[Solved] Multiple parameters to URL?

[vbmark]

I'm having some trouble with this custom Alias.

In FARR I want to be able to type something like this:

car ford fusion:2007

And have the results go to something like, for example...

http://www.car.com/q=ford+fusion%3a2007

I just can't figure out how to do the multiple parameters.

Thanks!

[Nod5]

Make a new alias called "car"
Under Regular Expression Pattern put

Code

^car (.*)$

and under results put

Code

car $$1|http://www.car.com/q=$$1

Then click Ok.

Type "car ford fusion:2007" in FARR to launch http://www.car.com/q=ford fusion:2007
Spaces isn't replaced with plus signs in the search but it may still work.

[mouser]

It may work exactly as Nod5 says, but you can also use
You can use $$u1 instead of $$1 to have FARR urlencode the spaces and special characters when doing a web search.

[vbmark]

Using $$u1 was what I needed to make it work how I wanted.  Thanks for that.

However, once I tried that, I see now that what I really need is for this:

car ford fusion:2007

to be turned into this:

...bla.com/?make=ford&model=fusion&year=2007

Is there a way to transform that FARR command into those parameters?

[mouser]

yep, you just need a regex with multiple capture groups.  see if you can figure it out, if not someone good at regex will be able to post a solution.

note that the farr alias group configuration where you specify the regular expression has a place where you can test things out -- you can use that to experiment.

[vbmark]

Sweet! Capture groups worked. And the testing place is great.

Thanks!!

[vbmark]

One other question.  When I use the Test place with capture groups the results box shows $$1= | $$2= | $$3=

Everything seems to go into $$1. What is the purpose of the other ones?

Thanks.

[skajfes]

The other ones ($$2 and $$3) are also capture groups, but what is actually captured depends on your regex. What did you put in regex?

[vbmark]

What did you put in regex?

^test (.*)\s?(.*)?:?(.*)?

[mouser]

Everything seems to go into $$1

then your regex is wrong, and you aren't capturing the different components into the different groups.

[Ath]

Should be \w* instead of .* within the round braces

[wjamoe]

if you define alias trigger (if you leave out the colon you should also leave it out of the regex)

  car make model:year

and define regex as car at the beginning of the line followed by two words separated by a single space followed by a  colon and a third word

  ^car (.*) (.*):(.*)

results where $$1 substitues the first word matching (.*)

  http://www.car.com/?make=$$1&model=$$2&year=$$3

or alternatively from your original request:
  http://www.car.com/q=$$1+$$2%3a$$3

when you type
 
  car ford fusion:2009

it turns into
  http://www.car.com/?make=ford&model=fusion&year=2009

or when you use the alternative
  http://www.car.com/q=ford+fusion%3a2009

(BTW www.car.com does not work with these paramaters?)

[Ath]

 ^car (.*) (.*):(.*)

results where $$1 substitues the first word matching (.*)

This only works when typing exactly 1 space after the first word.

A more reliable regex for capturing separate words, independent from the number of spaces, would be:

Code

^car (\w*)\s*(\w*)\s*:?\s*(\w*)

This way you can even have 0 or more spaces around the colon.

[vbmark]

(BTW www.car.com does not work with these paramaters?)

That domain was just an example.

A more reliable regex for capturing separate words, independent from the number of spaces, would be:

Code

^car (\w*)\s*(\w*)\s*:?\s*(\w*)

This way you can even have 0 or more spaces around the colon.

Good stuff!  Thanks!