Here is a link to the function in action:
http://dev.relatebase.com/devteam/development/site_spider/config_v200.php
Here is the function source code:
function enhanced_parse_url($parse){
/**
2008-07-19:
version 1.01. big changes in the parsing order; also files assumed to be files based on having an extension like html, php, etc.
**/
global $enhanced_parse_url, $sp;
if($parse==''){
return;
}else if($parse=='.'){
$enhanced_parse_url['file']='';
$enhanced_parse_url['path']='';
return true;
}else if($parse=='..'){
$enhanced_parse_url['file']='';
$enhanced_parse_url['uplevel']=1;
return true;
}
// handle fragment
$result= strrpos($parse, '#');
if(strlen($result)){
$pos= -(strlen($parse)-$result);
if($pos+1<0){
$enhanced_parse_url['fragment']=substr($parse,$pos+1);
$parse=substr($parse,0,$pos);
}else{
//0
$parse=substr($parse,0,strlen($parse)-1);
}
}
// handle querystring
$result= strpos($parse, '?');
if(strlen($result)){
if($result+1==strlen($parse)){
//no query string
$parse=substr($parse,0,strlen($parse));
}else{
$enhanced_parse_url['query']=substr($parse,$result+1);
$parse=substr($parse,0,$result);
}
}
//remove the protocol and domain from the front
#note: regex for file: needs to handle a triple-backslash character as well
$schema=array(
'file:///'=>'/^file:((\/\/\/)|((\\\){3}))/i',
'http://'=>'/^http:\/\//i',
'https://'=>'/^https:\/\//i',
'ftp://'=>'/^ftp:\/\//i',
'mailto:'=>'/^mailto:/i'
);
$schemaL=array(
'file:///'=>8,
'http://'=>7,
'https://'=>8,
'ftp://'=>6,
'mailto:'=>7
);
foreach($schema as $n=>$v){
if(preg_match($v,$parse)){
$enhanced_parse_url['protocol']=$n;
$parse=substr($parse, $schemaL[$n], strlen($parse)-$schemaL[$n]);
break;
}
}
if($enhanced_parse_url['protocol']=='http://' || $enhanced_parse_url['protocol']=='https://' || $enhanced_parse_url['protocol']=='ftp://'){
//extract the domain name
if($pos=strpos($parse,'/')){
$enhanced_parse_url['domain']=strtolower(substr($parse,0,$pos));
//this leaves the forward slash for uniformity
$parse=substr($parse,$pos,4096);
}else{
//only domain remains
$enhanced_parse_url['domain']=strtolower($parse);
return true;
}
}else if($enhanced_parse_url['protocol']=='mailto:'){
//this seems appropriate at this point, should not be any more components
$enhanced_parse_url['email']=trim($parse);
$enhanced_parse_url['identity']='Email address';
return true;
}
/******************
we could have:
file.php
/file.php
./file.php
../file.php
../../file.php
path/to/file.php
/path/to/file.php
./path/to/file.php
../path/to/file.php
../../path/to/file.php
path/to/
path/to
if the last character is a /, the file is default (directory_default)
******************/
//remove the level indicators
if(preg_match('/^([.\/]+)/',$parse,$a)){
$enhanced_parse_url['level_indicators']=$a[1];
if(substr($a[1],0,1)=='/'){
$enhanced_parse_url['positioning']='absolute';
}else if(substr($a[1],0,3)=='../'){
$enhanced_parse_url['positioning']='relative';
}else{
$enhanced_parse_url['positioning']='samefolder';
}
switch(true){
case $a[1]=='./':
$parse=substr($parse,2,strlen($parse)-2);
break;
case $a[1]=='/':
$parse=substr($parse,1,strlen($parse)-1);
break;
default:
// ../../../ etc.
$enhanced_parse_url['uplevel']=strlen($a[1])/3;
$parse=substr($parse,strlen($a[1]),strlen($parse)-strlen($a[1]));
}
}
//now we are left with path and file
if(strlen($parse)==0){
$enhanced_parse_url['file']='(directory_default)';
return true;
}else if(preg_match('/\/$/',$parse)){
$enhanced_parse_url['file']='(directory_default)';
$parse=substr($parse,0,strlen($parse)-1);
$haveFile=true;
}
if($haveFile){
$enhanced_parse_url['path']=$parse;
return true;
}
//analyze final position
$a=explode('/',$parse);
$possible=$a[count($a)-1];
if($enhanced_parse_url['query'] || $enhanced_parse_url['fragment'] || preg_match('/\.(htm|html|asp|php|jsp|asp|txt|pdf|xls|jpg|gif|png|css|js)([0-9]*)$/i',$possible)){
$enhanced_parse_url['file']=array_pop($a);
if(count($a)>0)$enhanced_parse_url['path']=implode('/',$a);
return true;
}
$enhanced_parse_url['path']=$parse;
return true;
}
please FIRST see how it works - it doesn't always return all nodes of the $enhanced_parse_url array unless they are present - then try and break it and email me or reply if this happens. You are free to use this.
Samuel
