On Thu, 4 May 2017 18:26:42 -0400
Alex wrote:
> Hi,
>
> >>> I found a local version which maybe did the trick
> >>>
> >>> redirector_pattern
> >>>
> >>> m'^https?:/*(?:\w+\.)?google(?:\.\w{2,3}){1,2}/url\?.*?(?<=[?&])q=(.*?)(?:$|[&\#])'i
> >>>
> >>
> Yes, but I don't understand how that equates to an eventual score.
I haven't used these, but by the look of it it's trying to identify
the encoded URI so that other rules can see it.
> Perhaps I don't understand regex's well enough, but I don't understand
> what it does with the redirector site portion and the target portion.
In rules you often see ?: used in brackets to prevent the matching
text being captured as this makes it more efficient, e.g. (?:\w+\.)
in the above. Towards the end of the pattern is (.*?) which is
used to capture the target uri.
.*? is like .*, but matches the shortest run of characters it can,
rather than the longest.
