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-1168 Clojure <p>First discovered by Tim McCormack with Clojure 1.5.0-beta13 and -RC15:</p> <p>$ java -Dclojure.read.eval=unknown -jar ~/.m2/repository/org/clojure/clojure/1.5.0-RC16/clojure-1.5.0-RC16.jar -e "(+ 1 2)"<br/> Exception in thread "main" java.lang.RuntimeException: Reading disallowed - <b>read-eval</b> bound to :unknown</p> <p>This would prevent a Leiningen user from putting the following line in their project.clj file in order to look for read or read-string calls that don't explicitly bind <b>read-eval</b> to true or false:</p> <p> :jvm-opts <span class="error">&#91;&quot;-Dclojure.read.eval=unknown&quot;&#93;</span></p> CLJ-1168

Setting clojure.read.eval to unknown on JVM cmd line causes --eval option to fail

Enhancement Minor Closed Completed Unassigned Andy Fingerhut Tue, 19 Feb 2013 19:23:08 -0600 Fri, 22 Feb 2013 14:12:22 -0600 Fri, 22 Feb 2013 14:12:22 -0600 Release 1.5 0 3 <p>Patch clj-1168-patch-v1.txt dated Feb 19 2013 has been tested manually with these results:</p> <p>% java -cp clojure.jar clojure.main</p> <p>(ran some simple tests to ensure repl still works, and obeys -Dclojure.read.eval=unknown setting)</p> <p>% java -Dclojure.read.eval=unknown -jar clojure.jar -e '(println (+ 1 2))'<br/> 3</p> <p>The patch works for me, but I think it would be safer for the patch to treat &#42;read-eval* :unknown as false for the purpose of -e. It's unlikely that anyone wants to -e '#=(boom)'. Also, there's no need for the let to capture the cur-read-eval. I suggest something like this for the patch (updated):</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java">(defn- read-never-unknown [read-fn &amp; args] (binding [*read-eval* (<span class="code-keyword">true</span>? *read-eval*)] (apply read-fn args)))</pre> </div></div> <p>I have asked on the Leiningen email list whether treating <b>read-eval</b> :unknown as false for the purpose of -e would break Leiningen functionality, and will report back here when I find out.</p> <p>I don't have an answer from Leiningen folks yet, but I do have another thought related to Steve's comment:</p> <p>The current behavior is to treat &#42;read-eval&#42; as true for the purposes of reading lines in the REPL, if -Dclojure.read.eval=unknown was specified on the command line. Why should expressions given after -e be any more restricted? Whoever is invoking the JVM has complete control.</p> <p>If they really want &#42;read-eval&#42; to be false when reading the expressions after -e, just use -Dclojure.read.eval=false instead of =unknown.</p> <p>You're right, the user has control. I was just thinking that the user might not appreciate the details and it would be safer to default to the false behavior. However, I failed to consider the REPL treatment &#8211; that's a good point.</p> <p>In any case, the logic for the binding in the patch could be simplified. I think this would work for the intended result:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java">(defn- read-never-unknown [read-fn &amp; args] (binding [*read-eval* (and *read-eval* <span class="code-keyword">true</span>)] (apply read-fn args)))</pre> </div></div> <p>Sorry, if my comments have descended into bike-shedding.</p> <p>I implemented "with-read-known" patch after input from Rich. Helpers that establish bindings should take a closure or fn body, rather than carrying around explicit fns and arg lists as the original 19 Feb patch does.</p> <p>Marking screened as of the Feb 22 with-read-known patch.</p> <p>Steve: More importantly, any time you call (eval (read ...)), there's no reason for &#42;read-eval&#42; to be anything other than true.</p> <p>Newer patch works. My verification procedure:</p> <ul> <li>Patched master (should probably use an RC instead)</li> <li>Compiled and installed (mvn install) with version = 1.5.0-clj1168 in pom.xml</li> <li>Set up fresh Leiningen project including these options: <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java">:dependencies [[org.clojure/clojure <span class="code-quote">"1.5.0-clj1168"</span>]] :jvm-opts [<span class="code-quote">"-Dclojure.read.eval=unknown"</span>]</pre> </div></div></li> <li>Verify: `lein repl` succeeds <ul> <li>Verify: <tt>=&gt; (read-string "foo")</tt> fails with "Reading disallowed - &#42;read-eval&#42; bound to :unknown"</li> <li>Verify: <tt>=&gt; #=(+ 1 2)</tt> prints <tt>3</tt> (read is fully available to REPL)</li> </ul> </li> <li>Write -main fn in project as <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java">(defn -main [&amp; args] (println *read-eval*) (println (read-string (first args))))</pre> </div></div></li> <li>Verify: `lein run foo` prints <tt>:unknown</tt> and then fails with "Reading disallowed" error.</li> <li>Bind &#42;read-eval&#42; to false in -main</li> <li>Verify: `lein run foo` prints <tt>foo</tt>.</li> </ul> Approval Ok Global Rank Patch Code