``hh`` + - ``hh:mm`` or ``hhmm`` + - ``hh:mm:ss`` or ``hhmmss`` + - ``hh:mm:ss.ssssss`` (Up to 6 sub-second digits) + + Midnight is a special case for `hh`, as the standard supports both + 00:00 and 24:00 as a representation. The decimal separator can be + either a dot or a comma. + + + .. caution:: + + Support for fractional components other than seconds is part of the + ISO-8601 standard, but is not currently implemented in this parser. + + Supported time zone offset formats are: + + - `Z` (UTC) + - `±HH:MM` + - `±HHMM` + - `±HH` + + Offsets will be represented as :class:`dateutil.tz.tzoffset` objects, + with the exception of UTC, which will be represented as + :class:`dateutil.tz.tzutc`. Time zone offsets equivalent to UTC (such + as `+00:00`) will also be represented as :class:`dateutil.tz.tzutc`. + + :param dt_str: + A string or stream containing only an ISO-8601 datetime string + + :return: + Returns a :class:`datetime.datetime` representing the string. + Unspecified components default to their lowest value. + + .. warning:: + + As of version 2.7.0, the strictness of the parser should not be + considered a stable part of the contract. Any valid ISO-8601 string + that parses correctly with the default settings will continue to + parse correctly in future versions, but invalid strings that + currently fail (e.g. ``2017-01-01T00:00+00:00:00``) are not + guaranteed to continue failing in future versions if they encode + a valid date. + + .. versionadded:: 2.7.0 + """ + components, pos = self._parse_isodate(dt_str) + + if len(dt_str) > pos: + if self._sep is None or dt_str[pos:pos + 1] == self._sep: + components += self._parse_isotime(dt_str[pos + 1:]) + else: + raise ValueError('String contains unknown ISO components') + + if len(components) > 3 and components[3] == 24: + components[3] = 0 + return datetime(*components) + timedelta(days=1) + + return datetime(*components) + + @_takes_ascii + def parse_isodate(self, datestr): + """ + Parse the date portion of an ISO string. + + :param datestr: + The string portion of an ISO string, without a separator + + :return: + Returns a :class:`datetime.date` object + """ + components, pos = self._parse_isodate(datestr) + if pos < len(datestr): + raise ValueError('String contains unknown ISO ' + + 'components: {!r}'.format(datestr.decode('ascii'))) + return date(*components) + + @_takes_ascii + def parse_isotime(self, timestr): + """ + Parse the time portion of an ISO string. + + :param timestr: + The time portion of an ISO string, without a separator + + :return: + Returns a :class:`datetime.time` object + """ + components = self._parse_isotime(timestr) + if components[0] == 24: + components[0] = 0 + return time(*components) + + @_takes_ascii + def parse_tzstr(self, tzstr, zero_as_utc=True): + """ + Parse a valid ISO time zone string. + + See :func:`isoparser.isoparse` for details on supported formats. + + :param tzstr: + A string representing an ISO time zone offset + + :param zero_as_utc: + Whether to return :class:`dateutil.tz.tzutc` for zero-offset zones + + :return: + Returns :class:`dateutil.tz.tzoffset` for offsets and + :class:`dateutil.tz.tzutc` for ``Z`` and (if ``zero_as_utc`` is + specified) offsets equivalent to UTC. + """ + return self._parse_tzstr(tzstr, zero_as_utc=zero_as_utc) + + # Constants + _DATE_SEP = b'-' + _TIME_SEP = b':' + _FRACTION_REGEX = re.compile(b'[\\.,]([0-9]+)') + + def _parse_isodate(self, dt_str): + try: + return self._parse_isodate_common(dt_str) + except ValueError: + return self._parse_isodate_uncommon(dt_str) + + def _parse_isodate_common(self, dt_str): + len_str = len(dt_str) + components = [1, 1, 1] + + if len_str < 4: + raise ValueError('ISO string too short') + + # Year + components[0] = int(dt_str[0:4]) + pos = 4 + if pos >= len_str: + return components, pos + + has_sep = dt_str[pos:pos + 1] == self._DATE_SEP + if has_sep: + pos += 1 + + # Month + if len_str - pos < 2: + raise ValueError('Invalid common month') + + components[1] = int(dt_str[pos:pos + 2]) + pos += 2 + + if pos >= len_str: + if has_sep: + return components, pos + else: + raise ValueError('Invalid ISO format') + + if has_sep: + if dt_str[pos:pos + 1] != self._DATE_SEP: + raise ValueError('Invalid separator in ISO string') + pos += 1 + + # Day + if len_str - pos < 2: + raise ValueError('Invalid common day') + components[2] = int(dt_str[pos:pos + 2]) + return components, pos + 2 + + def _parse_isodate_uncommon(self, dt_str): + if len(dt_str) < 4: + raise ValueError('ISO string too short') + + # All ISO formats start with the year + year = int(dt_str[0:4]) + + has_sep = dt_str[4:5] == self._DATE_SEP + + pos = 4 + has_sep # Skip '-' if it's there + if dt_str[pos:pos + 1] == b'W': + # YYYY-?Www-?D? + pos += 1 + weekno = int(dt_str[pos:pos + 2]) + pos += 2 + + dayno = 1 + if len(dt_str) > pos: + if (dt_str[pos:pos + 1] == self._DATE_SEP) != has_sep: + raise ValueError('Inconsistent use of dash separator') + + pos += has_sep + + dayno = int(dt_str[pos:pos + 1]) + pos += 1 + + base_date = self._calculate_weekdate(year, weekno, dayno) + else: + # YYYYDDD or YYYY-DDD + if len(dt_str) - pos < 3: + raise ValueError('Invalid ordinal day') + + ordinal_day = int(dt_str[pos:pos + 3]) + pos += 3 + + if ordinal_day < 1 or ordinal_day > (365 + calendar.isleap(year)): + raise ValueError('Invalid ordinal day' + + ' {} for year {}'.format(ordinal_day, year)) + + base_date = date(year, 1, 1) + timedelta(days=ordinal_day - 1) + + components = [base_date.year, base_date.month, base_date.day] + return components, pos + + def _calculate_weekdate(self, year, week, day): + """ + Calculate the day of corresponding to the ISO year-week-day calendar. + + This function is effectively the inverse of + :func:`datetime.date.isocalendar`. + + :param year: + The year in the ISO calendar + + :param week: + The week in the ISO calendar - range is [1, 53] + + :param day: + The day in the ISO calendar - range is [1 (MON), 7 (SUN)] + + :return: + Returns a :class:`datetime.date` + """ + if not 0 < week < 54: + raise ValueError('Invalid week: {}'.format(week)) + + if not 0 < day < 8: # Range is 1-7 + raise ValueError('Invalid weekday: {}'.format(day)) + + # Get week 1 for the specific year: + jan_4 = date(year, 1, 4) # Week 1 always has January 4th in it + week_1 = jan_4 - timedelta(days=jan_4.isocalendar()[2] - 1) + + # Now add the specific number of weeks and days to get what we want + week_offset = (week - 1) * 7 + (day - 1) + return week_1 + timedelta(days=week_offset) + + def _parse_isotime(self, timestr): + len_str = len(timestr) + components = [0, 0, 0, 0, None] + pos = 0 + comp = -1 + + if len_str < 2: + raise ValueError('ISO time too short') + + has_sep = False + + while pos < len_str and comp < 5: + comp += 1 + + if timestr[pos:pos + 1] in b'-+Zz': + # Detect time zone boundary + components[-1] = self._parse_tzstr(timestr[pos:]) + pos = len_str + break + + if comp == 1 and timestr[pos:pos+1] == self._TIME_SEP: + has_sep = True + pos += 1 + elif comp == 2 and has_sep: + if timestr[pos:pos+1] != self._TIME_SEP: + raise ValueError('Inconsistent use of colon separator') + pos += 1 + + if comp < 3: + # Hour, minute, second + components[comp] = int(timestr[pos:pos + 2]) + pos += 2 + + if comp == 3: + # Fraction of a second + frac = self._FRACTION_REGEX.match(timestr[pos:]) + if not frac: + continue + + us_str = frac.group(1)[:6] # Truncate to microseconds + components[comp] = int(us_str) * 10**(6 - len(us_str)) + pos += len(frac.group()) + + if pos < len_str: + raise ValueError('Unused components in ISO string') + + if components[0] == 24: + # Standard supports 00:00 and 24:00 as representations of midnight + if any(component != 0 for component in components[1:4]): + raise ValueError('Hour may only be 24 at 24:00:00.000') + + return components + + def _parse_tzstr(self, tzstr, zero_as_utc=True): + if tzstr == b'Z' or tzstr == b'z': + return tz.UTC + + if len(tzstr) not in {3, 5, 6}: + raise ValueError('Time zone offset must be 1, 3, 5 or 6 characters') + + if tzstr[0:1] == b'-': + mult = -1 + elif tzstr[0:1] == b'+': + mult = 1 + else: + raise ValueError('Time zone offset requires sign') + + hours = int(tzstr[1:3]) + if len(tzstr) == 3: + minutes = 0 + else: + minutes = int(tzstr[(4 if tzstr[3:4] == self._TIME_SEP else 3):]) + + if zero_as_utc and hours == 0 and minutes == 0: + return tz.UTC + else: + if minutes > 59: + raise ValueError('Invalid minutes in time zone offset') + + if hours > 23: + raise ValueError('Invalid hours in time zone offset') + + return tz.tzoffset(None, mult * (hours * 60 + minutes) * 60) + + +DEFAULT_ISOPARSER = isoparser() +isoparse = DEFAULT_ISOPARSER.isoparse |