Skip to content
Permalink
master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

address-splitter/src/AddressSplitter.php /
Jump to
Code definitions
Code navigation index up-to-date

Go to file
  • Go to file
  • Copy path
  • Copy permalink
@cjuner
cjuner Fix parsing streets starting the same as number prefixes (#29) (#29)
Latest commit 5aa9c01 Oct 24, 2019 History 
Resolves: <#30>
5 contributors

Users who have contributed to this file

@andrewisplinghoff @cjuner @windaishi @emilv @bcremer
355 lines (328 sloc) 15 KB
Raw Blame
Open in GitHub Desktop
<?php
namespace VIISON\AddressSplitter;
use VIISON\AddressSplitter\Exceptions\SplittingException;
/**
* @copyright Copyright (c) 2017 VIISON GmbH
*/
class AddressSplitter
{
private static $generalNumberPrefixes = array(
// English
'No [.:]?',
'Nos ([.:]|\s)',
'Number \s',
// German
'Nr [.:]?',
'Nummer \s',
// Other
'№ [.:]?',
'Nº [.:]?',
'n° [.:]?'
);
private static $additionalHouseNumberPrefixes = array(
// German
'Hausnummer \s',
'Hausnr [.:]?'
);
private static function getHouseNumberPrefixes()
{
return self::getRegexForNumberPrefixes(self::$generalNumberPrefixes + self::$additionalHouseNumberPrefixes);
}
private static function getGeneralNumberPrefixes()
{
return self::getRegexForNumberPrefixes(self::$generalNumberPrefixes);
}
/**
* Creates an anonymous group matching the supplied alternative regular expressions for number prefixes.
*
* The generated regular expression
*
* * assumes 'x' and 'u' pattern modifiers,
*
* * assumes each alternative must either be at the start of the string or be preceded by whitespace,
*
* * assumes each alternative to be case insensitive,
*
* * checks that the prefix will be succeeded by a number (following optional spaces).
*
* @param array $numberPrefixes a list of regular expressions (assuming 'x' and 'u' flags)
* @return string a regular expression for an anonymous group matching the alternatives supplied in the argument
*/
private static function getRegexForNumberPrefixes(array $numberPrefixes)
{
return self::concatenateRegexAlternatives(array_map(
function ($numberPrefix) {
return '(?: ^ | (?<=\s)) (?i: ' . $numberPrefix . ' (?= \s* \pN+) )';
},
$numberPrefixes
));
}
private static function concatenateRegexAlternatives(array $alternatives)
{
return '(?:' . implode('|', $alternatives) . ')';
}
/**
* This function splits an address line like for example "Pallaswiesenstr. 45 App 231" into its individual parts.
* Supported parts are additionToAddress1, streetName, houseNumber and additionToAddress2. AdditionToAddress1
* and additionToAddress2 contain additional information that is given at the start and the end of the string, respectively.
* Unit tests for testing the regular expression that this function uses exist over at https://regex101.com/r/vO5fY7/1.
* More information on this functionality can be found at http://blog.viison.com/post/115849166487/shopware-5-from-a-technical-point-of-view#address-splitting.
*
* @param string $address
* @return array
* @throws SplittingException
*/
public static function splitAddress($address)
{
$houseNumberPrefixes = self::getHouseNumberPrefixes();
$numberPrefixes = self::getGeneralNumberPrefixes();
/* Matching this group signifies the following text is part of
* additionToAddress2.
*
* See [1] for some of the English language stop words and abbreviations.
*
* [1] <https://web.archive.org/web/20180410130330/http://maf.directory/zp4/abbrev.html>
*/
$addition2Introducers = '(?:
# {{{ Additions relating to who (a natural person) is addressed
\s+ [Cc] \s* \/ \s* [Oo] \s
| ℅
| \s+ care \s+ of \s+
# German, Swiss, Austrian
| \s+ (?: p|p.\s*|per\s+ ) (?: A|A.|Adr.|(?<=\s)Adresse ) \s
| \s+ p. \s* A. \s
| \s+ (?: z | z.\s* | zu\s+ ) (?: Hd|Hd.|(?<=\s)Händen|(?<=\s)Haenden|(?<=\s)Handen) \s+
## o. V. i. A. = oder Vertreter im Amt
| \s+ (?: o | o.\s* | oder\s+ )
(?: V | V.\s* | (?<=\s)Vertreter\s+ )
(?: i | i.\s* | (?<=\s)im\s+ )
(?: A | A.\s* | (?<=\s)Amt\s+ )
# }}}
# {{{ Additions which further specify more precisely the location
| \s+ (?: Haus ' . $numberPrefixes . '? ) \s
| \s+ (?: WG | W\.G\. | WG\. | Wohngemeinschaft ) ' . $numberPrefixes .'? ($ | \s)
| \s+ (?: [Aa]partment | APT \.? | Apt \.? ) ' . $numberPrefixes .'? \s
| \s+ (?: [Ff]lat ) ' . $numberPrefixes .'? \s
| (?: # Numeric-based location specifiers (e.g., "3. Stock"):
\s+
(?:
[\p{N}]+ # A number, …
(?i: st | nd | rd | th)? # …, optionally followed by an English number suffix
\.? # …, followed by an optional dot,
\s* # …, followed by optional spacing
)?
(?: # Specifying category:
(?i: Stock | Stockwerk)
| App \.? | Apt \.? | (?i: Appartment | Apartment)
)
# At the end of the string or followed by a space
(?: $ | \s)
)
| (?:
\s+ (?:
# English language stop words wrt location from source [1]
# (extracted only those which may not be _exclusively_ part of
# street names):
| ANX \.? | (?i: ANNEX)
| APT \.? | (?i: APARTMENT)
| ARC \.? | (?i: ARCADE)
| AVE \.? | (?i: AVENUE)
| BSMT \.? | (?i: BASEMENT)
| BLDG \.? | (?i: BUILDING)
| CP \.? | (?i: CAMP)
| COR \.? | (?i: CORNER)
| CORS \.? | (?i: CORNERS)
| CT \.? | (?i: COURT)
| CTS \.? | (?i: COURTS)
| DEPT \.? | (?i: DEPARTMENT)
| DV \.? | (?i: DIVIDE)
| EST \.? | (?i: ESTATE)
| EXT \.? | (?i: EXTENSION)
| FRY \.? | (?i: FERRY)
| FLD \.? | (?i: FIELD)
| FLDS \.? | (?i: FIELDS)
| FLT \.? | (?i: FLAT)
| FL \.? | (?i: FLOOR)
| FRNT \.? | (?i: FRONT)
| GDNS \.? | (?i: GARDEN)
| GDNS \.? | (?i: GARDENS)
| GTWY \.? | (?i: GATEWAY)
| GRN \.? | (?i: GREEN)
| GRV \.? | (?i: GROVE)
| HNGR \.? | (?i: HANGER)
| HBR \.? | (?i: HARBOR)
| HVN \.? | (?i: HAVEN)
| KY \.? | (?i: KEY)
| LBBY \.? | (?i: LOBBY)
| LCKS \.? | (?i: LOCK)
| LCKS \.? | (?i: LOCKS)
| LDG \.? | (?i: LODGE)
| MNR \.? | (?i: MANOR)
| OFC \.? | (?i: OFFICE)
| PKWY \.? | (?i: PARKWAY)
| PH \.? | (?i: PENTHOUSE)
| PRT \.? | (?i: PORT)
| RADL \.? | (?i: RADIAL)
| RM \.? | (?i: ROOM)
| SPC \.? | (?i: SPACE)
| SQ \.? | (?i: SQUARE)
| STA \.? | (?i: STATION)
| STE \.? | (?i: SUITE)
| TER \.? | (?i: TERRACE)
| TRAK \.? | (?i: TRACK)
| TRL \.? | (?i: TRAIL)
| TRLR \.? | (?i: TRAILER)
| TUNL \.? | (?i: TUNNEL)
| VW \.? | (?i: VIEW)
| VIS \.? | (?i: VISTA)
# Custom custom additions:
| (?i: Story | Storey)
| LVL \.? | (?i: Level)
)
' . $numberPrefixes . '?
# May optionally be followed directly by a number+letter
# combination (e.g., "LVL3C"):
(?: [\p{N}]+[\p{L}]* )?
# Occurs at the end of the string or followed by a space:
($ | \s)
)
# Heuristic to match location specifiers. These must not be
# conflated with house number extensions as in "12 AB". Hence
# our heuristic is at least 3 letters with the first letter being
# spelled as a capital. E.g., it would match "Haus", "Gebäude" or
# "Arbeitspl.", but not "AAB".
| \s+ ( [\p{Lu}\p{Lt}] [\p{Ll}\p{Lo}]{2,} \.? ) ($ | \s)
# }}}
)';
$regex = '
/\A\s*
(?: #########################################################################
# Option A: [<Addition to address 1>] <House number> <Street name> #
# [<Addition to address 2>] #
#########################################################################
(?:(?P<A_Addition_to_address_1>.*?),\s*)? # Addition to address 1
(?: ' . $houseNumberPrefixes . ' \s*)?
(?P<A_House_number_match>
(?P<A_House_number_base>
\pN+(\s+\d+\/\d+)?
)
(?:
\s*[\-\/\.]?\s*
(?P<A_House_number_extension>(?:[a-zA-Z\pN]){1,2})
\s+
)?
)
\s*,?\s*
(?P<A_Street_name>(?:[a-zA-Z]\s*|\pN\pL{2,}\s\pL)\S[^,#]*?(?<!\s)) # Street name
\s*(?:(?:[,\/]|(?=\#))\s*(?!\s* ' . $houseNumberPrefixes . ')
(?P<A_Addition_to_address_2>(?!\s).*?))? # Addition to address 2
| #########################################################################
# Option B: [<Addition to address 1>] <Street name> <House number> #
# [<Addition to address 2>] #
#########################################################################
(?:(?P<B_Addition_to_address_1>.*?),\s*(?=.*[,\/]))? # Addition to address 1
(?!\s* ' . $houseNumberPrefixes . ')
(?P<B_Street_name>[^0-9# ]\s*\S(?:[^,#](?!\b\pN+\s))*?(?<!\s)) # Street name
\s*[\/,]?\s*(?:\s ' . $houseNumberPrefixes . ')?\s*
(?P<B_House_number_match>
(?P<B_House_number_base>
\pN+
)
(?:
# Match house numbers that are (optionally) amended
# by a dash (e.g., 12-13) or slash (e.g., 12\/A):
(?: \s*[\-\/]\s* )*
(?P<B_House_number_extension>
(?:
# Do not match "care-of"-like additions as
# house numbers:
(?!' . $addition2Introducers .')
\s*[\pL\pN]+
)
# Match any further slash- or dash-based house
# number extensions:
(?:
# Do not match "care-of"-like additions as
# house numbers:
(?!' . $addition2Introducers .')
# Match any (optionally space-separated)
# additionals parts of house numbers after
# slashes or dashes.
\s* [\-\/] \s*
[\pL\pN]+
)*
)
)?
) # House number
(?:
(?:\s*[-,\/]|(?=\#)|\s)
\s*
(?!\s* ' . $houseNumberPrefixes . ' )
\s*
(?P<B_Addition_to_address_2>(?!\s).*?)
)?
)
\s*\Z/xu';
$result = preg_match($regex, $address, $matches);
if ($result === 0) {
throw new SplittingException(SplittingException::CODE_ADDRESS_SPLITTING_ERROR, $address);
} elseif ($result === false) {
throw new \RuntimeException(sprintf('Error occurred while trying to split address \'%s\'', $address));
}
if (!empty($matches['A_Street_name'])) {
return array(
'additionToAddress1' => $matches['A_Addition_to_address_1'],
'streetName' => $matches['A_Street_name'],
'houseNumber' => $matches['A_House_number_match'],
'houseNumberParts' => array(
'base' => $matches['A_House_number_base'],
'extension' => isset($matches['A_House_number_extension']) ? trim($matches['A_House_number_extension']) : ''
),
'additionToAddress2' => (isset($matches['A_Addition_to_address_2'])) ? $matches['A_Addition_to_address_2'] : ''
);
} else {
return array(
'additionToAddress1' => $matches['B_Addition_to_address_1'],
'streetName' => $matches['B_Street_name'],
'houseNumber' => $matches['B_House_number_match'],
'houseNumberParts' => array(
'base' => $matches['B_House_number_base'],
'extension' => isset($matches['B_House_number_extension']) ? trim($matches['B_House_number_extension']) : ''
),
'additionToAddress2' => isset($matches['B_Addition_to_address_2']) ? $matches['B_Addition_to_address_2'] : ''
);
}
}
/**
* @param string $houseNumber A house number string to split in base and extension
* @return array
* @throws SplittingException
*/
public static function splitHouseNumber($houseNumber)
{
$regex =
'/
\A\s* # Trim white spaces at the beginning
(?: ' . self::getHouseNumberPrefixes() . '?\s*)?
(?:\#\s*)? # Trim #
(?<House_number_base>
[\pN]+ # House Number base (only the number)
)
\s*[\/\-\.]?\s* # Separator (optional)
(?<House_number_extension> # House number extension (optional)
.*? # Here we allow every character. Everything after the separator is interpreted as extension
)
\s*\Z # Trim white spaces at the end
/xu'; // Option (u)nicode and e(x)tended syntax
$result = preg_match($regex, $houseNumber, $matches);
if ($result === 0) {
throw new SplittingException(SplittingException::CODE_HOUSE_NUMBER_SPLITTING_ERROR, $houseNumber);
} elseif ($result === false) {
throw new \RuntimeException(sprintf('Error occurred while trying to house number \'%s\'', $houseNumber));
}
return array(
'base' => $matches['House_number_base'],
'extension' => $matches['House_number_extension']
);
}
}