Question

I have strings like these in my Python program and I added the wanted result as requested:

"Sat 1 Dec - 11h + 14h / Sun 2 Dec - 12h30"
("Sat 1 Dec 11h", "Sat 1 Dec 14h", "Sun 2 Dec 12h30")
"Tue 27 + Wed 28 Nov - 20h30"
("Tue 27 Nov 20h30", "Wed 28 Nov 20h30")
"Fri 4 + Sat 5 Jan - 20h30"
("Fri 4 Jan 20h30", "Sat 5 Jan 20h30")
"Wed 23 Jan - 20h"
("Wed 23 Jan 20h")
"Sat 26 Jan - 11h + 14h / Sun 27 Jan - 11h"
("Sat 26 Jan 11h", "Sat 26 Jan 14h", "Sun 27 Jan 11h")
"Fri 8 and Sat 9 Feb - 20h30 + thu 1 feb - 15h"
("Fri 8 Feb 20h30", "Sat 9 Feb 20h30", "Thu 1 feb 15h")
"Sat 2 Mar - 11h + 14h / Sun 3 Mar - 11h"
("Sat 2 Mar 11h", "Sat 2 Mar 14h", "Sun 3 Mar 11h")
"Wed 12, Thu 13, Fri 14 and Sat 15 Jun - 19h + Sun 16 Jun - 12h30"
("Wed 12 Jun 19h", "Thu 13 Jun 19h", "Fri 14 Jun 19h", "Sat 15 Jun 19h", "Sun 16 Jun 12h30") 

and with these two regex I can finde the 3 dates of the first string:

(Mon|Tue|Wed|Thu|Fri|Sat|Sun)\s([0-9]{1,2}\s(?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec))(?:.*?)([0-9]{1,2}[uh\:](?:[0-9]{2})?)

(Mon|Tue|Wed|Thu|Fri|Sat|Sun)\s([0-9]{1,2}\s(?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec))(?:.*?\+\s)([0-9]{1,2}[uh\:](?:[0-9]{2})?)

Is it possible to get all the dates from these strings with one or two regex patterns (to match all of them). So I think what it needs to do: finding the first following month for each date if not given, get the corresponding time and if followed by multiple hours make multiple datetimes per date.
Formatting is not that important.

Answer 1

I got you started. This is my interpretation of your problem. I leave the implementation of parse_complex up to you.

class DateParser(object):
    """parse dates according to the custom rules here:

    >>> DateParser("Sat 1 Dec - 11h + 14h / Sun 2 Dec - 12h30").parse()
    ("Sat 1 Dec 11h", "Sat 1 Dec 14h", "Sun 2 Dec 12h30")
    >>> DateParser("Tue 27 + Wed 28 Nov - 20h30").parse()
    ("Tue 27 Nov 20h30", "Wed 28 Nov 20h30")
    >>> DateParser("Fri 4 + Sat 5 Jan - 20h30").parse()
    ("Fri 4 Jan 20h30", "Sat 5 Jan 20h30")
    >>> DateParser("Wed 23 Jan - 20h").parse()
    ("Wed 23 Jan 20h")
    >>> DateParser("Sat 26 Jan - 11h + 14h / Sun 27 Jan - 11h").parse()
    ("Sat 26 Jan 11h", "Sat 26 Jan 14h", "Sun 27 Jan 11h")
    >>> DateParser("Fri 8 and Sat 9 Feb - 20h30 + thu 1 feb - 15h").parse()
    ("Fri 8 Feb 20h30", "Sat 9 Feb 20h30", "Thu 1 feb 15h")
    >>> DateParse("Sat 2 Mar - 11h + 14h / Sun 3 Mar - 11h").parse()
    ("Sat 2 Mar 11h", "Sat 2 Mar 14h", "Sun 3 Mar 11h")
    >>> DateParser("Wed 12, Thu 13, Fri 14 and Sat 15 Jun - 19h + Sun 16 Jun - 12h30").parse()
    ("Wed 12 Jun 19h", "Thu 13 Jun 19h", "Fri 14 Jun 19h", "Sat 15 Jun 19h", "Sun 16 Jun 12h30")
    """

    def __init__(self, line):
        self.date  = line
        self.dates = self.split_dates(line)
        self.final = []

        self.days = ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun']
        self.mons = ['jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug', 'sep', 'oct', 'nov', 'dec']

    def parse(self):
        if self.is_complex():
            self.parse_complex()
        else:
            self.parse_simple()

        return tuple(self.final)

    def parse_simple(self):
        """typical formats: 
        Day 00 + Day 01 Mon - 00h00
        Day 00 Mon - 00h00 + 01h00
        Day 00 Mon - 00h00 / Day 02 Mon - 00h00
        """

        for date in self.dates:
            mods = self.split_modifiers(date)

            date_mods = []
            for mod in mods:
                if self.is_complete(mod):
                    #only *one* base_date
                    base_date, time = self.split_time(mod)
                    date_mods.append(time)
                else:
                    date_mods.append(mod)

            for mod in date_mods:
                if self.is_hour(mod):
                    #Sat 1 Dec - 11h + 14h
                    self.final.append(' '.join([base_date, mod]))
                else:
                    #Fri 4 + Sat 5 Jan - 20h30
                    self.final.append(' '.join([mod, self.extract_month(base_date), time]))

    def parse_complex(self):
        """typical format:
        Day 00, Day 01 and Day 02 Mon - 00h00 + Day 03 Mon 01h00
        """
        raise NotImplementedError()


    def is_complex(self):
        """presence of the complex date attribute requires special parsing"""
        return self.date.find(' and ') > -1

    def is_complete(self, section):
        """section has format `Day 00 Mon - 00h00`
        must have no modifiers to determine completeness
        """
        sections = map(lambda x: x.lower(), section.split())

        try:
            dow, dom, moy, dash, time = sections
        except ValueError, e:
            return False

        return all([dow in self.days, moy in self.mons])


    def is_hour(self, section):
        return section[0].isdigit()

    def is_day(self, section):
        return section[:3].lower() in self.days


    def extract_month(self, section):
        """return the month present in the string, if any"""
        for mon in self.mons:
            if section.lower().find(mon) > -1:
                found = section.lower().index(mon)
                return section[found : found + 3]
        return None


    def split_dates(self, section):
        """split individual dates from a list of dates"""
        return section.split(' / ')

    def split_time(self, section):
        """split individual times from a complete date"""
        return section.split(' - ')

    def split_modifiers(self, section):
        """extend a date by implying that they share a date or a time
        modifiers change their meaning when parsing a complex date
        """
        return section.split(' + ')

>>> DateParser("Fri 4 + Sat 5 Jan - 20h30 / Sat 1 Dec - 11h + 14h + 16h / Sun 2 Dec - 12h30").parse()
('Fri 4 Jan 20h30', 'Sat 5 Jan 20h30', 'Sat 1 Dec 11h', 'Sat 1 Dec 14h', 'Sat 1 Dec 16h', 'Sun 2 Dec 12h30')

If you have questions about the way I’ve documented this class, feel free to get back to me, and I can help you out some more. This problem was a little more complex than I first thought, I need to get some other stuff done first.