simple Array question (just learning)

Maryanne

I'm trying to learn simple arrays but I'm thinking I used an extra step and made it too complicated. 😃

My goal: to regex a string of animal names and prices, and end up with a simple array of animal=>price.

Here's what I did:

$string = 'the "bear" is 4.00 the "goat" is 2.00 and the "cat" is 12.00';

I use preg_match to find the animal and price (the "animal" is always between quotes, and the price is always a normal dollar amount):

preg_match_all("/\"(.*?)\".*?(\d{1,2}\.\d{2})/",$string,$matches);

using "print_r" I see that $matches[1] holds the animals and $matches[2] holds the prices.

Array
(
    [0] => Array
        (
            [0] => "bear" is 4.00
            [1] => "goat" is 2.00
            [2] => "cat" is 12.00
        )

[1] => Array
    (
        [0] => bear
        [1] => goat
        [2] => cat
    )

[2] => Array
    (
        [0] => 4.00
        [1] => 2.00
        [2] => 12.00
    )

)

Now I create a new array to load the aforementioned "$matches array" into:

$animal_price = array();

for ($i=0;$i<count($matches);$i++){
$animal_price[$matches[1][$i]]=$matches[2][$i];
}

Now I'm done. When I loop through that final array, I get the simple [bear] => 4.00 and [goat] => 2.00

My question: did I do create that last array and then loop through the first array to fill it? Or is there something more efficient that I can do?

Thank you!!

p.s. sorry, I am just learning. I've spent about a week trying to figure out this on my own, I sure could use your input. 🙂

Weedpacket

Just just checking that no animal will cost more than 99.99 because [font=monospace]\d{1,2}[/font] won't catch the hundreds place.

If the text is always strictly [font=monospace]the "animal" is price[/font], then the preg_match_all will run faster if you explicitly mention the words (especially the [font=monospace]is[/font] part, because then it doesn't have to keep checking if it's found a digit yet or keep going):

/the "([^"]+)" is (\d{1,2}\.\d{2})/

I've also changed the first capturing group because (and I haven't bothered to test) I suspect it's easier to "match as many characters as possible before a double quote" than "match as few characters as possible, but match more if necessary until a double-quote is found". Certainly it used to be quicker, but I haven't kept up with the state of the art in PCRE internals.

The expression is short enough not to really bother, but there's the option of naming the capture groups and using those names to locate them in the [font=monospace]$matches[/font] array: [font=monospace]..."(?<animal>[^{"]+)"...[/font]} and [font=monospace]$matches['animal'][$i][/font].

And for the actual pairing up of animals and prices I'd use [man]array_combine[/man] instead of the loop:

$animal_price = array_combine($matches['animal'], $matches['price']);

Maryanne

Thank you! But you lost me at where you said:

"The expression is short enough not to really bother, but there's the option of naming the capture groups and using those names to locate them in the $matches array: ..."(?<animal>[^"]+)"... and $matches['animal'][$i]."

How do you name the capture groups?

Maryanne

Hi I figured it out :-)
I know how to name capture groups now :-)

I notice both ways work:
a.) my "loop"
b.) using named captures

Is the usage of "named captures" quicker, faster, more efficient? Or just easier to read?

My code will usually have no more than 10 array elements...

Weedpacket

Maryanne wrote:
Is the usage of "named captures" quicker, faster, more efficient? Or just easier to read?

It makes no difference to the actual operation of the regexp, but they can be easier to read and work with for a couple of reasons:

In such a short regexp as this, it doesn't really make much difference, but in more complex ones it can get a bit hard to tell which pairs of parentheses are supposed to be capturing which bits of text.
When working with the resulting array of matches, it can be easier to identify captures by name than have to count parentheses to figure out their numbers...
...Especially if you alter the pattern in the future to add or remove capturing parentheses, since that would mess up the numbering of the rest.