RegExp Language Routing - php

Question

Good Day,
I am in need of a Route-RegExp based on client-language for a website.
It should be like this:
Relative URL / Route:
/(No-Language) -> /?lng=(someDefaultLanguage)
/(No-Language)/ -> /?lng=(someDefaultLanguage)
/lngCode/page -> /page/?lng=lngCode
/lngCode/page/ -> /page/?lng=lngCode
/lngCode/pageL1/pageL2 -> /pageL1/pageL2/?lng=lngCode
/language/page?param=Value -> /page/?lng=lngCode&param=Value
(Notice the trailing slashes on some lines)
Tree structure is, ...well infinite :)
There are cases with single and multiple URL-Params.
I'm absolute no regex wizard, I managed this result in uhm, ...hours:
/^\/([a-z]{2})(?:(.*[^\?])|^$)((?:[\/\?]).*|^$)/
Please don't ask me what I was trying to route there. I am sooooo new to regex.
Thank you in advance
--
Edit for clarification (I hope):
Basically it is this concept: (It is internal routing, no redirection if I didnt mention.)
The language-parameter (as directory-style) must be grabbed from the 1st url and attached as a real parameter named, "lng". The directory-parameter should disappear.
If there are already other parameters, they need to be attached as well (?/&-case).
If there is no language given (=default-language), there is no directory-style-parameter in the url. Would be nice if still a ?lng=en parameter can be attached.
Examples:
localhost/blogpage/coolentry (default language)
localhost/de/blogpage/coolentry
localhost/es/blogpage/coolentry
localhost/blogpage/ -> localhost/blogpage/?lng=en
localhostde/de/blogpage/ -> localhost/blogpage/?lng=de
localhost/blogpage/coolentry/ -> localhost/blogpage/coolentry/?lng=en
localhost/de/blogpage/coolentry/ -> localhost/blogpage/coolentry/?lng=de
localhost/de/blogpage/coolentry/?entryPage=1 -> localhost/blogpage/coolentry/?lng=de&entryPage=1
It gets routed always with a real language parameter.
I have as well edited the first post, there was a confusing typo in it.

Answer 1

Sorry for the delay #hcm.
Well, I gotta tell ya. I spent 10 minutes writing the original regex.
Tested in Perl, everything worked great.
Then I go to dump it into an online php tester and I get this "Undefined offset" error/warning.
Capture groups 2,3,4 are optional, so I do a (?: ( capture ) )? but php is such a mess
you can't even test for undefined group.
Update
#hcm - Ok, figured it out. Those online testers don't translate CRLF's to LF's.
Therefore, when using multiline mode $ is boundry before a newline or end of string.
^ is after a newline or beginning of a string, which is no problem.
So, $ won't match before a CR only before a LF. A workaround is probably using the
\R any linebreak construct but that is not a boundry, its an actual character.
What I did was to cure this is to use (?: $ | (?= \r) ) outside of assertions, and
(?: $ | \r ) inside assertions. This cures all problems.
After reading your message, I've changed the regex so that every part is optional, but
still positional.
The 4 optional parts are as follows.
1. Before the lang code.
2. The lang code.
3. After the lang code.
4. The parameters.
No part will run over the other.
All leading /'s are taken out of each part (not part of the match),
while internal slashes are left in place.
All that's left is to construct the new url as you wish.
Let me know how this turns out or if you need a little tweak.
PHP code:
// Go to this website:
// http://writecodeonline.com/php/
// Cut & paste this into the code box, hit run.
$text = '
invalid
/
/de
/de/coolentry
localhost/
localhost/blogpage
localhost/blogpage/
localhost/blogpage/de/
/localhost/blogpage/coolentry/famous invalid
/root/blog/page/cool/entry/?entryPage=1&var1=A&var2=B
localhost/blogpage/coolentry/
localhost/blogpage/de/coolentry/
localhost/blogpage/de/coolentry/
localhost/blogpage/de/coolentry/?entryPage=1
localhost/blogpage/coolentry/?entryPage=2
';
$str = preg_replace_callback('~^(?![^\S\r\n]*(?:\r|$))(?|(?!/[a-z]{2}[^\S\r\n]*(?:/|(?:\r|$)))/?((?:(?!/[a-z]{2}[^\S\r\n]*(?:/|(?:\r|$))|\?|/[^\S\r\n]*(?:\r|$))\S)*)|())(?|/([a-z]{2})(?=/|[^\S\r\n]*(?:\r|$))|())(?|/((?:(?!/[^\S\r\n]*(?:\r|$))[^?\s])+)|())(?|/\?((?:(?!/[^\S\r\n]*(?:\r|$))\S)*)|())/?[^\S\r\n]*(?:$|(?=\r))~m',
function( $matches )
{
///////////////// URL //////////////////
$url = '';
// Before lang code -- Group 1
if ( $matches[1] != '' ) {
$url .= '/' . $matches[1];
}
// After lang code -- Group 3
if ( $matches[3] != '' ) {
$url .= '/' . $matches[3];
}
///////////////// PARAMS //////////////////
$params = '/?lng=';
// Lang code -- Group 2
if ( $matches[2] != '' ) {
$params .= $matches[2];
}
else {
$params .= 'en'; // No lang given, set default
}
// Other params
if ( $matches[4] != '') {
$params .= '&' . $matches[4];
}
///////////////// Check there is a Url //////////////////
if ( $url == '' ) { // No url given, set a default
$url = '/language'; // 'language', 'localhost', etc...
}
///////////////// Put the pieces back together //////////////////
$NewURL = $url . $params;
return $NewURL;
},
$text);
print $str;
output:
invalid
/language/?lng=en
/language/?lng=de
/coolentry/?lng=de
/localhost/?lng=en
/localhost/blogpage/?lng=en
/localhost/blogpage/?lng=en
/localhost/blogpage/?lng=de
/localhost/blogpage/coolentry/famous invalid
/root/blog/page/cool/entry/?lng=en&entryPage=1&var1=A&var2=B
/localhost/blogpage/coolentry/?lng=en
/localhost/blogpage/coolentry/?lng=de
/localhost/blogpage/coolentry/?lng=de
/localhost/blogpage/coolentry/?lng=de&entryPage=1
/localhost/blogpage/coolentry/?lng=en&entryPage=2
Regex
# '~^(?![^\S\r\n]*(?:\r|$))(?|(?!/[a-z]{2}[^\S\r\n]*(?:/|(?:\r|$)))/?((?:(?!/[a-z]{2}[^\S\r\n]*(?:/|(?:\r|$))|\?|/[^\S\r\n]*(?:\r|$))\S)*)|())(?|/([a-z]{2})(?=/|[^\S\r\n]*(?:\r|$))|())(?|/((?:(?!/[^\S\r\n]*(?:\r|$))[^?\s])+)|())(?|/\?((?:(?!/[^\S\r\n]*(?:\r|$))\S)*)|())/?[^\S\r\n]*(?:$|(?=\r))~m'
^ # BOL
(?! # Not a blank line, remove to generate a total default url
[^\S\r\n]*
(?: \r | $ )
)
(?| # BEFORE lang code
(?!
/ [a-z]{2} [^\S\r\n]* # not lang code
(?:
/
| (?: \r | $ )
)
)
/? # strip leading '/'
( # (1 start)
(?:
(?!
/ [a-z]{2} [^\S\r\n]* # not lang code
(?:
/
| (?: \r | $ )
)
|
\? # not parms
|
/ [^\S\r\n]* # not final slash
(?: \r | $ )
)
\S
)*
) # (1 end)
|
( ) # (1)
)
(?| # LANG CODE
/ # strip leading '/'
( [a-z]{2} ) # (2)
(?=
/
| [^\S\r\n]*
(?: \r | $ )
)
|
( ) # (2)
)
(?| # AFTER lang code
/ # strip leading '/'
( # (3 start)
(?:
(?! # not final slash
/ [^\S\r\n]*
(?: \r | $ )
)
[^?\s] # not parms
)+
) # (3 end)
|
( ) # (3)
)
(?| # PARAMETERS
/ \? # strip leading '/?'
( # (4 start)
(?:
(?! # not final slash
/ [^\S\r\n]*
(?: \r | $ )
)
\S
)*
) # (4 end)
|
( ) # (4)
)
/?
[^\S\r\n]* # EOL
(?:
$
| (?= \r )
)