Crafting a regex for mod_rewrite with more than nine parameters

Question

Say for example, I have a URL in this format:

http://example.com/(mobile-or-desktop-flag)/(major-venue-category)/(smaller-venue-category)/(venue-name)/(month)-(day)-(year)/(picture-number)/(user-action-on-picture)/(request-ajax-or-not)

where the text in parenthesis represent the parameter data.

and some example URLs are:

http://example.com/desktop/big/nightclubs/bloke/jan-1-2015/1/download/no-ajax
http://example.com/desktop/small/lounges/oko-blu/feb-2-2014/6/share/no-ajax
http://example.com/mobile/tiny/bars/honest-lawyer/mar-3-2013/9/report/ajax

With mod_rewrite, I know I can use $1 through $9 for the first nine parameters, but here, I have 10 parameters, and $10 is considered first parameter with a zero added to it. See RewriteCond at http://httpd.apache.org/docs/current/mod/mod_rewrite.html

I'm just wondering, is there any way one can include a variable 10th parameter in a rewrite rule, or will I have to merge two parameters into one and extract them into two via a server-side script?

Answer 1

is there any way one can include a variable 10th parameter in a rewrite rule

As far as I'm aware you can only use the backreferences $0..$9 ($1 to $9 references the parenthesised/captured groups and $0 references the entire pattern that is matched) in Apache regex (eg. mod_rewrite, etc) to reference captured groups in the RewriteRule pattern (%0..%9 for the last matched CondPattern).

Workarounds:

1. Like you suggest, you could merge two or more parameters into one and let your server-side script parse the URL. The date (month-day-year) is the obvious one. You could simply use mod_rewrite to validate the URL and let your server-side script parse the entire URL into its component parts (a front-controller).

2. As an extension to the above. You could merge two or more params and rewrite the URL a second time, combining the params into the correct format required by the script you are rewriting to. For example, take the month-day-year parameter(s). If the script you were rewriting to required the date in universal date format (year-month-day) then you could first capture and convert this into one parameter (together with the remainder of the URL). For example:

   RewriteRule ^(\w+/[\w-]+/[\w-]+/[\w-]+)/([a-z]{3})-(\d{1,2})-(\d{4})/(.*) /$1/$4-$2-$3/$5
   RewriteRule ^(desktop|mobile)/([\w-]+)/([\w-]+)/([\w-]+)/(\d{4}-[a-z]{3}-\d{1,2})/(\d{1,3})/(download|share|report)/(no-ajax|ajax) /script.php?platform=$1&majorcat=$2&minorcat=$3&vname=$4&date=$5&picnum=$6&action=$7&ajax=$8 [L]
   

3. Avoid capturing groups for any groups you don't need in the current rewrite. ie. Make it a non-capturing group by prefixing with ?:, for example: (?:desktop|mobile).

Answer 2

You can also utilize te behavior of .htaccess, remembering that the generated url will be passed again and process the url in cycles, in the first you can deal with a group of parameters, and in the second with the other group...

The image is Taken from here

# This is a .htaccess file
# This will match the full pattern, that has to be adapted to any case, in this particular case will match "one/two/three/four/five/six/seven/eight/nine/ten/" ten parameters path
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_URI} !(all_parameters_matched|processed_first_group)
RewriteRule ^((?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/(?:.+)\/?)$ all_parameters_matched/$1 [QSA,L]

# This deals with the first group of parameters
RewriteRule ^all_parameters_matched\/([^\/]+)\/([^\/]+)\/([^\/]+)\/(.*)$ processed_first_group/$4?param1=$1&param2=$2&param3=$3 [QSA]

# This deals with the second group of parameters
RewriteRule ^processed_first_group\/([^\/]+)\/([^\/]+)\/(.*)$ /script.php?param4=$1&param5=$2&the_rest=$3 [QSA]

In the example above it processes only five parameters in two parts, but you could get the idea.