URL Strings Being Forced to Lower Case

registered · Posted: Sat Oct 15, 2005 4:06 pm

Hi again, unfortunately, I am struggling with another problem. I am trying to Google-Tap the NSN GR Downloads module -- with quite a bit difficulty given that it is very GT unfriendly in a few areas.

What I seem to be finding is that although I can see links like downloads-NewDownloads.html in the brower and in the status bar, but Apache seems to be lower casing everything so that it comes to the re-write engine all lower case as downloads-newdownloads.html. I am able to get around this issue by making many more specific .htaccess directives, but unfortunately, the way paging is done and the search function, it comes out all wrong.

With the paging, I just gave up as I didn't want to modify code. In terms of Search, though, its a problem because the search is broken by the last GT replace of downloads.html. Bob has buried the link reference in an OPTION element as a POST, which when GT'd, it comes out with an empty value in the $query parameter, basically making the search non-functional.

GR Downloads is not googletap friendly. (Bob: I would suggest making this more friendly in future revisions as it is really a nice module otherwise.)

Here is the code that is writing the search form:

Code:




function SearchForm() {


    global $module_name, $query;


    echo "<table border='0' cellspacing='0' cellpadding='0' align='center'>\n";


    echo "<form action='modules.php?name=$module_name&amp;op=search&amp;query=$query' method='post'>\n";


    echo "<tr><td><font class='content'><input type='text' size='25' name='query' value='$query'> <input type='submit' value='"._DL_SEARCH."'></td></tr>\n";


    echo "</form>\n";


    echo "</table>\n";


}

For those of you who are familiar with GT, basically, this will come out looking like:

download-search-.html

So, the query string is lost. I have been unable to figure out a way to fix this. What I was thinking of doing was changing the GT line:

Code:




"'(?<!/)modules.php\?name=Downloads'"

to something that would STOP after the last "s" so that if there was another & it would not change it. Then, I would just live with the search not being GT'd by removing all the GT statements for changing op=search*.

Any suggestions? I can't seem to find anything that will do it because it is most likely not at the end of a line.

TIA,
montego

Site Admin/Owner Joined: Aug 27, 2002 Posts: 17088

Apache doesn't modify that. Are you sure that in header.php the outurl isn't lower casing it?

Posted: Sat Oct 15, 2005 5:18 pm

Actually, Raven, I am not certain, but I didn't modify the GT-NextGEn code. Anyways, just in case, here is the GT-Downloads that I have for NSN GR Downloads (note that it has the offending statements commented out right now since this is now "live"):

Code:




$urlin = array(


"'(?<!/)modules.php\?name=Downloads&amp;op=modifydownloadrequest&amp;lid=([0-9]*)'",


"'(?<!/)modules.php\?name=Downloads&amp;op=NewDownloads&amp;newdownloadshowdays=([0-9]*)'",


"'(?<!/)modules.php\?name=Downloads&amp;op=NewDownloadsDate&amp;selectdate=([a-zA-Z0-9+]*)'",


"'(?<!/)modules.php\?name=Downloads&amp;op=getit&amp;lid=([0-9]*)'",


"'(?<!/)modules.php\?name=Downloads&amp;op=(MostPopular)&amp;ratenum=([0-9]*)&amp;ratetype=(num|percent)'",


"'(?<!/)modules.php\?name=Downloads&amp;op=(NewDownloads|MostPopular)'",


"'(?<!/)modules.php\?name=Downloads&amp;cid=([0-9]*)&amp;orderby=([a-zA-Z0-9+]*)'",


//"'(?<!/)modules.php\?name=Downloads&amp;min=([0-9]*)&amp;cid=([0-9]*))'",   //GR Downloads not GT-Friendly!!


//"'(?<!/)modules.php\?name=Downloads&amp;op=search&amp;query=([/:\-\'{}()\,\._&amp;a-zA-Z0-9+=]*)&amp;min=([0-9]*)&amp;orderby=([a-zA-Z0-9+]*)&amp;show=([0-9]*)'",   //GR Downloads not GT-Friendly!!


//"'(?<!/)modules.php\?name=Downloads&amp;op=search&amp;query=([/:\-\'{}()\,\._&amp;a-zA-Z0-9+= ]*)&amp;orderby=([a-zA-Z0-9+]*)'",   //GR Downloads not GT-Friendly!!


//"'(?<!/)modules.php\?name=Downloads&amp;op=search&amp;query=([/:\-\'{}\(\)\,\._&amp;a-zA-Z0-9+=]*)'",   //GR Downloads not GT-Friendly!!


"'(?<!/)modules.php\?name=Downloads&amp;op=gfx&amp;random_num=([0-9]*)'",


"'(?<!/)modules.php\?name=Downloads&amp;cid=([0-9]*)'",


"'(?<!/)modules.php\?name=Downloads'"


);





$urlout = array(


"download-mod-\\1.html",


"download-shownew-\\1.html",


"download-seldate-\\1.html",


"download-file-\\1.html",


"download-\\1-\\2-\\3.html",


"downloads-\\1.html",


"download-sort-\\1-orderby-\\2.html",


//"download-paging-\\1-\\2.html",   //GR Downloads not GT-Friendly!!


//"download-search-\\1-\\2-\\3-\\4.html",   //GR Downloads not GT-Friendly!!


//"download-search-\\1-\\2.html",   //GR Downloads not GT-Friendly!!


//"download-search-\\1.html",   //GR Downloads not GT-Friendly!!


"download-gfx-\\1.html",


"downloads-cat\\1.html",


"downloads.html"


);

The header.php code is:

Code:




#Start of Google Tap Header 


global $nextgen_name, $prefix, $db; 





// Google Tap On or Off, 1=On, 0=Off 


$next_gen_ob = 1; 





//Used for main module or a module without a name 


if ($nextgen_name == "") { 


$sql = "SELECT main_module FROM ".$prefix."_main"; 


$result = $db->sql_query($sql); 


$row = $db->sql_fetchrow($result); 


$mainmod_name = $row[main_module]; 


$nextgen_name = $mainmod_name; 


} 





//Make the path to the GT-NExtGEn Files 


$nextgen_path = "GoogleTap/GT-".$nextgen_name.".php"; 


if ($next_gen_ob == 1) { 


   if (file_exists($nextgen_path)) { 


   ob_start(); 


   } else { 


   $next_gen_ob = 0; 


   } 


} 





function replace_for_mod_rewrite(&$s, $nextgen_path) { 





//Check to see if file exists before continuing 


     if (file_exists($nextgen_path)) { 


     include($nextgen_path); 


     } 


     $s = preg_replace($urlin, $urlout, $s); 


     return $s; 


} #End of Google Tap Header

And the footer.php code is:

Code:




#Start of GoogleTap Footer 


if ($next_gen_ob == 1) { 


// Store Buffer in $contents 


$contents = ob_get_contents(); 


//Replace all standalone &'s with &amp's 


$contents = preg_replace("(&(?!([a-zA-Z]{2,6}|[0-9\#]{1,6})[\;]))", "&amp;", $contents); 


//Fix up for && in script and bad coding of middot;'s, and  ;'s 


$contents = str_replace(array("&amp;&amp;", "&amp;middot", "&amp;nbsp"), array("&&", "·", " ;"), $contents); 


//delete output buffer and stop buffering 


ob_end_clean(); 


//display modified buffer to screen 


echo replace_for_mod_rewrite($contents, $nextgen_path); 


} else { 


//if module has unusual method of including footer, make NExtGEn variables 


//global and try it again. If variables have values for NExtGEn it will succeed 


//else it will go to normal footer. 


global $next_gen_ob, $nextgen_path; 


   if ($next_gen_ob == 1 AND $nextgen_path != "") { 


   $contents = ob_get_contents(); 


   $contents = preg_replace("(&(?!([a-zA-Z]{2,6}|[0-9\#]{1,6})[\;]))", "&amp;", $contents); 


   $contents = str_replace(array("&amp;&amp;", "&amp;middot", "&amp;nbsp"), array("&&", "·", " ;"), $contents); 


   ob_end_clean(); 


   echo replace_for_mod_rewrite($contents, $nextgen_path); 


   } 


} 


#End of GoogleTap Footer

This is straight out of the 0.4 version of GT-NG.

Also, when I look at the GT'd links, they show up correctly when hovered over. But, when I click them, no dice. The Apache error log shows everything lower cased. This leads me to believe that either Apache is doing it.

Regards,
montego

Posted: Sat Oct 15, 2005 5:24 pm

Just another follow-up, I had a hunch that it might be the browser changing it, but both IE and FF worked the same.

Note, that it is this re-write rule that won't work:

Code:




RewriteRule ^downloads-(AddDownload|NewDownloads|MostPopular|TopRated).html modules.php?name=Downloads&op=$1

Bummer...
montego

Posted: Sat Oct 15, 2005 5:24 pm

Those are all lowercase on output. I need to see the .htaccess lines. That's where mod_rewrite does its magic.

Posted: Sat Oct 15, 2005 5:31 pm

I just tried the New on your site and it shows http://montegoscripts.com/downloads-NewDownloads.html So, as usual, I'm all confused Laughing

I need to see an example of what you're talking about.

Posted: Sat Oct 15, 2005 5:58 pm

I had to break apart the mod rewrite statements like this in order to get it to work (because it does come in lower case -- I can prove it from my access logs):

RewriteRule ^downloads-(NewDownloads|newdownloads).html modules.php?name=Downloads&op=NewDownloads

If you see, I had to hard-code the op variable. I wouldn't have to do that if I could simple pass the \\1 into it and it would retain the original incoming case.

Regards,
montego

Posted: Sat Oct 15, 2005 6:04 pm

Sorry, i meant to say "error logs" because it cannot find downloads-NewDownloads.html.

Posted: Sat Oct 15, 2005 7:33 pm

It may be coming in lower case but it is not Apache. Take out all of the download urlin and urlout and .htaccess except the one condition. Let's resolve this once and for all.

Posted: Sat Oct 15, 2005 8:42 pm

Ok, that did it! I have only two entries now and those two are working properly. I will need to work on the others and try and figure out just what is stepping on what!

thanks!
montego

Posted: Sat Oct 15, 2005 9:05 pm

I think you'll find that your urlin/out was stepping on itself Wink

Posted: Sat Oct 15, 2005 10:19 pm

yeah, just wish I knew POSIX better... I came from the "evil GUI world". Laughing

Thanks again!
montego

Posted: Mon Oct 17, 2005 8:31 am

I spoke too fast! Although it is working fine on my montegoscripts.com domain hosted on a *nix server, it is still NOT working on my development windows server! I am really needing both to work the same so that I can properly test before publishing. Now I am thinking again that it is either Windows or the combination of Apache on Windows, because here are the "facts":

1. Development is on Windows2000/Apache 1.3.x and Production is on *nix/Apache 1.3.x.

2. I copied the working .htaccess and GT-Downloads.php scripts from Production to Development so that I am using the exact same directives.

3. Development and Production Nuke files are exactly the same as I had just done an upgrade to the lastest patchset (3.1) and Sentinel so I always replace ALL my files accordingly.

4. Same browsers (IE and FF) used to test both in Development and Production.

5. "New" (NewDownloads) works fine in Production, but in Development, it gives a 404 file not found. The url coming in is downloads-NewDownloads.html and what is being shown in the error log in Development is "Unable to find downloads-newdownloads.html".

So, back to my original premise! Something is forcing the URL to lower case!!!!

Uuuugggghhhhhh.....

montego

Posted: Mon Oct 17, 2005 10:04 am

I run my development on WindowsXP, same Apache. That code above works on mine. I can't think of anything else right now except that Windows is not case sensitive. But, as to Apache, Apache will not convert on its own. If a conversion is being made it's being made in code that passes to Apache, regardless of how it appears.

Since you're only using 1 rule right now, please post the urlin, urlout, and .htaccess mod_rewrite rule that you are using on Windows.

Posted: Mon Oct 17, 2005 12:51 pm

hhhhmmmm... yes, this is strange. Same setup. My PC at home is WindowsXP and same issue as my laptop with Windows2000. All running the same Apache and GT/.htaccess/etc. Ok here are the lines:

Code:







$urlin = array(


"'(?<!/)modules.php\?name=Downloads&amp;op=(NewDownloads|MostPopular)'"


);





$urlout = array(


"downloads-\\1.html"


);





RewriteRule ^downloads-(AddDownload|NewDownloads|MostPopular|TopRated).html modules.php?name=Downloads&op=$1

In addition, here is the line from my Apache error log file:

Code:




[Mon Oct 17 11:51:08 2005] [error] [client 127.0.0.1] File does not exist: d:/http/montegoscripts/html/downloads-newdownloads.html

TIA,
montego