All,
Is anyone else experiencing problems using danielz/system and/or
stuartsierra/component? I'm getting the following:
Caused by: java.lang.IllegalArgumentException: Call to clojure.core/ns did
not conform to spec:
In: [1] val: ((require [com.stuartsierra.component :as component]
[clojure.tools.namespace.track :as track] [system.reload :as reload]
[clojure.stacktrace :as st] [io.aviso.ansi :refer [bold-red bold-yellow]]))
fails at: [:args] predicate: (cat :docstring (? string?) :attr-map (? map?)
:clauses :clojure.core.specs/ns-clauses), Extra input
:clojure.spec/args (system.repl (require [com.stuartsierra.component :as
component] [clojure.tools.namespace.track :as track] [system.reload :as
reload] [clojure.stacktrace :as st] [io.aviso.ansi :refer [bold-red
bold-yellow]]))
I noticed Alex's PR for aviso but I'm not using it explicitly - it must be
coming in transiently but I don't think that is the problem:
I can see the use of (require ...) in the implicated ns-form, here:
(system.repl (require [com.stuartsierra.component :as component] ...
but I searched my project for any "(require" and didn't find any.
There must be some macro or namespace magic happening in either of those
two libraries. I've been searching them for any use of (require ...) vs
(:require ...) and only found a couple in danielz/system, here:
src/system/reload.clj
src/system/repl.clj
but I'm not exactly sure how those would cause my error. The second one
looks like it is related.
Alex - does a previously compiled library, like system, pulled from clojars
uses "(require" be subject to the spec conformance? How much of the ns-form
is resolved at compile time and how much happens at runtime?
Clearly I've not done my deep-dive homework on that form (yet) and how it
works... I know those two libraries do a bit of namespace manipulation to
deal with reloading repls and whatnot so it very well could be something
they are doing that is beyond my current understanding and pay grade.
Alan
On Friday, August 19, 2016 at 11:35:31 AM UTC-7, Alex Miller wrote:
>
> One thing to note is that the new specs for clojure.core means that there
> is a reasonable amount of (broken) code in the wild that now does not
> compile.
>
> I have filed PRs in a number of commonly used projects and while those
> were all merged, most have not yet been released into new versions yet. Any
> project that uses any of those projects via dependencies will also have
> issues. So, expect to see some breakage if using the latest alpha. In
> particular, ClojureScript itself had a bad ns declaration through version
> 1.9.93 so anything not on a fresh CLJS will likely see that.
>
> The three most common things I saw were:
>
> 1. (ns ... (require ...) (import ...) ) instead of (ns ... (:require ...)
> (:import ...) )
>
> 2. (fn a.b [] ...) - fully-qualified symbol names in fn (these are pretty
> much always the result of a sloppy macro that expands to fn)
>
> 3. (let [{ ... :or {:a 1}} {}] ... ) - :or keys that are not simple
> symbols. There were some accidental cases where this did something for
> fully-qualified keyword keys in the past, but those were not intended and
> are no longer accepted as of Clojure 1.9.0-alpha8. All of the cases I found
> were actually typos though where the keyword was used instead of the symbol.
>
>
>
>
> On Friday, August 19, 2016 at 1:15:06 PM UTC-5, Alex Miller wrote:
>>
>> Clojure 1.9.0-alpha11 is now available.
>>
>> Try it via
>>
>> - Download:
>> https://repo1.maven.org/maven2/org/clojure/clojure/1.9.0-alpha11
>> - Leiningen: [org.clojure/clojure "1.9.0-alpha11"]
>>
>> 1.9.0-alpha11 includes the following changes since 1.9.0-alpha10:
>>
>> Clojure now has specs for the following clojure.core macros: let, if-let,
>> when-let, defn, defn-, fn, and ns. Because macro specs are checked during
>> macroexpansion invalid syntax in these macros will now fail at compile time
>> whereas some errors were caught at runtime and some were not caught at all.
>>
>> - CLJ-1914 - Fixed race condition in concurrent range realization
>> - CLJ-1870 - Fixed reloading a defmulti removes metadata on the var
>> - CLJ-1744 - Clear unused locals, which can prevent memory leaks in some
>> cases
>> - CLJ-1423 - Allow vars to be invoked with infinite arglists (also,
>> faster)
>> - CLJ-1993 - Added *print-namespace-maps* dynamic var that controls
>> whether to use namespace map syntax for maps with keys from the same
>> namespace. The default is false, but standard REPL bindings set this to
>> true.
>> - CLJ-1985 - Fixed with-gen of conformer losing unform fn
>> - Fixed clojure.spec.test/check to skip spec'ed macros
>> - Fixed regression from 1.9.0-alpha8 where type hints within
>> destructuring were lost
>> - Fixed clojure.spec/merge docstring to note merge doesn't flow conformed
>> values
>> - Fixed regex ops to use gen overrides if they are used
>>
>
--
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.
