Re: Problem with ghc on Windows ME
It's called 'raw' because it is supposed to get the arguments through *unmodified* to the called program. No file globbing, no escape stuff, nothing.
That's exactly what I'm worried about: it seems that rawSystem is *not* passing the arguments unmodified, but tries to compensate for windows interpretation, from a unix perspective (*). As long as "raw" means "unmodified", I just have the problem of finding the appropriate documentation for the system I'm working on (after all, that's why it's in System). But if there are two interpreters fighting with each other for some unstable balance, things do not look so promising. For instance, I can't use rawSystem in ghc-6.x because there are already two different versions in circulation, and a third one planned. So, instead of simplifying my code, I'd have to triplicate it, and add version testing, which adds preprocessing.. Therefore, my suggestion would be to keep the rawSystem from ghc-6.0.1 (which doesn't seem to do any interpretation?), and to provide a system-specific escape function System.Cmd.escape :: String -> String -> String -- (System.Cmd.escape chars string) escapes occurrences of -- chars in string, according to convention on the current -- system If really necessary, there could be a convenience function somewhat like: -- try to do "the right thing" System.Cmd.rawSystem' :: String -> [String] -> IO ExitCode System.Cmd.rawSystem' path args = rawSystem $ concat (path:[' ':(escape "\\\"" a) | a <- args]) Since the implementation of rawSystem in the current binary release appears to be buggy, there is still a chance to declare the original behaviour as the intended one and have the same code work in all versions of ghc-6 that may be installed out there.. Claus (*) I can't think of any case where I could pretend that a rawSystem call on unix could be the same as one on any other system - not with all the system variations around on windows alone (mingw ghc/cygwin ghc/international variants of windows/different versions of windows/..). So what I really need is a way to specify different code, depending on what system I'm on (which seems to be the purpose of System.Info). And for that it doesn't really help me if I have to think about what a unix system call would do while writing my windows system call - it is well-meant, but it just adds complications.
C.Reinke wrote:
Therefore, my suggestion would be to keep the rawSystem from ghc-6.0.1 (which doesn't seem to do any interpretation?), and to provide a system-specific escape function
System.Cmd.escape :: String -> String -> String -- (System.Cmd.escape chars string) escapes occurrences of -- chars in string, according to convention on the current -- system
If really necessary, there could be a convenience function somewhat like:
-- try to do "the right thing" System.Cmd.rawSystem' :: String -> [String] -> IO ExitCode System.Cmd.rawSystem' path args = rawSystem $ concat (path:[' ':(escape "\\\"" a) | a <- args])
That sounds is if it makes sense, but it doesn't. At least, not on
Unix. On Unix, there are two basic options for invoking another
program:
1. Use the shell, i.e. system(). Pass a single string which will
undergo all of the various forms of processing which the shell
performs.
2. Don't use the shell. Pass a list of strings which become the
argv[i] of the called program.
If you don't want any processing, the correct solution is to bypass
the shell altogether, *not* to attempt to subject the string to an
inverse transformation in the hope that the shell will eventually end
up passing the desired argv[i] to the called program.
--
Glynn Clements
participants (2)
-
C.Reinke -
Glynn Clements