It is checking of Querystring key null instead of its value

[tsibelman]

It is checking of Querystring key null instead of its value

[tsibelman]

Hi, thanks for the contribution but that was not a mistake. It was a real case I encountered.

Consider URLs in the form of
https://www.test.com/?param

Folowing code will print "Key is:null value is:param"
var paramCollection=HttpUtility.ParseQueryString(new Uri("https://www.test.com/?param").Query);
Console.WriteLine("Key is:"+(paramCollection.AllKeys[0]??"null")+" value is:"+paramCollection.GetValues(0)[0]);

[alex-piccione]

I see the difference now, thanks for the example.
The "Handles keys without values" didn't make me think about that scenario.
For the benefit of other readers:
?param is parsed as a null key with "param" value
?param= is parsed as "param" key with null value

So, the querystring parser is consideringnull the key and "param" the value, but the code (in the key == null branch) is doing the other way (using the value "param" as key name and setting null/empty the value):
values.Add(Uri.EscapeDataString(querystring[key]), new[] { $"{Uri.EscapeDataString(querystring[key])}=" });
that is "param" the key and null the value.

A quick test:
string url = "https://www.test.com/?aaa=1&aaa=2&value1&value2";
result in:
querystring[0] (key = "aaa"): "1,2"
querystring[1] (key is null) : "value1,value2"

I'm not saying it is wrong or not working, just an analysis to understand the behavior.

[tsibelman]

I think it very weird way of handling such cases if I would design NameValueCollection I would handling key and key= cases as the same one instead of treating one as key without value and the other as value without the key, but this an option we have right now.