HttpContext Request.QuerystringEntries Wrong about watsonwebserver

[jchristn]

Hi @rangsutu88 thanks for your note. I will look into it ASAP hopefully today.

[jchristn]

Hi @rangsutu88 - these are being URL decoded. The / character is a reserved character which has to be encoded. Please refer to p2.2 of the RFC: https://tools.ietf.org/html/rfc3986#page-12 (there's a healthy discussion about this on stackoverflow here: https://stackoverflow.com/questions/1455578/characters-allowed-in-get-parameter)

These are being processed and decoded by WebUtility.UrlDecode (see https://docs.microsoft.com/en-us/dotnet/api/system.net.webutility.urldecode?view=netcore-3.1) to accomplish this.

Cheers, Joel

[rangsutu88]

hi @jchristn .
I think this issue come from WatsonWebserver Request module. at below code
at the line 243 in HttpRequest.cs. Sorry if i'm wrong

region Process-Raw-URL-and-Populate-Raw-URL-Elements

                while (RawUrlWithoutQuery.Contains("//"))
                {
                    RawUrlWithoutQuery = RawUrlWithoutQuery.Replace("//", "/");
                }

[jchristn]

Hi @rangsutu88 that is being done over the raw URL i.e. /hello, not the query string entries i.e. ?key=test//test. The query string is handled by the decode function:

tempKey = WebUtility.UrlDecode(tempKey);
if (!String.IsNullOrEmpty(tempVal)) tempVal = WebUtility.UrlDecode(tempVal);

Please see lines 326-327, 358-360, 375-376, and 384-386 in HttpRequest.cs.

Which reminds me I have some refactoring to do in that particular method. Thanks!

[rangsutu88]

hi @jchristn ,
Thank your for your explanation.