http://www.couchbase.com/issues This file is an XML representation of an issue en-us 5.2.4 845 26-12-2012 http://www.couchbase.com/issues/browse/SPY-57 Spymemcached Java Client The current API forces the user to quote and encode the values that they&#39;re passing as parameters to the view. This should be simpler for the user, but it is not. <br/> <br/> For instance, if a user wanted the string asd&quot;asd as their couch &quot;key&quot; start then they&#39;d have to enter it &quot;\&quot;asd\&quot;asd\&quot;&quot;, which is a big burden on the developer. SPY-57

view setKey and friends should be quoting/escaping Strings passed in

Bug Minor Resolved Fixed Public Mike Wiederhold Matt Ingenthron Sat, 1 Oct 2011 12:25:07 -0500 Thu, 5 Apr 2012 12:34:40 -0500 Wed, 14 Dec 2011 15:19:53 -0600 2.8.0-dp 2.8.0-dp2 library 0 0 This was filed incorrectly. On String arguments for query parameters I automatically add quotes.<br/> <br/> private String getArg(String key, Object value) {<br/> &nbsp;&nbsp;&nbsp;&nbsp;// Special case<br/> &nbsp;&nbsp;&nbsp;&nbsp;if (key.equals(STARTKEYDOCID)) {<br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return key + &quot;=&quot; + value;<br/> &nbsp;&nbsp;&nbsp;&nbsp;}<br/> &nbsp;&nbsp;&nbsp;&nbsp;if (value instanceof Boolean) {<br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return key + &quot;=&quot; + ((Boolean) value).toString();<br/> &nbsp;&nbsp;&nbsp;&nbsp;} else if (value instanceof Integer) {<br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return key + &quot;=&quot; + ((Integer) value).toString();<br/> &nbsp;&nbsp;&nbsp;&nbsp;} else if (value instanceof Stale) {<br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return key + &quot;=&quot; + ((Stale) value).toString();<br/> &nbsp;&nbsp;&nbsp;&nbsp;} else {<br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return key + &quot;=\&quot;&quot; + value + &quot;\&quot;&quot;;<br/> &nbsp;&nbsp;&nbsp;&nbsp;}<br/> &nbsp;&nbsp;} This needs to url encode the string, not just quote it. it will be an API change, but it&#39;s the right thing to do. We also need to document that we&#39;re expecting the string to be in the platform&#39;s default character encoding, just to match what is commonly done in the JDK[1]. <br/> <br/> For example, someone could call setKey() with the following valid JSON: <br/> [&quot;val1&quot;,&quot;val2&quot;] <br/> And I think we just need to verify that we&#39;re converting it to URL encoded before sending the HTTP GET to the sserver: <br/> %5B%22val1%22%2C%22val2%22%5D <br/> <br/> The Java method call itself would look like: setKey(&quot;[\&quot;val1\&quot;, \&quot;val2\&quot;]&quot;); <br/> <br/> Or, more likely, they&#39;re using a JSON library to generate the string that&#39;s getting passed in to setKey(). In fact, our test should do that. <br/> <br/> I think this is the right behavior for setKey(String key) at the moment. We may later want to add a setRawKey() or a setKey(String key, CouchbaseClient.encoding encoding), where that second param is some kind of encoding scheme we come up with if there&#39;s not a good standard one to use. That does bring up that we should probably have setKey() that takes a String and a Charset argument, since they may not supply UTF-8. <br/> <br/> [1]: <a href="http://download.oracle.com/javase/6/docs/api/java/net/URLEncoder.html">http://download.oracle.com/javase/6/docs/api/java/net/URLEncoder.html</a> Lowering priority because this isn&#39;t the cause of a bug. It might be nice to add this in the future, but at the moment it is not a mandatory fix. Also, URLEncoder is deprecated so if this is fixed we should probably use something else for the encoding. I&#39;m going to close this issue because it appears that this is taken care of by the underlying http library. Rank 9537 Time In Status