From: Groepaz (groepaz_at_gmx.net)
Date: 2003-09-30 22:05:16
On Tuesday 30 September 2003 20:54, Ullrich von Bassewitz wrote:
> The character codes are translated into PETSCII, which explains the effect
> you've noticed.
ooooh. mmmh :O)
can you (briefly :=P) tell what the standard says on that then? (ie, what are
the properties of escaped numerics?) i thought that escaped numeric
characters shouldnt be translated like this.... mmmmh. is this even defined
at all? or is it implementation defined?
i gotta modify the test then so it will only check if those can be translated
at all....and maybe if they lead to unique binary values.... however, i am
still not quite sure if this translation should take place... what about
binary data that's embedded into the c-program, translation would be fatal
here? (maybe such thing is simply considered non portable by the standard?)
if what you suggest is ok (ie, the translation should take place), character
constants would have the following properties (for which i could check :))
a) each character has exactly one corrosponding escaped numeric. since both
will be translated, a check like ('A' == '\101') must work on all targets
(ie, its portable)
b) the same as a), but for escaped characters ('\n' etc)
c) a binary value does not need to corrospond to a character in the target
characterset. comparison and/or assignments of binary values with/to
characters/strings is dependend on the target characterset and thus not
portable. (like the tests i have pasted :=P)
mmmh, maybe you can tell some more? (and if its really like this :=))
gpz
----------------------------------------------------------------------
To unsubscribe from the list send mail to majordomo_at_musoftware.de with
the string "unsubscribe cc65" in the body(!) of the mail.
This archive was generated by hypermail 2.1.3 : 2003-09-30 22:10:35 CEST