Best Effort OAuth1 token parsing

The Oauth1 standard requires the server to respond with Content-Type: application/x-www-form-urlencoded  However the code in qoauthoobreplyhandler.cpp:

void QOAuthOobReplyHandler::networkReplyFinished

Handles a number of other Content-Types, such as text/html, application/javascript, text/json, etc.

SUGGESTION:

Instead of issuing a warning and returning when the content type is unknown, attempt to parse the data and only return if the parsing failed:

diff --git a/src/oauth/qoauthoobreplyhandler.cpp b/src/oauth/qoauthoobreplyhandler.cpp
index 0409274..a2c03cd 100644
--- a/src/oauth/qoauthoobreplyhandler.cpp
+++ b/src/oauth/qoauthoobreplyhandler.cpp
@@ -92,7 +92,9 @@ void QOAuthOobReplyHandler::networkReplyFinished(QNetworkReply *reply)
{{ ret = object.toVariantMap();}}
{{ } else {}}
{{ qCWarning(lcReplyHandler, "Unknown Content-type: %s", qPrintable(contentType));}}
- return;
+ ret = parseResponse(data);
+ if(ret.isEmpty())
+ return;
{{ }}}

{{ Q_EMIT tokensReceived(ret);}}

This was discovered while attempting to authenticate against OpenStreetMap.org's auth server.  It returns "text/plain" and the url encoded oauth credentials resulting in failed authentication by QOAuth1.  The content type was ultimately set by the Ruby oauth-plugin project which is used by OSM.