[SOLVED] Problem running my code when the directory name has spaces.

[OM2]

I created a bit of code with help from answers given to other posts.

The code takes a PHP page and gives me the PHP output source code for the resulting HTML page.

 

The code is:

<?php
    $url = "http://" . $_SERVER[ 'HTTP_HOST' ] . urldecode( $_GET[ "u" ] );
    $lines = file( $url );
?>
<?php
$wholePage = "";

    foreach( $lines as $line_num => $line ) 
{
	$line = htmlspecialchars( $line );
	$line = str_replace( "<", '<span><', $line );
	$line = str_replace( ">", '></span>', $line );
	$line = str_replace( "<!--", '<em><!--', $line );
	$line = str_replace( "-->", '--></em>', $line );

	$wholePage = $wholePage . $line;
}

echo nl2br($wholePage);
?>

 

Unfortunately, I have a problem: if the directory that my PHP file sits in has a space.

 

The way the code is being called is:

 

http://www.domain.com/dirs/users/myName/source.php?u=/dirs/users/myName/index.php

 

What am I doing wrong?

 

Thanks.

 

 

OM

[trq]

You might want to use urlencode in the code that generates the links but really, directory names should not contain spaces on *nix.

[OM2]

u mean, instead of:

$url = "http://" . $_SERVER[ 'HTTP_HOST' ] . urldecode( $_GET[ "u" ] );

i should use:

$url = "http://" . $_SERVER[ 'HTTP_HOST' ] . urlencode( $_GET[ "u" ] );

?

 

that's funny because, i would have thought, urldecode and urlencode are opposites.

looked both up in php.net... aparently not (though it could be me misundersatnding and misreading).

 

let me know what u think.

thanks.

[trq]

No, thats not what I meant.

 

Have you got some code that generates these links?

 

http://www.domain.com/dirs/users/myName/source.php?u=/dirs/users/myName/index.php

 

Its there that you need to run through urlencode. If you are simply accessing these pages manually you will need to use %20 in place of spaces.

[OM2]

oh... i see what u mean.

yes, i have code like what u say.

i have:

<a href="source.php?u=<?= $_SERVER[ "PHP_SELF" ] ?>">View Source</a>

the only thing is, after failing to get it to work, i manually added the %20 in the url, but this didn't solve the problem.

still confused why it's not wokring.

[trq]

The entire problem is that *nix does not like directories of filenames to contain spaces. For instance, in the Linux shell if I wanted to cd into a directory called /this is foo I would need to use...

 

$ cd /this\ is\ foo

 

Why have your directories got spaces in there names in the first place?

[OM2]

just to make sure i wasnt missing anything, i did as u suggested.

 

the error i got was:

 

"The requested URL /dirs/users/name/my was not found on this server"

 

i've quoted the error exactly as i got it.

the actual directory name was 'my dir' - this isn't printed above, it seems to be missing all of the letters after the space.

i'm scratching my head over this... cant figure it out.

[OM2]

The entire problem is that *nix does not like directories of filenames to contain spaces. For instance, in the Linux shell if I wanted to cd into a directory called /this is foo I would need to use...

 

$ cd /this\ is\ foo

 

Why have your directories got spaces in there names in the first place?

well... i'm trying to develop a system where users can duplicate directories and give those directories ANY name they wish - with a space if necessary.

if it was for me only, then i'd be happy to just get passsed the problem and do without having spaces.

but the problem is users: i'd rather not stick in an artificial rule not allowing them to have spaces in the names.

 

let me know what you think.

thanks.

[trq]

But spaces shouldn't be in directory / file names. This is a irregularity inherited via a faulty design in windows.

 

Anyway, your going to need to explain exactly what it is your trying to do. The file() function asks for a filename, not a url. While this will work on some servers, by default most servers will have url_wrappers disabled and your code will fail.

 

What exactly is this codes intended purpose?

[OM2]

hmmm... those damn spaces!!

 

what i want to do: the php produces html that u see on your browser.

i want the source of the html code (NOT the source php code).

 

after getting help from a few questions i posted, i came up with the above script.

can u think of a better way to get the source code?

thanks.

[OM2]

just thought i'd let u know: i found a solution in the end.

it was much simpler than the code i had.

i've got the following code:

 

<?php 
$rawPageCode = htmlspecialchars(file_get_contents('http://' . $_SERVER['HTTP_HOST'] . $_SERVER['SCRIPT_NAME'])); 
$cleanPageCode = preg_replace("/<!--\s*Admin Code Start\s*-->.*?<!--\s*Admin Code End\s*-->/is", "", $rawPageCode);
echo $cleanPageCode;
?>

 

the above does exactly what i need it to do and spaces don't seem to be an issue anymore.

thanks for all the replies - much appreciated.

[OM2]

hmmm... i still had problems with spaces!

(and also the code itself wasn't quite right.)

i've fixed now.  now the following works:

 

<?php 
$newURL = str_replace(" ", "%20", 'http://' . $_SERVER['HTTP_HOST'] . dirname($_SERVER['PHP_SELF']));
$rawPageCode = htmlspecialchars(file_get_contents($newURL)); 
$cleanPageCode = preg_replace("/<!--\s*Admin Code Start\s*-->.*?<!--\s*Admin Code End\s*-->/is", "", $rawPageCode);
echo $cleanPageCode;
?>

(the above summes u r interested in the index page.  else, u can specifiy whatever page u want.)