Afficher les réponses par date
On 5-Jul-09, at 5:10 PM, Pierre Bernatchez wrote:
My goal is to maximize the portability of my scheme source code.
Here is what I think I should be doing:
Confine most of my code to r5rs scheme, and, with srfi-0 support, use SRFI's when I need them. Similarly use the full gsi scheme language on occasion when I need that.
Presuming gsi is a superset of r5rs, I could just write everything in that language and use self discipline, but It would be nice to get support from the system inhibiting me from stepping outside the bounds of r5rs in a given module.
Here's how you can do that. Write your code like this (the first two lines are what is important):
(##namespace ("not-r5rs#")) ;; prefix to use on all global names (##include "~~lib/r5rs#.scm") ;; except the R5RS ones
(define (foo) (display "hello") (newline))
(define (bar) (println "world"))
(foo) (bar)
If you call procedures that are not in r5rs (like the Gambit specific procedure "println" in the above code) then you will get an error message. For example this is the output when running the above code (which is stored in the file "limited-to-r5rs.scm"):
% gsi limited-to-r5rs.scm hello *** ERROR IN not-r5rs#bar, "limited-to-r5rs.scm"@9.4 -- Unbound variable: not-r5rs#println
Use (##include "~~lib/r4rs#.scm") if you want to limit to R4RS or (##include "~~lib/gambit#.scm") to get all the Gambit goodness.
Note that this does not catch all portability problems, because some "standard" procedures may have implementation dependent behavior. So for example in Gambit (sqrt 9) will evaluate to 3 (the exact integer 3). Some implementations of Scheme will evaluate this to 3.0 (an inexact real). Although there are many such discrepancies, they seldom cause problems (but when there is a problem it can be very hard to debug!).
Marc