ensure top level req before processing query

[majuscule]

DuckPAN is forwarding the query parameter to multiple calls for some reason (is this intended behavior?):

127.0.0.1 - - [07/Oct/2013:03:48:54 -0400] "GET /?duckduckhack_ignore=1&q=perl%20jobs%20in%20san%20francisco&l=us-en&p=1&s=0 HTTP/1.1" 200 0 "http://0.0.0.0:5000/" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.76 Safari/537.36"
127.0.0.1 - - [07/Oct/2013:03:48:55 -0400] "GET /y.js?s=1&q=perl%20jobs%20in%20san%20francisco&l=us-en HTTP/1.1" 200 0 "http://0.0.0.0:5000/" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.76 Safari/537.36"
127.0.0.1 - - [07/Oct/2013:03:48:56 -0400] "GET /y.js?x=1&q=perl%20jobs%20in%20san%20francisco&l=us-en&safe=1 HTTP/1.1" 200 0 "http://0.0.0.0:5000/" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.76 Safari/537.36"
127.0.0.1 - - [07/Oct/2013:03:48:56 -0400] "GET /?duckduckhack_ignore=1&q=perl%20jobs%20in%20san%20francisco HTTP/1.1" 200 0 "http://0.0.0.0:5000/" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/29.0.1547.76 Safari/537.36"

Whatever the reason for that, Web.pm is not properly checking that a query is placed against the proper resource /. This additional check avoids the duplicated processing, and with it, prevents the confusingly repeated Data::Printer output.

Resolves issue #17.

[yegg]

It really shouldn't be writing y.js at all to this as well -- should open a PR internally to prevent that for these special duckduckhack queries.

[yegg]

Made the internal PR.

[moollaza]

LGTM 👍

[yegg]

released