Check if variable starts with 'http'

Question

I'm sure this is a simple solution, just haven't found exactly what I needed.

Using php, i have a variable $source. I wanna check if $source starts with 'http'.

if ($source starts with 'http') {
 $source = "<a href='$source'>$source</a>";
}

Thanks!

Solution 1

if (strpos($source, 'http') === 0) {
    $source = "<a href=\"$source\">$source</a>";
}

Note I use ===, not == because strpos returns boolean false if the string does not contain the match. Zero is falsey in PHP, so a strict equality check is necessary to remove ambiguity.

Reference:

http://php.net/strpos

http://php.net/operators.comparison

Solution 2

You want the substr() function.

if(substr($source, 0, 4) == "http") {
   $source = "<a href='$source'>$source</a>";
}

Solution 3

if(strpos($source, 'http') === 0)
    //Do stuff

Solution 4

Use substr:

if (substr($source, 0, 4) === 'http')

Solution 5

As of PHP 8.0 there is method str_starts_with implemented:

if (str_starts_with($source, 'http')) {
    $source = "<a href='$source'>$source</a>";
} 

View more solutions

Comments

[Jsowa]

PHP 8.0 solution - stackoverflow.com/a/64160271/7082164

[Ben]

You confused the haystack with the needle.

[Jonah]

Oops, needs to be 0, 4, not 4. 4 would get everything but the http :)

[Mikeon]

@Jonah: I was editing that while you were writing the comment. "Shit! I always do that. I always mess up some mundane detail." -- Michael Bolton

[Itay Moav -Malimovka]

You should test it, I believe it will fetch the entire string from position 4

[Robert]

Probably the most elegant way to check if a string starts with some substring in PHP.

[Mikeon]

@Itay: I already fixed it. I just typed it so fast that I didn't notice I missed the second parameter.

[Jonah]

@Agent: it's a rush here on StackOverflow. This place is very competitive; StackExchange did an amazing job of figuring out how to motivate people.

[Robert]

Lolwut? You really ask why and state that strpos (what I just commented to be the most elegant way) is more clear?

[Mikeon]

@Jonah: I know. Today has been a busy day for me. I got 'Mortarboard' today (yay!), then decided to see just how high I can go. :)

[Jonah]

@Robert: Oh, hehe, I mis-read. I thought you were saying that the substr solution was better. My bad :)

[Jonah]

@Agent: Cool, what is the daily cap? I might be almost there... 200 so far today.

[Robin]

Worked like a charm. Thanks so much for the quick reply :) And the extra links to read about strpos

[Mikeon]

@Jonah: You get the badge for hitting 200 rep, but you get rep from upvotes until you get 200 from upvotes only. Even after that, bonuses like accepted answers are exempt. Since SO goes based on UTC, "yesterday" ended 11 minutes ago. I got 265 rep for the day, just under double my previous record of 140.

[John Carter]

It is more visually clean than using substr(), but the inefficiency niggles at me - it'll be scanning the whole of $source in order to find the position of 'http'. Of course 99% of the time this won't matter, but still...

[Jonah]

@therefromhere: You have an intriguing point. Although, it seems like it would stop searching as soon as it has a match, wouldn't it? I don't know... we'd have to run a performance test.

[John Carter]

Also note that you've actually introduced a bug that wasn't in the question sample code - you've got ' instead of ". Fixing this myself.

[Mikeon]

@Jonah: I just looked at the reputation tab for your account. If I added it up right, you should have got 210 rep "yesterday". Expect the mortarboard badge sometime within the next few hours, and congratulations.

[John Carter]

@Jonah, yes it would stop as soon as it got a match, but for all the strings that don't start with 'http' it'd scan the whole string. Whereas if you used substr it'd only ever look at the letters we care about (ie the first 4).

[Mikeon]

@therefromhere: There's no bug in that code. The whole string is encapsulated within single quotes, so the double quotes inside of it are treated as normal characters (i.e. they don't close the string). Try it yourself if you don't believe me.

[Jonah]

Ahh, I get it now. I actually ran a test :) strpos took 0.0000003287 seconds, while substr took 0.0000003881. But that's with "http" at the beginning.

[Jonah]

Okay, now I'm really creeped out. Without "http" at the beginning, strpos took 0.0000003440 seconds, while substr took 0.0000004285 seconds. What's with this? Here's the test code if you want to try it yourself: pastebin.com/LdYBbaS3

[John Carter]

@Agent - yes, there is a bug. The single quotes mean $source won't be replaced with the variable value, so you'll get literally '<a href="$source">$source</a>' in the string.

[John Carter]

@Jonah props for actually testing! That is surprising. My guess is that strpos is faster since it returns an int whereas substr needs to do some memory allocation to return the string. If I could be bothered I'd run this on valgrind to check, but it's late :) It would be interesting to see if the same result occurs on a more varied data set.

[Jonah]

@therefromhere: Yes, it surprised me too. What you said sounds right. The comparison is being run in PHP with substr, but it's being run natively with strpos.

[squarecandy]

-1 - this would include strings with 'http' in the middle. The question asks how to check for strings that start with 'http'.

[Jonah]

@squarecandy I'm fairly sure it will only match the beginning. strpos returns the position of the needle ("http") within the haystack, and then the code above checks to make sure it's at position 0 (the beginning of the haystack). Perhaps there's a bug here that I've missed? Can you clarify?

[squarecandy]

ok, nevermind - i guess I misunderstood how you were using it. Does work with ===0

[badmadrad]

It would be helpful to explain your solution. Specifically the preg_match function and regex. This will help others understand how they could possibly use these tools for other problems.