Question

What would be the best approach to finding a date in a freeform text? A post where a user may place a date in it in several different ways such as:

1. July 14th & 15th
2. 7/14 & 7/15
3. 7-14 & 7-15
4. Saturday 14th and Sunday 15th
5. Saturday July 14th and 15th

and so on. Is regex my best choice for this type of thing with preg_match? I would also like to search if there are two dates, one for a start date and a second for an end date, but in the text I’m searching there may be one date or two.

This is my PHP code so far:

$dates1 = '01-01';
$dates2 = 'July 14th & 15th';
$dates3 = '7/14 & 7/15';
$dates4 = '7-14 & 7-15';
$dates5 = 'Saturday 14th and Sunday 15th';
$dates6 = 'Saturday July 14th and 15th';

$regexes = array(
        '/\s(1|2|3|4|5|6|7|8|9|10|11|12)\/\d{1,2}/',  //finds a date
        '/\s(1|2|3|4|5|6|7|8|9|10|11|12)-\d{1,2}/',  //finds another date
        '%\b(0?[1-9]|[12][0-9]|3[01])[- /.](0?[1-9]|1[012])\b%', //finds date format dd-mm or dd.mm
        );
foreach($regexes as $regex){
preg_match($regex,$dates,$matches);
}
var_dump($matches);

Answer 1

PHP has a class called DateTime that manages timestamps. It lets you convert between strings and DateTime objects pretty easily… Providing that your strings use the formats PHP gives you.

For example,

$date = DateTime::createFromFormat('d-m', '01-01');
$date = DateTime::createFromFormat('F d', 'July 14');
$date = DateTime::createFromFormat('d-M-Y', '15-Feb-2009');

That said, here’s what I’d do:

Create an array of acceptable formats, in order of priority:

$formats = array("d-m", "j-M-Y" ... );

Work with RegEx to massage your inputs so that they match your formats.

// Add the current year to this one:
$dates1 = '01-01';

// Split these into "July 14" and "July 15", and add the year
//  (this one will be the toughest)
$dates2 = 'July 14th & 15th';

// Split these into "7/14" and "7/15", and add the year
$dates3 = '7/14 & 7/15';

// Split these into "7-14" and "7-15", and add the year
$dates4 = '7-14 & 7-15';

// Split these, and add a month and year
$dates5 = 'Saturday 14th and Sunday 15th';

// Split these, and add a year:
$dates6 = 'Saturday July 14th and 15th';

Try to construct a DateTime object:

$date = false;
foreach ($formats as $format)
{
    $date = DateTime::createFromFormat($format, $dateString);
    if ($date) break;
}