I have just been working for the last day on adding code to the Eastwood Clojure lint tool [1] for checking these and a few other things about wrong ns forms. The documentation I just wrote for what it will do is here [2].
No released version of Eastwood makes these checks yet. I will probably release the first Eastwood version that does within the next 1 to 3 weeks. There are instructions in Eastwood's documentation for using the latest unreleased version if you are interested on being on the bleeding edge [3]. If anyone has thoughts on other things to warn about in wrong ns forms, please reply to this thread, or create an Eastwood issue on GitHub. The only other change I have in mind right now is perhaps not warning if someone uses :include-macros in :require, because it appears that has snuck into the :require's of several projects that are both Clojure/Java and ClojureScript. I will first verify that this option is ignored by Clojure/Java before eliminating such warnings. Andy [1] https://github.com/jonase/eastwood [2] https://github.com/jonase/eastwood/blob/master/README.next.md#wrong-ns-form [3] https://github.com/jonase/eastwood#for-eastwood-developers On Sun, Dec 14, 2014 at 1:58 AM, Petr <petrg...@gmail.com> wrote: > > Hello. > > I stumbled on a problem that caused me fair amount of frustration and I > suspect that it could affect other users who start using tools.namespace. > > The problem is that clojure.code/ns macro is more lenient than what > follows from it’s documentation. For example ‘use’ (or require) clause in > vector is accepted e.g. “[:use ....]” (see > http://dev.clojure.org/jira/browse/TNS-21) or use clause with plain > symbol (not keyword) is also OK, e.g. “(use ....)" (as in my case > http://dev.clojure.org/jira/browse/TNS-30). And the way ns macro is > expanded section in form “(aaaa ....)” would be replaced with > “(clojure.core/aaaa ... )”. However tools.namespace parses dependencies > strictly according to doc. E.g.TNS-21 was closed as won’t fix for that > reason. > > This way there could be ns declarations that have no compilation errors > and work as one may think but are ignored by c.t.n’s parser. Hence there > could be actual dependencies that are silently ignored without any error or > warning. > > It seems that proper way to resolve this would be to rewrite ns macro to > be more strict and accept only what’s described in documentation. But I > suspect that this would lead to backward compatibility issues - some of > existing code will break. > > There is also possibility to rewrite > clojure.tools.namespace.parse/deps-from-ns-decl so it does macroexpand and > then looks for references to clojure.core/refer, clojure.core/require and > clojure.core/use. This would match now ns actually works but one might not > like this “implementation as specification”. > > So I think that the good solution for now would be to have some > validation function forns form. This function would issue a warning when > some parts of namespace declaration could be ignored. This way unexpected > behavior of clojure.tools.namespace.repl/refresh would not go unnoticed and > this will make this library more user-friendly. > > What do you think would be a good solution for this? > > > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to the Google Groups > "Clojure" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
