Extracting data tools - coordinates

I will use regular expressions to extract data from raw text and named tuples for convenient way to access tuple items, so first I import  required modules:

import re
from collections import namedtuple

I am going to build desired regular expression to extract coordinates from text based on coordinates order (lat-lon or lon-lat), coordinates formats and separator character that separates latitude and longitude within coordinate pair not one coordinate pair from another one.

# Coordinates order
ORDER_LATLON = 'LATLON'
ORDER_LONLAT = 'LONLAT'

# Coordinate format constants
DMSH_COMP = 'DMSH_COMP'
HDMS_COMP = 'HDMS_COMP'
DMSH_SEP = 'DMSH_SEP'
HDMS_SEP = 'HDMS_SEP'

# Separators constants - separates latitude and longitude in pair, not pairs!
SEP_NULL = r''
SEP_SPACE = r' '
SEP_HYPHEN = r'-'
SEP_SLASH = r'/'
SEP_BACKSLASH = r'\\'

Now, it's time to define named tuple - coordinate pair. It will be convenient to used this tuple in dictionary with patterns for regex:

coord_pair = namedtuple('coord_pair', 'lat lon')

And dictionary with regex patterns:

coord_formats = {DMSH_COMP: coord_pair(r'\d{6}\.\d+[NS]|\d{6}[NS]', r'\d{7}\.\d+[EW]|\d{7}[EW]'),
                 HDMS_COMP: coord_pair(r'[NS]\d{6}\.\d+|[NS]\d{6}', r'[EW]\d{7}\.\d+|[EW]\d{7}'),
                 DMSH_SEP: coord_pair(r'''\d{1,2}\W\d{2}\W\d{1,2}\.\d+\W{1,2}[NS]|\d{1,2}\W\d{1,2}\W\d{1,2}\W{1,2}[NS]''',
                                      r'''\d{1,3}\W\d{1,2}\W\d{1,2}\.\d+\W{1,2}[EW]|\d{1,3}\W\d{1,2}\W\d{1,2}\W{1,2}[EW]'''),
                 HDMS_SEP: coord_pair(r'''[NS]\d{1,2}\W\d{2}\W\d{1,2}\.\d+\W{1,2}|[NS]\d{1,2}\W\d{1,2}\W\d{1,2}\W{1,2}''',
                                      r'''[EW]\d{1,3}\W\d{1,2}\W\d{1,2}\.\d+\W{1,2}|[EW]\d{1,3}\W\d{1,2}\W\d{1,2}\{1,2}W''')}

Notice that for DMSH_SEP format I used \W special character which matches any character which is not a word character. I make an assumption here that DMS format is separated by degrees, minutes and seconds symbols such as in examples:
85°52'55"W, 33°45'14"N, 48°54'04.05"N, 002°21'10"E
It is simple way to avoid using of special characters for degrees symbol for example and issue with different encodings. (utf-8, ASCII etc.). Because it is not a 'strict' way to check if given text is in DMS separated format, it may lead to unexpected result.

class CoordRegexBuilder:
    def __init__(self, coord_order, coord_format, coord_sep):
        self.coord_order = coord_order
        self.coord_format = coord_format
        self.coord_sep = coord_sep
        self.coord_regex_str = self.create_regex_str()

Methods that create regular expression string(pattern) and compile it into regex object:

def create_regex_str(self):
    regex_str = ''
    lat_format = coord_formats.get(self.coord_format).lat
    sep_format = self.coord_sep
    lon_format = coord_formats.get(self.coord_format).lon

    if self.coord_order == ORDER_LATLON:
        regex_str = r'(?P<lat>' + lat_format + ')' + \
                    '(?P<sep>' + sep_format + ')' + \
                    '(?P<lon>' + lon_format + ')'

    elif self.coord_order == ORDER_LONLAT:
        regex_str = r'(?P<lon>' + lon_format + ')' + \
                    '(?P<sep>' + sep_format + ')' + \
                    '(?P<lat>' + lat_format + ')'

    return regex_str

def get_coord_regex(self):
    return re.compile(self.coord_regex_str)