diff options
author | Michael Twomey <michael.twomey@fieldaware.com> | 2013-08-29 15:01:13 +0100 |
---|---|---|
committer | Michael Twomey <michael.twomey@fieldaware.com> | 2013-08-29 15:01:13 +0100 |
commit | 488def46bf9fbc38f54b0ae6908364561116d430 (patch) | |
tree | 261a85a943989c3ed5705bec4c1d2c18d461796b /iso8601 | |
download | pyiso8601-488def46bf9fbc38f54b0ae6908364561116d430.tar.gz |
Importing r23 / 0.1.4 from https://code.google.com/p/pyiso8601/
Diffstat (limited to 'iso8601')
-rw-r--r-- | iso8601/__init__.py | 1 | ||||
-rw-r--r-- | iso8601/iso8601.py | 102 | ||||
-rw-r--r-- | iso8601/test_iso8601.py | 111 |
3 files changed, 214 insertions, 0 deletions
diff --git a/iso8601/__init__.py b/iso8601/__init__.py new file mode 100644 index 0000000..e72e356 --- /dev/null +++ b/iso8601/__init__.py @@ -0,0 +1 @@ +from iso8601 import * diff --git a/iso8601/iso8601.py b/iso8601/iso8601.py new file mode 100644 index 0000000..f923938 --- /dev/null +++ b/iso8601/iso8601.py @@ -0,0 +1,102 @@ +"""ISO 8601 date time string parsing + +Basic usage: +>>> import iso8601 +>>> iso8601.parse_date("2007-01-25T12:00:00Z") +datetime.datetime(2007, 1, 25, 12, 0, tzinfo=<iso8601.iso8601.Utc ...>) +>>> + +""" + +from datetime import datetime, timedelta, tzinfo +import re + +__all__ = ["parse_date", "ParseError"] + +# Adapted from http://delete.me.uk/2005/03/iso8601.html +ISO8601_REGEX = re.compile(r"(?P<year>[0-9]{4})(-(?P<month>[0-9]{1,2})(-(?P<day>[0-9]{1,2})" + r"((?P<separator>.)(?P<hour>[0-9]{2}):(?P<minute>[0-9]{2})(:(?P<second>[0-9]{2})(\.(?P<fraction>[0-9]+))?)?" + r"(?P<timezone>Z|(([-+])([0-9]{2}):([0-9]{2})))?)?)?)?" +) +TIMEZONE_REGEX = re.compile("(?P<prefix>[+-])(?P<hours>[0-9]{2}).(?P<minutes>[0-9]{2})") + +class ParseError(Exception): + """Raised when there is a problem parsing a date string""" + +# Yoinked from python docs +ZERO = timedelta(0) +class Utc(tzinfo): + """UTC + + """ + def utcoffset(self, dt): + return ZERO + + def tzname(self, dt): + return "UTC" + + def dst(self, dt): + return ZERO +UTC = Utc() + +class FixedOffset(tzinfo): + """Fixed offset in hours and minutes from UTC + + """ + def __init__(self, offset_hours, offset_minutes, name): + self.__offset = timedelta(hours=offset_hours, minutes=offset_minutes) + self.__name = name + + def utcoffset(self, dt): + return self.__offset + + def tzname(self, dt): + return self.__name + + def dst(self, dt): + return ZERO + + def __repr__(self): + return "<FixedOffset %r>" % self.__name + +def parse_timezone(tzstring, default_timezone=UTC): + """Parses ISO 8601 time zone specs into tzinfo offsets + + """ + if tzstring == "Z": + return default_timezone + # This isn't strictly correct, but it's common to encounter dates without + # timezones so I'll assume the default (which defaults to UTC). + # Addresses issue 4. + if tzstring is None: + return default_timezone + m = TIMEZONE_REGEX.match(tzstring) + prefix, hours, minutes = m.groups() + hours, minutes = int(hours), int(minutes) + if prefix == "-": + hours = -hours + minutes = -minutes + return FixedOffset(hours, minutes, tzstring) + +def parse_date(datestring, default_timezone=UTC): + """Parses ISO 8601 dates into datetime objects + + The timezone is parsed from the date string. However it is quite common to + have dates without a timezone (not strictly correct). In this case the + default timezone specified in default_timezone is used. This is UTC by + default. + """ + if not isinstance(datestring, basestring): + raise ParseError("Expecting a string %r" % datestring) + m = ISO8601_REGEX.match(datestring) + if not m: + raise ParseError("Unable to parse date string %r" % datestring) + groups = m.groupdict() + tz = parse_timezone(groups["timezone"], default_timezone=default_timezone) + if groups["fraction"] is None: + groups["fraction"] = 0 + else: + groups["fraction"] = int(float("0.%s" % groups["fraction"]) * 1e6) + return datetime(int(groups["year"]), int(groups["month"]), int(groups["day"]), + int(groups["hour"]), int(groups["minute"]), int(groups["second"]), + int(groups["fraction"]), tz) diff --git a/iso8601/test_iso8601.py b/iso8601/test_iso8601.py new file mode 100644 index 0000000..ff9e273 --- /dev/null +++ b/iso8601/test_iso8601.py @@ -0,0 +1,111 @@ +import iso8601 + +def test_iso8601_regex(): + assert iso8601.ISO8601_REGEX.match("2006-10-11T00:14:33Z") + +def test_timezone_regex(): + assert iso8601.TIMEZONE_REGEX.match("+01:00") + assert iso8601.TIMEZONE_REGEX.match("+00:00") + assert iso8601.TIMEZONE_REGEX.match("+01:20") + assert iso8601.TIMEZONE_REGEX.match("-01:00") + +def test_parse_date(): + d = iso8601.parse_date("2006-10-20T15:34:56Z") + assert d.year == 2006 + assert d.month == 10 + assert d.day == 20 + assert d.hour == 15 + assert d.minute == 34 + assert d.second == 56 + assert d.tzinfo == iso8601.UTC + +def test_parse_date_fraction(): + d = iso8601.parse_date("2006-10-20T15:34:56.123Z") + assert d.year == 2006 + assert d.month == 10 + assert d.day == 20 + assert d.hour == 15 + assert d.minute == 34 + assert d.second == 56 + assert d.microsecond == 123000 + assert d.tzinfo == iso8601.UTC + +def test_parse_date_fraction_2(): + """From bug 6 + + """ + d = iso8601.parse_date("2007-5-7T11:43:55.328Z'") + assert d.year == 2007 + assert d.month == 5 + assert d.day == 7 + assert d.hour == 11 + assert d.minute == 43 + assert d.second == 55 + assert d.microsecond == 328000 + assert d.tzinfo == iso8601.UTC + +def test_parse_date_tz(): + d = iso8601.parse_date("2006-10-20T15:34:56.123+02:30") + assert d.year == 2006 + assert d.month == 10 + assert d.day == 20 + assert d.hour == 15 + assert d.minute == 34 + assert d.second == 56 + assert d.microsecond == 123000 + assert d.tzinfo.tzname(None) == "+02:30" + offset = d.tzinfo.utcoffset(None) + assert offset.days == 0 + assert offset.seconds == 60 * 60 * 2.5 + +def test_parse_invalid_date(): + try: + iso8601.parse_date(None) + except iso8601.ParseError: + pass + else: + assert 1 == 2 + +def test_parse_invalid_date2(): + try: + iso8601.parse_date("23") + except iso8601.ParseError: + pass + else: + assert 1 == 2 + +def test_parse_no_timezone(): + """issue 4 - Handle datetime string without timezone + + This tests what happens when you parse a date with no timezone. While not + strictly correct this is quite common. I'll assume UTC for the time zone + in this case. + """ + d = iso8601.parse_date("2007-01-01T08:00:00") + assert d.year == 2007 + assert d.month == 1 + assert d.day == 1 + assert d.hour == 8 + assert d.minute == 0 + assert d.second == 0 + assert d.microsecond == 0 + assert d.tzinfo == iso8601.UTC + +def test_parse_no_timezone_different_default(): + tz = iso8601.FixedOffset(2, 0, "test offset") + d = iso8601.parse_date("2007-01-01T08:00:00", default_timezone=tz) + assert d.tzinfo == tz + +def test_space_separator(): + """Handle a separator other than T + + """ + d = iso8601.parse_date("2007-06-23 06:40:34.00Z") + assert d.year == 2007 + assert d.month == 6 + assert d.day == 23 + assert d.hour == 6 + assert d.minute == 40 + assert d.second == 34 + assert d.microsecond == 0 + assert d.tzinfo == iso8601.UTC |