Re: The performance of plain map vs. defrecord

In fact they are symbols, but print as keywords: user=> (symbol ":a") :a user=> (symbol? (symbol (str :a))) true 2017-10-09 16:26 GMT+08:00 Peter Hull : > Slightly off-topic, but why doesn't using the 'incorrect' alphabet-macro > give an error? If I try and define a

Re: The performance of plain map vs. defrecord

Slightly off-topic, but why doesn't using the 'incorrect' alphabet-macro give an error? If I try and define a Record with keyword keys it tells me very clearly not to: user=> (defrecord Wrong [:a]) CompilerException java.lang.AssertionError: defrecord and deftype fields must be symbols,

Re: The performance of plain map vs. defrecord

Ooh, thanks for your tips. On Sun, Oct 8, 2017 at 7:19 PM, dennis zhuang wrote: > First, the alphabet-macro is wrong, it should use name instead of str in > comp: > > user=> (macroexpand-1 '(alphabet-macro)) > (do (clojure.core/defrecord Alphabet [:a :b :c :d :e :f :g :h

Re: The performance of plain map vs. defrecord

First, the alphabet-macro is wrong, it should use name instead of str in comp: user=> (macroexpand-1 '(alphabet-macro)) (do (clojure.core/defrecord Alphabet [:a :b :c :d :e :f :g :h :i :j :k :l :m :n :o :p :q :r :s :t :u :w :v :x :y :z]) (def dummy-record (Alphabet. 0 1 2 3 4 5 6 7 8 9 10 11 12

The performance of plain map vs. defrecord

As I read JoyOfClojure, it says one of reasons that prefer record over plain map is speed, out of curiosity I test this using https://github.com/hugoduncan/criterium, but the result is just contrary to my expectation. record is slower than map. Is there any optimizations made to clojure 1.8.0