Using the [Take New] restart in SBCL - error-handling

When I try to re-define a package in SBCL in such a way that causes name conflicts, I get a NAME-CONFLICT error with the restarts
0: [KEEP-OLD] Keep symbols already accessible FOO (shadowing others).
1: [TAKE-NEW] Make newly exposed symbols accessible in FOO, uninterning old ones.
2: [RESOLVE-CONFLICT] Resolve conflict.
3: [RETRY] Retry SLIME REPL evaluation request.
4: [*ABORT] Return to SLIME's top level.
5: [ABORT] Abort thread (#<THREAD "new-repl-thread" RUNNING {10060E47B3}>)
I'd like to write something that would automatically invoke the TAKE-NEW restart, so that I could do something like
(force (defpackage :foo (:use :cl :bar :baz :mumble)))
The result of this should be the same as calling defpackage, followed by manually invoking the TAKE-NEW restart. The problem is,
CL-USER> (handler-case
(defpackage :foo (:use :cl :bar :baz :mumble))
(error (e) (compute-restarts e)))
(#<RESTART SWANK::RETRY {1006DC40F3}> #<RESTART ABORT {10068007E3}>
#<RESTART ABORT {10060C7F93}>)
CL-USER>
I don't seem to have access to that particular restart. As confirmed when I try to invoke it:
CL-USER> (handler-case
(defpackage :foo (:use :cl :bar :baz :mumble))
(name-conflict (e) (invoke-restart 'take-new)))
No restart TAKE-NEW is active.
[Condition of type SB-INT:SIMPLE-CONTROL-ERROR]
Restarts:
0: [RETRY] Retry SLIME REPL evaluation request.
1: [*ABORT] Return to SLIME's top level.
2: [ABORT] Abort thread (#<THREAD "new-repl-thread" RUNNING {10060E47B3}>)
Any ideas?

handler-case exits the scope of the restarts before it transfers control to a handler; you need handler-bind. The two other differences between the constructs are that handler-bind has a different syntax (demonstrated below), and handler-bind doesn't transfer control -- that is, for a handler-bind handler to return a value it needs to explicitly transfer control, e.g. with return-from or invoke-restart.
Something like the following should work:
(handler-bind ((sb-ext:name-conflict
(lambda (c)
(when (find-restart 'sb-impl::take-new c)
(invoke-restart 'sb-impl::take-new)))))
(defpackage :foo (:use :mumble)))

Related

How do I deactivate logs in a Clojure test?

Question
I would like to black hole print like behaviors within my test bodies in order to keep my log output looking clean and tidy.
(deftest some-test
(testing "something"
(logless
(is (= 22 (test-thing 14))))))
I expect test-thing to call println and other similar calls to *out* and would like those to stop polluting my test output.
Is there a recognized way to do this in general?
I found this guy (with-out-str) but it's capturing the string, not quite what I'm looking for.
Background
I'm fairly new to Clojure, coming largely from a javascript world. Having a blast so far! But there's lots left for me to learn.
In Clojure, not Clojure.script (if it matters)
Just use with-out-str and then ignore the string.
Note that this will not capture error messages or messages from Java libraries. If you want to capture and/or suppress this output, I have written 3 additional functions in the Tupelo library that you may find useful:
with-err-str
with-system-err-str
with-system-out-str
The code looks like this:
(defmacro with-system-err-str
"Evaluates exprs in a context in which JVM System/err is bound to a fresh
PrintStream. Returns the string created by any nested printing calls."
[& body]
`(let [baos# (ByteArrayOutputStream.)
ps# (PrintStream. baos#)]
(System/setErr ps#)
~#body
(System/setErr System/err)
(.close ps#)
(.toString baos#)))
If you wanted, you could make a new macro like so:
(defmacro with-printing-suppressed
"Evaluates exprs in a context in which JVM System/err and System/out captured & discarded."
[& body]
`(let [baos# (ByteArrayOutputStream.)
ps# (PrintStream. baos#)
s# (new java.io.StringWriter)]
(System/setErr ps#)
(System/setOut ps#)
(binding [*err* s#
*out* s#]
(let [result# (do ~#body)]
(.close ps#)
(System/setErr System/err)
(System/setOut System/out)
result#))))
(defn stuff []
(println "***** doing stuff *****")
42)
and then test it:
(println "calling - before")
(is= 42 (with-printing-suppressed
(stuff)))
(println "calling - after")
with result:
calling - before
calling - after
Use default logging and logback.xml for output configuration.
default clojure logging

REPL broke out of nowhere

I use IntelliJ + Cursive.
I was doing some testing on a Genetic Algorithm I'm writing, and got a NPE. It wasn't clear where it was coming from, so I started up my debugger. As it was starting up however, a RuntimeException break was activated, and it pointed here in clojure.main:
(defn main
[& args]
(try
(if args
(loop [[opt arg & more :as args] args inits []]
(if (init-dispatch opt)
(recur more (conj inits [opt arg]))
((main-dispatch opt) args inits))) ; <---- Points Here
(repl-opt nil nil))
(finally
(flush))))
The exception reads:
Exception in thread "main" java.lang.RuntimeException: Unable to resolve symbol: create in this context, compiling:(C:\Users\slomi\AppData\Local\Temp\form-init2101329611502965041.clj:4223:33
I didn't see a create symbol anywhere in the code it pointed to (which would have been weird since it's clojure.main), and I checked my entire project and that symbol isn't used anywhere. I checked the path that the exception pointed to, and found this mess:
(.deleteOnExit (java.io.File. "C:\\Users\\slomi\\AppData\\Local\\Temp\\form-init2101329611502965041.clj")) (do (set! *warn-on-reflection* nil) (do (try (clojure.core/doto (quote ai-retry.genetic_algorithm.main) clojure.core/require clojure.core/in-ns) (catch java.lang.Exception e__11808__auto__ (clojure.core/println e__11808__auto__) (clojure.core/ns ai-retry.genetic_algorithm.main))) (try (clojure.core/require (quote clojure.tools.nrepl.server)) (catch java.lang.Throwable t__11809__auto__ (clojure.core/println "Error loading" (clojure.core/str (quote clojure.tools.nrepl.server) ":") (clojure.core/or (.getMessage t__11809__auto__) (clojure.core/type t__11809__auto__))))) (try (clojure.core/require (quote complete.core)) (catch java.lang.Throwable t__11809__auto__ (clojure.core/println "Error loading" (clojure.core/str (quote complete.core) ":") (clojure.core/or (.getMessage t__11809__auto__) (clojure.core/type t__11809__auto__))))) nil) (clojure.core/let [server__11804__auto__ (clojure.tools.nrepl.server/start-server :bind "127.0.0.1" :port 0 :ack-port 58282 :handler (clojure.tools.nrepl.server/default-handler (clojure.core/with-local-vars [wrap-init-ns__11765__auto__ (clojure.core/fn [h__11766__auto__] (clojure.core/with-local-vars [init-ns-sentinel__11767__auto__ nil] (clojure.core/fn [{:as msg__11768__auto__, :keys [session]}] (clojure.core/when-not ((clojure.core/deref session) init-ns-sentinel__11767__auto__) (clojure.core/swap! session clojure.core/assoc (var clojure.core/*ns*) (try (clojure.core/require (quote ai-retry.genetic_algorithm.main)) (clojure.core/create-ns (quote ai-retry.genetic_algorithm.main)) (catch java.lang.Throwable t__11769__auto__ (clojure.core/create-ns (quote user)))) init-ns-sentinel__11767__auto__ true)) (h__11766__auto__ msg__11768__auto__))))] (clojure.core/doto wrap-init-ns__11765__auto__ (clojure.tools.nrepl.middleware/set-descriptor! {:expects #{"eval"}, :requires #{(var clojure.tools.nrepl.middleware.session/session)}}) (clojure.core/alter-var-root (clojure.core/constantly (clojure.core/deref wrap-init-ns__11765__auto__))))))) port__11805__auto__ (:port server__11804__auto__) repl-port-file__11806__auto__ (clojure.core/apply clojure.java.io/file ["C:\\Users\\slomi\\IdeaProjects\\ai-retry" ".nrepl-port"]) legacy-repl-port__11807__auto__ (if (.exists (clojure.java.io/file "C:\\Users\\slomi\\IdeaProjects\\ai-retry\\target\\base+system+user+dev")) (clojure.java.io/file "C:\\Users\\slomi\\IdeaProjects\\ai-retry\\target\\base+system+user+dev" "repl-port"))] (clojure.core/when true (clojure.core/println "nREPL server started on port" port__11805__auto__ "on host" "127.0.0.1" (clojure.core/str "- nrepl://" "127.0.0.1" ":" port__11805__auto__))) (clojure.core/spit (clojure.core/doto repl-port-file__11806__auto__ .deleteOnExit) port__11805__auto__) (clojure.core/when legacy-repl-port__11807__auto__ (clojure.core/spit (clojure.core/doto legacy-repl-port__11807__auto__ .deleteOnExit) port__11805__auto__)) (clojure.core/deref (clojure.core/promise))))
Searching through it, there isn't any create symbol anywhere. The closest match is two instances of create-ns. I'm assuming this is a file auto-generated by the REPL for whatever reason, so I wouldn't expect it to be buggy anyways.
If I delete the file, a new one is created, and the same error happens in the new file.
Now, if I open a normal non-debugging REPL, I get the same error too. This happened completely out of nowhere. Before I started the debugger for the first time, I already had a REPL running, and it started fine. The error only started after I started the debugger, now it won't go away.
I tried lein clean, deleting the temp files manually, restarting IntelliJ and my computer, and ensuring none of my code has a weird compiler error that may be manifesting in an odd way.
Can anyone think of what this might be? Why is it thinking there's an unresolved create symbol in a file that doesn't even contain create?
I found out what it was. Thanks #Alan Thompson.
When I ran lein run, I got an error that it couldn't find anything to run in my main file, which was weird since my main had a proper main method. I looked over my project.clj, and realized that since I had started the REPL last, I had changed the main location, and I typed it wrong.
I had
:main ai-retry.genetic_algorithm.main
When it should have been
:main ai-retry.genetic-algorithm.main
Changing the underscore to a dash allowed everything to start properly.
How that possibly lead to the error it gave is beyond me though. I'm going to leave this here in case anyone has to deal with this in the future.
This happened to me. Same symptom, same cause (mangled namespaces), but slightly different manifestation:
(ns my-project.my-namespace=
^-- accidental errant character
If this happens to you, review all your namespace declarations.

Clojure test: global fixtures

I've got some fixtures that boot up and close the database in my project.
Now it looks something like this:
(use-fixtures :once with-embedded-db)
while in the fixture itself I've got a dynamic variable that I use in different places:
(def ^:dynamic *db*)
(defn with-embedded-db [f]
(binding [*db* (db/connect args)]
(f)
(finally
(db/clean-up *db)))
Now, assume that db/connect and db/clean-up take some time.
PROBLEM:
When I run tests using lein test, it takes very long time, unnecessarily spending time on connecting and disconnecting to the db for every namespace.
QUESTION:
Is there a way to set up global fixtures so that when I run lein test, it calls it only once for all the test namespaces?
Thanks!
It would have been better if that feature was added to leiningen itself. At least a ticket should be opened, if not a PR.
The following solution is dirty, but you can get the idea and transform it into something more intelligent.
;; profect.clj
:profiles
{:dev {:dependencies [[robert/hooke "1.1.2"]]
:injections [(require '[robert.hooke :as hooke])
(defn run-all-test-hook [f & nss]
(doall (map (fn [a]
(when (intern a '*db*)
(intern a '*db* "1234"))) nss))
(apply f nss))
(hooke/add-hook #'clojure.test/run-tests #'run-all-test-hook)
]}}
Note: leiningen itself uses robert/hooke in its core.
And then somewhere in tests:
(ns reagenttest.cli
(:require [clojure.test :refer :all]))
(def ^:dynamic *db*) ;; should be defined in every NS where it is needed
(deftest Again
(testing "new"
(prn *db*)))
Use circleci.test, it supports :global-fixtures:
... you can define global fixtures that are only run once for the entire test run, no matter how many namespaces you run.

How to pass a value from a fixture to a test with clojure.test?

When using clojure.test's use-fixture, is there a way to pass a value from the fixture function to the test function?
A couple of good choices are dynamic binding and with-redefs. You could bind a var from the test namespace in the fixture and then use it in a test definition:
core.clj:
(ns hello.core
(:gen-class))
(defn foo [x]
(inc x))
test/hello/core.clj:
(ns hello.core-test
(:require [clojure.test :refer :all]
[hello.core :refer :all]))
(def ^:dynamic *a* 4)
(defn setup [f]
(binding [*a* 42]
(with-redefs [hello.core/foo (constantly 42)]
(f))))
(use-fixtures :once setup)
(deftest a-test
(testing "testing the number 42"
(is (= *a* (foo 75)))))
You can tell that it works by comparing calling the test directly, which does not use fixtures, to calling it through run-tests:
hello.core-test> (a-test)
FAIL in (a-test) (core_test.clj:17)
testing the number 42
expected: (= *a* (foo 75))
actual: (not (= 4 76))
nil
hello.core-test> (run-tests)
Testing hello.core-test
Ran 1 tests containing 1 assertions.
0 failures, 0 errors.
{:test 1, :pass 1, :fail 0, :error 0, :type :summary}
This approach works because fixtures close over the tests they run, though they don't get to actually make the calls to the test functions directly (usually) so it makes sense to use closures to pass information to the test code.
Perhaps not a direct answer, but if your fixture was an :each fixture anyway, or you can tolerate it being an :each fixture, you can just cop out and create a set-up function returning the relevant state and call it as the first line of your test, instead of using a fixture. This may be the best approach for some circumstances.
(defn set-up [] (get-complex-state))
(deftest blah
(let [state (set-up)]
(frobnicate)
(query state)
(tear-down state)))

Clojure concurrency : Automating SQL Queries

I have a small program that is supposed to read SQL queries/commands one by one and execute them against a database.
If a query executes successfully, the next query is executed.
If there is an error executing one query, the program should stop executing all together.
I have the code, except that the query still continues execution even when there is an exception.
(defn main
[]
(loop [queries (get-all-queries)
querycount 1]
(let [q (first queries)]
(println (format "currently processing query %s", querycount))
(cond (nil? q) (println "All Queries ran successfully.")
:else (do
(cond (= (:status (process-query q querycount)) "OK")
(recur (rest queries) (+querycount 1)))
:else (println "An error occured while running queries")))))))
(defn process-query
[query query-count]
(let [{query-body :query-body, is-query-running? :is-query-running?} query
my-agent (agent
{:error false, :query-count query-count}
:error-handler handler-fn)]
(send my-agent (fn[_]
(execute-query! db query-body)))))
(loop [is-query-running? (is-query-running?)
error? (:error #my-agent)]
(cond error? (do (println "Error")
{:status "ERROR" :error-msg (:error-msg #my-agent)})
(and (not is-query-running?) (not error?)) (do (println "Success")
{:status "OK"})
(:else (do
(Thread/sleep 2000)
(recur (is-query-running?) (:error #my-agent)))))))
(defn handler-fn
[agent exception]
(println (format "an exception occured : %s" exception))
(if (instance? java.sql.BatchUpdateException exception)
(println (.getNextException exception)))
(send agent (? [_] {:error true, :error-message exception}))
(throw exception))
The reason why I'm using an agent is that I have some queries that take 4 hours to run.
and when that happens, the database does not notify the program that the query has been completed. instead, the program is stuck. so, instead, I constantly poll to check if the query is done already.
Is this the best way to accomplish what I'm trying to do ?
Should I be using any other concurrency primitives ?
Do I even need concurrency primitives ?
I've been thinking about this for a long time now.
I think you will need to use core.async to resolve this kind of workflow
Take a look at http://clojure.com/blog/2013/06/28/clojure-core-async-channels.html
This lib will let you to check your conditions with the related asynchronous tasks involved
A few resources that may help you
http://www.infoq.com/news/2013/07/core-async
https://www.youtube.com/watch?v=AhxcGGeh5ho
The main problem seems to be: on the one hand, you write that the long queries never return, i.e. they don't even throw exceptions. On the other hand, your error detection mechanism for the agent is based on catching an exception.
I think what you need to do is not check (primarily) whether an exception was caught, but whether execute-query has actually returned a valid result when is-query-running? returns false.
Regarding the right concurrency primitive, I would suggest using a future instead of an agent. They are simpler than agents, since they can only return a single value, instead of changing their state multiple times, and their way of error handling is to simply return the exception instead of the regular return value.
You can then follow this implementation idea: in the loop, do a deref with timeout on the future. If the return value of the deref is whatever execute-query! returns regularly, return "OK"(resp. add a second expression to the future body as a clearly identifiable return value, e.g. the keyword :ok). Otherwise, if the return value of the deref is an exception, return "ERROR" with the :error-msg from the exception like you do now. Finally, if the return value is the timeout value you gave to the deref, call is-query-running?. If it's true, loop another time, if it's false, return ERROR with a special :error-msg which communicates that your query ended without either returning nor throwing an exception. (And probably call future-cancel so you don't leak threads of never-ending execute-query! calls.)