Detecting /n, & presenting links

mghfaust

Hi,
I want to be able to allow administrators for a CMS to enter a URL and have it appear as a link. But, of course, I don't want to allow any old html.

I wrote a short function that can do this, but I have 2 concerns:

Is there a better way than what I've done (see code below)? It seems pretty dang clunky.
How can I detect text line breaks?

The reason for 2. is that my function looks for the first space after a substring beginning with "http://" to determine where the URL ends. However, if someone entered text with a line break right after the URL, my script thinks the next word is part of the URL

function text_tags($text)
{
# THIS FUNCTION TAKES TEXT, PREPARES IT FOR WEB PRESENTATION USING htmlspecialchars AND MAKING 
# ANYTHING BEGINNING WITH 'http://' A LINK
  $new_text = $text;
  $text_length = strlen($text); 
  $end_text = $text;
  $a = strpos($end_text, "http://");
  #LOOP AS MANY TIMES AS THERE ARE URLs IN THE TEXT:
  for ($i = 1; $i <= substr_count($text, "http://"); $i++)
  {
    #FIND THE POSITION OF THE NEXT LINK AND CUT EVERYTHING BEFORE THAT FROM $end_text:
  	$pos = strpos($end_text, "http://");
    $end_text = substr($end_text, $pos);
	#FIND THE END OF THE URL BY LOOKING FOR A SPACE 
	# -- IF THERE IS NONE, THE LINK GOES TO THE END OF THE STRING:
	$sp_pos = strpos($end_text, " ");
	if ($sp_pos < 1)
	  $sp_pos = strlen($end_text);
	#ISOLATE THE URL:
	$link = substr($end_text, 0, $sp_pos);
	#REMOVE THE LINK FROM $end_text (WHERE WHITTLING AWAY AT IT):
	$end_text = substr($end_text, $sp_pos +1);
	#CREATE A REPLACEMENT STRING FOR THE LINK, WHICH WILL GET BY HTMLSPECIALCHARS:
	$new_link = str_replace("http://", "{A HREF}", $link);
	$new_link = $new_link."{A MID}".$link."{A OFF}";
	#NOW USE THAT TO REPLACE THE LINK IN A COPY OF THE STRING:
    $new_text = str_replace($link, $new_link, $new_text);
  }
  $new_text = htmlspecialchars($new_text);
  #REPLACE THE DUMMY LINK PARTS WITH REAL html:
  $new_text = str_replace("{A HREF}", "<a href='http://", $new_text);
  $new_text = str_replace("{A MID}", "'>", $new_text);
  $new_text = str_replace("{A OFF}", "</a>", $new_text);
  return ($new_text);
}

benbox

for the url u cud just do what a lot of forums do, like this 1 for e.g and have them writing blahblah into the text area and have a 'replace' of some sort swapping for <a href=$link> and to </a>

benbox

cept these onces took me too seriously 😉
[ url ]
[ /url ]
(but with no spaces 😉)

samscripts

Hi, you could probably just use a regular expression to convert the http://whateverthelinkis.com to a <a href="blah"> link.

Something along the lines of:

<?php

$newtext = preg_replace("|http://([^\s ])|is", '<a href="http://$1">$1</a>', $text);

Should work (I haven't tested it though).

At any rate, I recommend learning about regular expressions. They can save you ages when it comes to things like this.

benbox wrote
for the url u cud just do what a lot of forums do, like this 1 for e.g and have them writing blahblah into the text area and have a 'replace' of some sort swapping for <a href=$link> and to </a>

I have a simple php class I wrote that does something similar to the way these forums work (with the tags). see http://www.samscripts.com/scripts/bbtags

Theres also one on phpclasses.org here which looks quite good, although I haven't tried it myself.

Hope this helps, Sam

mghfaust

I see
Very useful.

Still, I need to find an easier way to find the link within a long field of submitted text. Any ideas?

-mike

samscripts

Ok, I just checked the preg_replace I gave earlier, and I realise it doesn't work properly.

This function takes the text, runs htmlspecialchars over it (so they can't enter any htmltags), then it finds all http://whatever.com and converts them to <a href="http://whatever.com">http://whatever.com</a>, and then, just for good measure, returns the text with newlines converted to <br> ready for output.

If this example doesn't do what you want, then I've misunderstood what you are asking for.

<?php

$text = 'This is a load of text. In it, there are some links in the form http://www.phpbuilder.com
and http://www.php.net . Theres also some nasty html tags like <b>these<i>ones</i></b>.

That was 2 newlines just then. I think this function should also handle cases when someone
writes a url with a full stop or other punctuation mark after it (ie not just when it ends with a space),
like http://www.apache.org.';

function text_tags($text){
	$notagstext = htmlspecialchars($text);
	$convertedlinkstext = preg_replace("|http://([^\s ]*[a-z])|is", '<a href="http://$1">$1</a>', $notagstext);
	$linebreakedtext = nl2br($convertedlinkstext);
	return $linebreakedtext;
}

echo text_tags($text);

?>

What the preg_replace function does, is looks for a string beginning with http://, followed by any number of characters that are not white space (newlines, spaces, tabs). It keeps on grabbing the characters until it gets to something that is not a letter. Then it replaces all of this with the html <a> tag stuff neccessary.

If you just want to extract all the links from a bit of text, try the [man]preg_match_all/man function.

Hope this helps,

Sam

mghfaust

This function takes the text, runs htmlspecialchars over it (so they can't enter any htmltags), then it finds all http://whatever.com and converts them to <a href="http://whatever.com">http://whatever.com</a>, and then, just for good measure, returns the text with newlines converted to <br> ready for output.

If this example doesn't do what you want, then I've misunderstood what you are asking for.

That's exactly what I wanted it to do. Thank you very much!

One thing though:
Notice the text you used as $text. In it, the first link, [url]http://www.phpbuilder.com,[/url] is followed by a text-delimited line break. Your (much more efficient) function still sees the "and" that immediately follows as part of the link, and prints it into the link ulr and name, rendering it broken.

Any clues how to detect the line break so the function can recognize it as the end of the link?

This has been very helpful.

Thanks,
-mike

samscripts

Hi, the problem is that this forum is stripping out a backslash that should be appearing in the code.

The first argument to preg_replace should have a backslash \ before the 's':

|http://([^\s ]*[a-z])|is

The \s basically means any newline, etc character.

It was there when I posted it the first time, vbulletin must have eaten it. :mad:

mghfaust

Thanks!!!
It's now perfect.

-mike

Originally posted by samscripts
Hi, the problem is that this forum is stripping out a backslash that should be appearing in the code.

The first argument to preg_replace should have a backslash \ before the 's':

|[url]http://[/url]([^\s ]*[a-z])|is

The \s basically means any newline, etc character.

It was there when I posted it the first time, vbulletin must have eaten it. :mad: