lighttpd-Support

[PSIplus]

However... Are there any plans to ship it with future releases? And can I help?

[till]

Originally Posted by PSIplus

However... Are there any plans to ship it with future releases?

yes.

And can I help?

yes Please register as ISPConfig developer to get SVN write access to the latest ISPConfig dev releases.

http://server1.howtoforge.com/forums...read.php?t=135

[jeremyh]

A while ago I started writing a config parser for Lighttpd in PHP. It's been a while since I've worked on it, and is read-only. But it could be a good base if you need a full parser.

http://svn.jeremyh.net/r/LighttpdParser

[jameso]

Well, I am totally new to ISPConfig.

I managed to install it, and also use your instructions for using lighttpd instead of apache.

Everything seems ok so far, however I need to test things a lot more.

Thanks for the useful info PSIplus!

[Ovidiu]

wow. I'd love to see this in a more stable version, but I have to wait until edge comes out, make that upgrade and then I'll check back to see how far this project has proceeded...

[Ovidiu]

@PSIplus

can you give me a hand with a similar problem?
I would like to manually compile and install lighttpd on my debian box, have it lsitening on a different port i.e. 88 and import the vhosts from apache.

I plan on using mod_proxy on my apache2 so apache2 will stay the default webserver BUT have it forward static content to lighttpd (I have figured out the forwarding part, but need help with the compilation on debian (running 3.1)

If this works out I'd like to also use lighttpd for ruby on rails

can you help? I could open another thread? or you could send me a pm or whatever :-)

[jlp09550]

Thanks for the guide.. but can someone please convert it to an easier-to-read format? It's really hard to read what the guide says, muchless properly do it.

Thanks.

[justdave]

Hey guys I'd love to see ISPConfig have built in handling for lighttpd. I would be willing to donate some money towards making this a reality. I also know of one other person that would do the same.

If there are others here using Lighty perhaps we can all chip in?

[Croydon]

PSIplus, thanks for the great script

As lighttp does not recognize .htaccess files I ran into problems migrating to lighttp.
I extended your script so it now has a (very) rudimentary support for htaccess.

first I changed the modules section to include mod_auth and mod_rewrite

echo "server.modules=(\"mod_auth\",\"mod_access\",\"mod_cgi\",\"mod_accesslog\",\"mod_fastcgi\",\"mod_ssi\",\"mod_alias\",\"mod_rewrite\")\n";

then I added function calls to the end of the script

...

    echo "\n\t## Error-Handling ##\n";
    echo "\talias.url=(\"/error/\"=>\"$docroot/$prefix_web$parm[id]/web/error/\")\n";
    if($parm["error"])
    {
        /*
        ErrorDocument 400 /error/invalidSyntax.html
        ErrorDocument 401 /error/authorizationRequired.html
        ErrorDocument 403 /error/forbidden.html
        ErrorDocument 404 /error/fileNotFound.html
        ErrorDocument 405 /error/methodNotAllowed.html
        ErrorDocument 500 /error/internalServerError.html
        ErrorDocument 503 /error/overloaded.html
        AliasMatch ^/~([^/]+)(/(.*))? /var/www/web12/user/$1/web/$3
        AliasMatch ^/users/([^/]+)(/(.*))? /var/www/web12/user/$1/web/$3
        */
    }
    // search for .htaccess files and prevent direct access
    echo "\n\turl.access-deny = ( \".htaccess\", \".htpasswd\")\n";
    list($data, $rewdata) = recurse_htaccess("", "$docroot/$prefix_web$parm[id]/web");
    
    echo $data;
    echo $rewdata;

    echo "}\n\n";
...

finally I added my new functions to the end of the script

function recurse_htaccess($curdir, $basedir) {
    $dir = opendir($basedir . $curdir);
    
    $data = "";
    $rewdata = "";
    
    while (false !== ($file = readdir($dir))) {
        if ($file != "." && $file != "..") {
            if(is_dir($basedir . $curdir . "/" . $file)) {
                list($ndata, $rdata) = recurse_htaccess($curdir . "/" . $file, $basedir);
                $data .= $ndata;
                $rewdata .= $rdata;
            } elseif($file == ".htaccess") {
                list($ndata, $rdata) = get_htaccess($curdir . "/" . $file, $basedir);
                $data .= $ndata;
                $rewdata .= $rdata;
            }
        }
    }
    
    return array($data, $rewdata);
}

function get_htaccess($file, $basedir) {
    $data = "";
    
    $rewexp = array();
    
    $fpath = $basedir . $file;
    $file = str_replace(".htaccess", "", $file);

    $fp = fopen($fpath, "r");
    if(!$fp) return "";

    if(!file_exists($fpath)) return "";

    $lines = file($fpath);
    if(!is_array($lines) || count($lines) < 1) return "";

    $match = false;
    $data = "";
    $data .= "\t\$HTTP[\"url\"] =~ \"^$file\" {\n";
    $data .= "\t\tauth.backend = \"htpasswd\"\n";
    
    $rewbase = "";
    
    $reqdata = "\"$file\" => (\n";
    // method / realm / require...
    $first = true;
    foreach($lines as $line) {
        if(preg_match("'authtype\s+(\w+)'is", $line, $matches)) {
            $match = true;
            if(!$first) $reqdata .= ",\n";
            $reqdata .= "\t\t\t\"method\" => \"" . strtolower(trim($matches[1])) . "\"";
        } elseif(preg_match("'authname\s+\"?([^\"]+)\"?'is", $line, $matches)) {
            $match = true;
            if(!$first) $reqdata .= ",\n";
            $reqdata .= "\t\t\t\"realm\" => \"" . trim($matches[1]) . "\"";
        } elseif(preg_match("'require\s+\"?([^\"]+)\"?'is", $line, $matches)) {
            $match = true;
            if(!$first) $reqdata .= ",\n";
            $reqdata .= "\t\t\t\"require\" => \"" . trim($matches[1]) . "\"";
        } elseif(preg_match("'authuserfile\s+\"?([^\"]+)\"?'is", $line, $matches)) {
            $match = true;
            if(trim($matches[1]) != "") $data .= "\t\tauth.backend.htpasswd.userfile = \"" . trim($matches[1]) . "\"\n";
        } elseif(preg_match("'rewritebase\s+(\S+)'is", $line, $matches)) {
            $rewbase = trim($matches[1]);
        } elseif(preg_match("'rewriterule\s+(\S+)\s+(\S+)'is", $line, $matches)) {
            $srch = trim($matches[1]);
            $rewexp["$srch"] = trim($matches[2]);
        }
        
        $first = false;
    }
    $data .= "\t\tauth.require = ( ";
    $data .= $reqdata . ")\n\t\t)\n";
    $data .= "\t}\n";
    if($match == false) $data = ""; // reset it

    $rewdata = "";

    if(count($rewexp) > 0) {
        $rewdata = "\n\turl.rewrite-once = ( ";
        $first = true;
        foreach($rewexp as $search => $replace) {
            if($first == false) $rewdata .= ",";
            if($rewbase != "") {
                $search = preg_replace("'^\^'", "^$rewbase", $search);
            }
            $rewdata .= "\n\t\t\"$search\" => \"$replace\"";
            
            $first = false;
        }
        $rewdata .= ")\n";
    }
    
    

    return array($data, $rewdata);
}

Sorry for the missing code comments... But I had no time for these.

As I said this is VERY basic .htaccess support...
I managed to get all my .htaccess directory access limits running.
My basic rewrite rules are working, too (no rewrite conditions).

Host-based or file-based access limit in .htacess files does not yet work, maybe I'll get this to work later

[Murder4Al]

I just started using this script, and discoverd two flaws:
- it doesn't support wildcard domains, something I frequently use
- the .htaccess patch is to slow on big sites (and i mean BIG, over 100k folder really kills the perfornance)

Below is a patched version that adds wildcard support.
The .htaccess patch is included but disabled
I also added a few extra lighttpd options; especially a port parameter, so you can run lighttpt next to your apache2 to test the performance/validy of this script.

(had to rename file to .txt for the forum to accept it)