Re: Tiny fix for LispReader

2016-04-27 Thread Aaron Cummings
The clojure docs say numbers are "generally represented as per Java" with some additions (ratio, variable radix, M and N suffixes, etc.) For reference, the integer literals for Java are documented here: http://docs.oracle.com/javase/specs/jls/se8/html/jls-3.html#jls-3.10.1 There are some items fr

Re: Tiny fix for LispReader

2016-04-27 Thread Frank Pursel
I was acting with respect to what is said here: http://clojure.org/reference/reader. On Tuesday, April 26, 2016 at 10:19:59 PM UTC-7, Erik Assum wrote: > > In addition to what Andy said, this same behavior was recently implemented > in tools.reader (the Clojure impel of lisp reader) > http://

Re: Tiny fix for LispReader

2016-04-26 Thread Erik Assum
In addition to what Andy said, this same behavior was recently implemented in tools.reader (the Clojure impel of lisp reader) http://dev.clojure.org/jira/browse/TRDR-36 Erik. -- i farta > Den 27. apr. 2016 kl. 07.13 skrev Andy Fingerhut : > > Frank: > > I am pretty sure that the intent is t

Re: Tiny fix for LispReader

2016-04-26 Thread Andy Fingerhut
Frank: I am pretty sure that the intent is that strings like 077 are read as octal, and in Clojure 1.8.0 this is what happens, e.g. (read-string "077") evaluates to the integer 63 (decimal), not 77. Thus it is intentional that (read-string "08") and (read-string "09") throw exceptions. Andy On

Tiny fix for LispReader

2016-04-26 Thread Frank Pursel
I believe the following line should replace the current 1.8.0 jvm/clojure/lang/LispReader.java line 68: static Pattern intPat = Pattern.compile( "([-+]?)(?:(0)|(0*[1-9][0-9]*)|0[xX]([0-9A-Fa-f]+)|([0-7]{3})|([1-9][0-9]?)[rR]([0-9A-Za-z]+)|0[0-9]+)(N)?"); Without this ch