This is a universal, language-independent name parser.
Its purpose is to split a single string containing a full name, possibly including salutation, initials, suffixes etc., into meaningful parts like firstname, lastname, initials, and so on.
It is mostly tailored towards english names but works pretty well with non-english names as long as they use latin spelling.
E.g. Mr Anthony R Von Fange III is parsed to
This package has been used by The Iconic in production for years, successfully processing hundreds of thousands of customer names.
This parser is able to handle name patterns with and without comma:
... [firstname] ... [lastname] ...
... [lastname] ..., ... [firstname] ...
... [lastname] ..., ... [firstname] ..., [suffix]
More than 60 different successfully parsed name patterns can be found in the parser unit test.
composer require theiconic/name-parser
<?php $parser = new TheIconic\NameParser\Parser(); $name = $parser->parse($input); echo $name->getSalutation(); echo $name->getFirstname(); echo $name->getLastname(); echo $name->getMiddlename(); echo $name->getNickname(); echo $name->getInitials(); echo $name->getSuffix(); print_r($name->getAll()); echo $name;
An empty string is returned for missing parts.
You can retrieve last name prefixes and pure last names separately with
echo $name->getLastnamePrefix(); echo $name->getLastname(true); // true enables strict mode for pure lastnames, only
getNickname() returns the pure string of nick names. However, you can
true to have the same normalised parenthesis wrapping applied as in
echo $name->getNickname(); // The Giant echo $name->getNickname(true); // (The Giant)
You can re-print the parts that form a given name (that is first name, middle names and any initials)
in the order they were entered in while still applying normalisation
echo $name->getGivenName(); // J. Peter M.
You can re-print the full name, that is the given name as above followed by
any last name parts (excluding any salutations, nick names or suffixes)
echo $name->getFullName(); // J. Peter M. Schluter
$parser = new TheIconic\NameParser\Parser([ new TheIconic\NameParser\Language\English(), //default new TheIconic\NameParser\Language\German(), ])
$parser = new TheIconic\NameParser\Parser(); $parser->setNicknameDelimiters(['(' => ')']);
$parser = new TheIconic\NameParser\Parser(); $parser->setWhitespace("\t _.");
$parser = new TheIconic\NameParser\Parser(); $parser->setMaxSalutationIndex(2);
This will require salutations to appear within the first two words of the given input string. This defaults to half the amount of words in the input string, meaning that effectively the salutation may occur within the first half of the name parts.
$parser = new TheIconic\NameParser\Parser(); $parser->setMaxCombinedInitials(3);
Combined initials are combinations of several
uppercased letters, e.g.
separating spaces. The parser will treat such sequences
of uppercase letters (with optional dots) as combined
initials and parse them into individual initials.
This value adjusts the maximum number of uppercase letters
in a single name part are recognised as comnined initials.
Parts with more than the specified maximum amount of letters
will not be parsed into initials and hence will most likely
be parsed into first or middle names.
The default value is 2.
To disable combined initials support, set this value to 1;
If your input string consists of more than just the name and directly related bits like salutations, suffixes etc., any additional parts can easily confuse the parser. It is therefore recommended to pre-process any non-clean input to isolate the name before passing it to the parser.
We have not played with this, but you may be able to improve results by chaining several parses in sequence. E.g.
$parser = new Parser(); $name = $parser->parse($input); $name = $parser->parse((string) $name); ...
You can even compose your new input string from individual parts of a previous pass.
The parser is primarily built around the patterns of english names but tries to be compatible with names in other languages. Problems occur with different salutations, last name prefixes, suffixes etc. or in some cases even with the parsing order.
To solve problems with salutations, last name prefixes and suffixes you can create a separate language definition file and inject it when instantiating the parser, see 'Setting Languages' above and compare the existing language files as examples.
To deal with parsing order you may want to reformat the input string, e.g. by simply splitting it into words and reversing their order. You can even let the parser run over the original string and then over the reversed string and then pick the best results from either of the two resulting name objects. E.g. the salutation from the one and the lastname from the other.
The name parser has no in-built language detection. However, you may already ask the user for their nationality in the same form. If you do that you may want to narrow the language definition files passed into the parser to the given language and maybe a fallback like english. You can also use this information to prepare the input string as outlined above.
Alternatively, Patrick Schur as a PHP language detection library that seems to deliver astonishing results. It won't give you much luck if you run it over the the name input string only, but if you have any more text from the person in their actual language, you could use this to detect the language and then proceed as above.
Gender detection is outside the scope of this project. Detecting the gender from a name often requires large lists of first name to gender mappings.
However, you can use this parser to extract salutation, first name and nick names from the input string and then use these to implement gender detection using another package (e.g. this one) or service.
Writing different language files can not only be useful for parsing, but you can remap the normalised versions of salutations, prefixes and suffixes to transform them into something totally different.
E.g. you could map
princess of the kingdom of and then output
the parts in appropriate order to build a pipeline that automatically transforms
Ms. Louisa Lichtenstein into
Louisa, princess of the kingdom of Lichtenstein.
Of course, this is a silly and rather contrived example, but you get the
Of course this can also be used in more useful ways, e.g. to spell out
abbreviated titles, like
Professor etc. .
THE ICONIC Name Parser library for PHP is released under the MIT License.