PHP Forum

**Gilligan** · 05-27-2008, 05:33 PM

I know how to use nl2br, but how do I work backwards (change into new lines) ??

**JeremyMiller** · 06:13 PM

PHP Code:


			
$br2nl_string = str_replace('<br />',"\n",$original_string);

If you want a bit more flexibility on whether it's XHTML or HTML, this should work, but takes a bit more work for PHP

PHP Code:


			
$br2nl_string = preg_replace('/<br ?\/?>/',"\n",$original_string);

I didn't test that regex, but ? means 0 or 1, so it should work.

**Plugin-Developer** · 05-27-2008, 06:10 PM

If he wanted new lines in the new string, shouldn't he use the following?

PHP Code:


			
$br2nl_string = str_replace( '<br />', "\n", $original_string );

**JeremyMiller** · 05-27-2008, 06:13 PM

Quote:

Originally Posted by nickohrn

If he wanted new lines in the new string, shouldn't he use the following?

PHP Code:


			
$br2nl_string = str_replace( '<br />', "\n", $original_string );

Doh! Of course! lol. Thanks. TK coming your way.

**Gilligan** · 05-27-2008, 06:17 PM

its just displaying still

**JeremyMiller** · 05-27-2008, 06:22 PM

Can you give a code snippet? Since you made that remark, I created a test script:

PHP Code:


			
<?php
$original_string = 'ABC DEF<br />GHI JKL MNOP<br />';
$br2nl_string = str_replace( '<br />', "\n", $original_string );

print '<pre>'.$br2nl_string.'</pre>';

$br2nl_string = preg_replace('/<br ?\/?>/i',"\n",$original_string);
print '<pre>'.$br2nl_string.'</pre>';
?>

You'll notice I added an i in the preg_replace so that it can be or , etc.

**Gilligan** · 05-27-2008, 06:34 PM

So far there's nothing really in the script
so here's the whole thing

PHP Code:


			
<?php 



$index = fopen("index.html", "r+");

$index2 = fread($index, filesize("index.html"));

$heading = '<!-- main heading -->';
$headinge = "<!-- end main heading -->";
    list(, $data_split) = explode($heading, $index2, 2);
    list($ext2, ) = explode($headinge, $data_split, 2);
    
echo 'Heading: <input type="text" value="'.$ext2.'">';    
    
$para1 = '<!-- Start first -->';
$para1e = "<!-- End first -->";
    list(, $data_split) = explode($para1, $index2, 2);
    list($ext, ) = explode($para1e, $data_split, 2);
    
$extrep = preg_replace('/<br ?\/?>/',"\n", $ext);
echo '<textarea cols="60" rows="10">'.$extrep.'</textarea>';

fclose($index);



?>

So whats happening here is:

The script opens the file 'index.html', and the part wrapped in comments saying 'start first' and 'end first' are displayed in a textarea. Whats meant to happen is any tag inside the textarea should be replaced with new lines

**JeremyMiller** · 05-27-2008, 06:37 PM

I used

HTML Code:

<html>
  <head>
  </head>
  <body>
    <!-- main heading -->
    This is the main heading
    <!-- end main heading -->
    <!-- Start first -->
    ABC DEF GHI
    <br />
    JKL MNO
    <br />PQR STU VWX
    <br />YZ
    <!-- End first -->
</body>

For the index.html and got

Code:

    ABC DEF GHI
    

    JKL MNO
    
PQR STU VWX
    
YZ

as the result. Seems to be working (though, I'd add in that i I was talking about for greater flexibility).

**Gilligan** · 06:43 PM

And you used my exact code?? strange..

EDIT:

Grrr.. I kept refreshing and nothing changed, so then i done a complete refresh (CTRL + F5 for firefox) and it worked.

Thanks

**JeremyMiller** · 05-27-2008, 06:43 PM

Yep. Just copy-and-pasted. What is the HTML you're using (if gigantic, maybe a short snippet would be good).

**Gilligan** · 05-27-2008, 06:45 PM

Just realized the problem (read edit of last post to see)

**JeremyMiller** · 05-27-2008, 06:50 PM

Glad you caught it. Bit by the cache. We've all done that more than once! lol.

**JeremyMiller** · 06:57 PM

Just thought of a code optimization. I believe this will be faster, but you may want to do some benchmark testing:

PHP Code:


			
<?php
$index = fopen("index.html", "r+");

$index2 = fread($index, filesize("index.html"));

$heading = '<!-- main heading -->';
$headinge = "<!-- end main heading -->";
$start_position = strpos($index2,$heading);
$end_position = strpos($index2,$headinge,$start_position);

$ext2 = substr($index2,$start_position+21,$end_position - $start_position - 21);

echo 'Heading: <input type="text" value="'.$ext2.'">'.'<br />';

$para1 = '<!-- Start first -->';
$para1e = "<!-- End first -->";

$start_position = strpos($index2,$para1);
$end_position = strpos($index2,$para1e,$start_position);

$ext = substr($index2,$start_position+20,$end_position - $start_position - 20);

$extrep = preg_replace('/<br ?\/?>/i',"\n", $ext);
echo '<textarea cols="60" rows="10">'.$extrep.'</textarea>';

fclose($index);
?>

**JeremyMiller** · 05-27-2008, 07:06 PM

Ok. I must be having too much fun with this. If you change your tokens to require them of the form  and  where the ____ are equal, then this code may be helpful and more generalized:

PHP Code:


			
<?php
$index = fopen("index.html", "r+");

$index2 = fread($index, filesize("index.html"));

$successful = preg_match_all('/<!-- start ([a-z ]+?) -->(.+?)<!-- end (\1) -->/si',$index2,$matches);
//print '<pre>'.htmlentities(print_r($matches,true)).'</pre>';
if ((int)$successful > 0) {
  foreach ($matches[1] as $match_index=>$match_type) {
    switch ($match_type) {
      case 'main heading':
        echo 'Heading: <input type="text" value="'.trim($matches[2][$match_index]).'">'.'<br />';
        break;
      case 'first':
        $output_value = preg_replace('/<br ?\/?>/i',"\n", $matches[2][$match_index]);
        echo '<textarea cols="60" rows="10">'.trim($output_value).'</textarea>';
        break;
    }
  }
}
fclose($index);
?>

**Gilligan** · 05-27-2008, 07:15 PM

woah woah woah, thats way beyond me

Could you slow down and explain what you done a little more please?

**JeremyMiller** · 05-27-2008, 07:22 PM

lol. All I did was make a single regular expression to check the full document for all code blocks of the form described, and loop through them. The part your replace the _____'s with is what you should use in the case statement.

So, let's see some code instead of just verbiage:

HTML Code:

<html>
  <head>
  </head>
  <body>
    <!-- start main heading -->
    This is the main heading
    <!-- end main heading -->
    <!-- Start first -->
    ABC DEF GHI
    <br />
    JKL MNO
    <br />PQR STU VWX
    <br />YZ
    <!-- End first -->
</body>

is what I used to test the script above. You'll notice that there are 2 codeblocks and that the underscores in my format I gave you have been replaced with main heading and first, respectively.

Let's extend this a bit and say that you want to go ahead and add a second "first" block:

HTML Code:

<html>
  <head>
  </head>
  <body>
    <!-- start main heading -->
    This is the main heading
    <!-- end main heading -->
    <!-- Start first -->
    ABC DEF GHI
    <br />
    JKL MNO
    <br />PQR STU VWX
    <br />YZ
    <!-- End first -->
    <!-- Start first -->
    abc def ghi
    <br />
    jkl mno
    <br />pqr stu vwx
    <br />yz
    <!-- End first -->
</body>

That code will automatically process it. Now, let's say that you wanted to add a whole new type of section, say "navigation". You'd adjust the HTML like this:

HTML Code:

<html>
  <head>
  </head>
  <body>
    <!-- start main heading -->
    This is the main heading
    <!-- end main heading -->
    <!-- Start first -->
    ABC DEF GHI
    <br />
    JKL MNO
    <br />PQR STU VWX
    <br />YZ
    <!-- End first -->
    <!-- Start navigation -->
    <a href="index.php">Home</a>
    <a href="login.php">Login</a>
    <a href="faq.php">FAQ</a>
    <!-- End navigation -->
</body>

and then adjust the PHP to take care of another "case" like this (I added a couple of titles and line breaks for readability):

PHP Code:


			
<?php
$index = fopen("index.html", "r+");

$index2 = fread($index, filesize("index.html"));

$successful = preg_match_all('/<!-- start ([a-z ]+?) -->(.+?)<!-- end (\1) -->/si',$index2,$matches);
//print '<pre>'.htmlentities(print_r($matches,true)).'</pre>';
if ((int)$successful > 0) {
  foreach ($matches[1] as $match_index=>$match_type) {
    switch ($match_type) {
      case 'main heading':
        echo 'Heading: <input type="text" value="'.trim($matches[2][$match_index]).'">'.'<br />';
        break;
      case 'first':
        $output_value = preg_replace('/<br ?\/?>/i',"\n", $matches[2][$match_index]);
        echo 'Body: <br /><textarea cols="60" rows="10">'.trim($output_value).'</textarea><br />';
        break;
      case 'navigation':
        echo 'Navigation: <br /><textarea cols="60" rows="10">'.trim($matches[2][$match_index]).'</textarea><br />';
        break;
    }
  }
}
fclose($index);
?>

You can extend it as much as you please and handle as many tags as you want.