Fix query parsing when value has = characters

[jimmy-park]

The current query parser handles "?a=b=c&..." as ("a", "c") which should be ("a", "b=c").
We should not use split() when parsing query strings.
So I need to implement a string divider which just splits in half.
Here are possible outcomes of the string divider.

divide("", '=')      -> ("", "")
divide("=", '=')     -> ("", "")
divide(" ", '=')     -> (" ", "")
divide("a", '=')     -> ("a", "")
divide("a=", '=')    -> ("a", "")
divide("=b", '=')    -> ("", "b")
divide("a=b", '=')   -> ("a", "b")
divide("a=b=", '=')  -> ("a", "b=")
divide("a=b=c", '=') -> ("a", "b=c")

Note

Whitespaces aren't trimmed!

[yhirose]

@jimmy-park thanks for the pull request and the code looks good to me. Where do you find the information of "?a=b=c&..." in the HTML spec or RFCs?

[jimmy-park]

@yhirose RFC does not specify the format of query. Any string between '?' and '#' could be a query . I think key=value pair is just a well known convention that people use.
https://datatracker.ietf.org/doc/html/rfc3986#section-3.4

And I've seen some queries with '=' in the value actually being used at work.

[yhirose]

@jimmy-park, ok. Could you please check how other HTTP libraries handle these queries? If they do the same, I have no issue with this pull request.

[jimmy-park]

@yhirose I found one.
https://github.com/drogonframework/drogon/blob/96919df488e0ebaa0ed304bbd76bba33508df3cc/lib/src/HttpRequestImpl.cc#L145-L163

[yhirose]

@jimmy-park thanks for the improvement!