Currently gens are not lazy, so entire tree is generated. This is because a) test.check isn’t lazy either and b) we want errors when asking for gen, not using it. But it is rough for recursive specs as you see.
For now you can limit the recursion depth to control the branching: (binding [s/*recursion-limit* 3] (time (s/gen ::a))) > On May 23, 2016, at 6:34 PM, Ambrose Bonnaire-Sergeant > <abonnaireserge...@gmail.com> wrote: > > I'm observing mutually recursive regex ops taking a long time to generate > test.check generators. Is this expected? > > (s/def ::a (s/nilable (s/cat :a ::a > :b ::b > :c ::c))) > (s/def ::b (s/nilable (s/cat :a ::a > :b ::b > :c ::c))) > (s/def ::c (s/nilable (s/cat :a ::a > :b ::b > :c ::c))) > > (time (s/gen ::a)) > ;"Elapsed time: 3993.431793 msecs" > nil > > Thanks, > Ambrose > > > On Mon, May 23, 2016 at 6:12 PM, Nicola Mometto <brobro...@gmail.com> wrote: > Looks like it is, in the meanwhile this patch should fix it: > http://sprunge.us/XTiA > > > > On 23 May 2016, at 23:03, Nicola Mometto <brobro...@gmail.com> wrote: > > > > Possibly CLJ-1544 related? > > > >> On 23 May 2016, at 22:59, Ambrose Bonnaire-Sergeant > >> <abonnaireserge...@gmail.com> wrote: > >> > >> I'm having trouble calling `s/gen`, might be some sort of AOT compilation > >> error. > >> (I'm guessing this line has something to do with it). > >> > >> I'm using the latest master-SNAPSHOT. > >> > >> (ns gen-load.core > >> (:require [clojure.spec :as s])) > >> > >> (s/gen integer?) > >> ;CompilerException java.lang.NoClassDefFoundError: > >> clojure/spec/gen$gen_for_pred, compiling:(gen_load/core.clj:4:1) > >> > >> Thanks, > >> Ambrose > >> > >> On Mon, May 23, 2016 at 5:33 PM, Ambrose Bonnaire-Sergeant > >> <abonnaireserge...@gmail.com> wrote: > >> I see, thanks. > >> > >> On Mon, May 23, 2016 at 5:29 PM, Rich Hickey <richhic...@gmail.com> wrote: > >> fdef will not add doc metadata (see rationale re: not putting more stuff > >> in the namespaces/vars), but specs will be present when you call ‘doc’. > >> That doc enhancement was in a push later in the afternoon. > >> > >> https://github.com/clojure/clojure/commit/4c8efbc42efa22ec1d08a1e9fa5dd25db99766a9 > >> > >>> On May 23, 2016, at 5:26 PM, Ambrose Bonnaire-Sergeant > >>> <abonnaireserge...@gmail.com> wrote: > >>> > >>> Thanks Rich+team, this is awesome. > >>> > >>> Instrumented vars via `fdef` do not seem to add :doc metadata yet > >>> (which is advertised in the docstring for `fdef`). > >>> > >>> Am I missing something? > >>> > >>> Thanks, > >>> Ambrose > >>> > >>> On Mon, May 23, 2016 at 5:20 PM, Andrey Grin <andrey.g...@gmail.com> > >>> wrote: > >>> Thank you. Yes, it seems that for recursion I can use the same approach. > >>> As for inheritance currently I just include all parent attributes in > >>> every child schema. As for spec definition multi-spec, as I understand > >>> it, requires "defmethod" for every class which is probably a problem for > >>> my case since all inheritance hierarchies are also determined at > >>> run-time. I will need to find some approach that is both dynamic and > >>> supports precise error messages in case validation fails. > >>> > >>> > >>> Am Montag, 23. Mai 2016 22:38:24 UTC+2 schrieb Alex Miller: > >>> I think you could apply the same idea - use a regular translation between > >>> Java class and registered name and then always refer to the registered > >>> name. > >>> > >>> There are probably multiple approaches to modeling inheritance, so hard > >>> to recommend something without knowing more. In general, attributes > >>> coming from super classes can be modeled using the parent attribute name. > >>> And I would recommend looking into multi-spec for doing validation of a > >>> number of "types" that could be flowing into the same location by > >>> leveraging a multimethod that conditionally provides the spec to use. > >>> > >>> On Monday, May 23, 2016 at 3:09:46 PM UTC-5, Andrey Grin wrote: > >>> Thanks, Alex. I've taken random example from plumatic.schema wiki to > >>> quickly illustrate the question. In my actual use case I dynamically > >>> generate schemas based on hundreds of model Java classes (beans) with > >>> mutual references so I need recursion (and also some way to model > >>> inheritance, that was is achieved with "conditional" in plumatic). Will > >>> try to implement the same with spec. > >>> > >>> > >>> -- > >>> 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. > >> > >> -- > >> 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. > > > > -- > 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. -- 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.
