Query parameters are not removed when resolving URLs

[fidian]

When navigating to "test.html" from "http://example.com/somewhere.html?a=b#c", I expect the URL to be "http://example.com/test.html", not "http://example.com/test.html?a=b". Here's simple node REPL code to replicate it.

var URI = require('./URI');
URI('asdf', 'http://example.com/somewhere.html?a=b#c').toString();

[rodneyrehm]

Looking at examples in the RFC most of the following tests fail:

test("absoluteTo", function() {
    // http://tools.ietf.org/html/rfc3986#section-5.4
    var uri;
    var base = 'http://a/b/c/d;p?q';
    var map = {
        // normal
        // "g:h"           :  "g:h", // identified as URN
        "g"             :  "http://a/b/c/g",
        "./g"           :  "http://a/b/c/g",
        "g/"            :  "http://a/b/c/g/",
        "/g"            :  "http://a/g",
        "//g"           :  "http://g",
        "?y"            :  "http://a/b/c/d;p?y",
        "g?y"           :  "http://a/b/c/g?y",
        "#s"            :  "http://a/b/c/d;p?q#s",
        "g#s"           :  "http://a/b/c/g#s",
        "g?y#s"         :  "http://a/b/c/g?y#s",
        ";x"            :  "http://a/b/c/;x",
        "g;x"           :  "http://a/b/c/g;x",
        "g;x?y#s"       :  "http://a/b/c/g;x?y#s",
        ""              :  "http://a/b/c/d;p?q",
        "."             :  "http://a/b/c/",
        "./"            :  "http://a/b/c/",
        ".."            :  "http://a/b/",
        "../"           :  "http://a/b/",
        "../g"          :  "http://a/b/g",
        "../.."         :  "http://a/",
        "../../"        :  "http://a/",
        "../../g"       :  "http://a/g",
        // abnormal
        "../../../g"    :  "http://a/g",
        "../../../../g" :  "http://a/g"
    };

    for (var key in map) {
        uri = URI(key).absoluteTo(base);
        equal(r + "", map[key], 'reference resolution ' + key);
    }
});

@djcsdy have you got anything to say about query+hash in absoluteTo()? otherwise I'd just remove them…

[fidian]

I believe the query should stay the same if we are just resolving against a hash. See the "#s" example in your code.

[djcsdy]

@rodneyrehm Those tests derived from the spec look pretty complete.

The query component of a base URL should be retained if and only if:

• The relative URL to be resolved is the empty string (""); OR
• The relative URL to be resolved contains a hash component but no other components ("#foo").

Otherwise it should be discarded.

[rodneyrehm]

I agree. I'll look into fixing absoluteTo() this week. It will break relativeTo() tests - they'll have to be adjusted as well.

[rodneyrehm]

fixed in master, will be included in the next release