Hi Steve, On Tue, 17 Mar 2015 22:27:38 +0000 (UTC), Steve Graham <jsgraha...@yahoo.com> wrote:
>George, > Your unicode comment made me wonder. So I wrote out some code to >the repl and it "compiled" and ran fine. Next I copied that code to >Notepad and then copied and pasted it back to the repl and it "compiled" >and ran fine. Finally I google "scheme factorial" and copied what I >found to the repl and there I had problems. I have seen a few editions of DrRacket [including 6.1.1 release] hang or crash loading a source file that has an embedded UTF BOM header. CLI racket or raco will complain about the module language being unknown or not found [can't remember the exact message], so if the file otherwise looks correct, that is a quick test for the header being present. If DrRacket does manage to load the file, you'll see a couple of weird glyphs (that can't be removed!) at the beginning. Only solution is to use another editor to remove the header. May be just my luck, but I haven't yet found an editor that will remove an existing BOM header just by saving to a different format. I use the hex/bin function of my code editor to manually remove the header bytes. The compiler - at least in conjunction with DrRacket - also doesn't much like multi-byte characters. As you have discovered, sometimes things which visually appear to be correct just won't compile. My solution, if that happens, is to load the file in a different editor, resave it as Western (ISO-8859-1) to quickly clean up as much as possible and then retype in DrRacket any lines that still produce syntax errors. I have never investigated whether the compile problem is only in conjunction with the Dr editor or is pervasive. I generally do a lot of testing/tweaking in the debug environment before moving to CLI racket. By that time I no longer have the compile problem. Unfortunately, I don't encounter either unicode problem very often, and when I do it tends to be when I'm really busy, so I just deal with it and forget about bug reports. Yeah, I know. Selfish 8-( > : First saw this in the arc programming language, which is based >on Racket. > The funny thing is that a number of other Scheme implementations do >not act thusly. I wonder what can be done to remove this from Racket. If you have an example that fails reliably, you should file a bug report. >Steve George >--- >;; Typed into repl >> (define (fact n) > (if (< n 2) > 1 > (* n (fact (- n 1))))) >> (fact 5) >120 >;; Copied from Notepad into repl> (define (fact n) > (if (< n 2) > 1 > (* n (fact (- n 1))))) >> (fact 5) >120 >;; Copied from browser into repl> (define (factorial n) > (if (= n 0) > 1 > (* n (factorial (- n 1))))) >stdin::355: if: bad syntax; > has 4 parts after keyword > in: (if (= n 0) 1 R (* n (factorial (- n 1)))) > context...: > C:\Program Files\Racket\collects\racket\private\misc.rkt:87:7 >> >From: George Neuner <gneun...@comcast.net> > To: users@racket-lang.org > Sent: Tuesday, March 17, 2015 2:21 PM > Subject: Re: [racket] Error when pasting code > >On Tue, 17 Mar 2015 19:06:50 +0000 (UTC), Steve Graham ><jsgraha...@yahoo.com> wrote: > >>When running the following code, when manually typed in, there is no >>error. However, when the same is pasted from the keyboard buffer, >>the error indicated appears. >>Comments? >>Thanks, Steve > >Likely there are hidden control characters being carried when you cut >and paste. > >I've had troubles going back and forth between DrRacket and various >unicode aware code editors - everything from syntax errors compiling >to DrRacket hanging or even crashing trying to load source files. >Inevitably it is because some hidden character(s) snuck in. > >I haven't actually experienced it with cut-n-paste, but it would not >surprise me. > >George > >____________________ > Racket Users list: > http://lists.racket-lang.org/users ____________________ Racket Users list: http://lists.racket-lang.org/users
