http://dev.clojure.org/jira This file is an XML representation of an issue en-us 4.4 649 25-07-2011 http://dev.clojure.org/jira/browse/CLJ-950 Clojure <p> ((fn [] true)) ; true<br/> (#(true)) ; classcast exception</p> <p> ((fn [])) ; nil<br/> (#()) ; ()</p> <p> (some (fn <span class="error">&#91;_&#93;</span>_) <span class="error">&#91;nil false 0 1&#93;</span>) ; 0<br/> (some #(%) <span class="error">&#91;nil false 0 1&#93;</span>) ; NPE</p> CLJ-950

Function literals behavior differ from that of fns

Defect Major Closed Declined Unassigned Víctor M. Valenzuela reader Thu, 8 Mar 2012 06:13:29 -0600 Fri, 1 Mar 2013 12:46:59 -0600 Thu, 8 Mar 2012 12:28:16 -0600 0 0 <p>This is no defect. Function literals must have a function (or macro or special form) as first symbol.<br/> So your examples should be written like so:</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>user&gt; (#(do true)) true user&gt; (#(do)) nil user&gt; (some #(do %) [nil false 0 1]) 0 </pre> </div></div> <p>It makes sense. However (and correct me if I'm wrong) there should be little problem in making them fully equivalent to fns, resulting in a more concise and consistent API.</p> <p>Please consider re-opening the issue as a feature request.</p> <p>Regards - Víctor.</p> <p>The reader docs at <a href="http://www.clojure.org/reader">http://www.clojure.org/reader</a> say that #() is not a replacement for (fn [] ...). You can't make it more equivalent to fn without making it much harder to understand. Let me explain that with an example.</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>(defn get-time [] (System/currentTimeMillis)) (#(get-time)) ;; What's the result? </pre> </div></div> <p>What's the result of the funcall above? Clearly, right now, it is the current system time.</p> <p>So if we decided to allow to write #(true) as an alternative to (constantly true) <span class="error">&#91;which is a varargs fn&#93;</span> or #(do true) <span class="error">&#91;which is a fn of zero args&#93;</span>, then valid values of #(get-time) where both the current system time but also the function object for get-time. Functions are values, too.</p> <p>Ok, one could say that in the case of a function, #(function) is always a call, but it would make it harder to reason about what the code does for not much benefit.</p> <p>You're right - satisfying my request would require to change the average use of this feature to <tt>#((some_fn %1 %2))</tt> rather than just <tt>#(some_fn %1 %2)</tt>, if we wanted <tt>#(true)</tt> to be valid. Which indeed would be barely handy.</p> <p>Thank you.</p> Global Rank