Feliz Navidad.

I have not being reading the newsgroups, because of a few projects that I have going on.  I am trying to get D into my company.  It's not going to be easy, but I am positive. :-)

So I have this module that gets some environment variables from the
system.  Here is some code:
//-------------------------------------------------------
extern (C)
{
  char*   getenv  (char *);
  int     putenv  (char *);
}
char[] GetEnv(char[] pSymbol)
{
  return std.string.toString(getenv(pSymbol));
}

int PutEnv(char[] eVar, char[] eVal)
{
  int i = putenv(eVar ~ "=" ~ eVal);
  return i;
}
char[] GetServerName()
{
  // returns the Hostname on the local computer
  char[] sn = GetEnv("COMPUTERNAME");
  return sn;
}
//-------------------------------------------------------

I copied it from the group a few years ago. So, now, with this new .177 release I get this new error:

..\jic\libs\MyOSEnv.d(32): function jic.libs.MyOSEnv.getenv (char*)
does not match parameter types (char[])
..\jic\libs\MyOSEnv.d(32): Error: cannot implicitly convert
expression (pSymbol) of type char[] to char*
..\jic\libs\MyOSEnv.d(37): function jic.libs.MyOSEnv.putenv (char*)
does not match parameter types (char[])
..\jic\libs\MyOSEnv.d(37): Error: cannot implicitly convert
expression ((eVar) ~ "=" ~ (eVal)) of type char[] to char*

How can I fix this?  Will someone be willing to point me on the right direction?

Thanks,

josé

"jicman" <jicman@mail.com> wrote in message news:em2fih$1ks5$1@digitaldaemon.com...

> How can I fix this?  Will someone be willing to point me on the right direction?

The implicit conversion from D array to pointer has been removed in 0.177. All you need to do is add ".ptr" to convert from arrays to pointers now. Though since you have no way of knowing whether or not the passed-in strings are zero-terminated, you should use toStringz whenever passing D strings to C functions.

return toString(getenv(toStringz(pSymbol)));

== Quote from Jarrett Billingsley (kb3ctd2@yahoo.com)'s article
> "jicman" <jicman@mail.com> wrote in message news:em2fih$1ks5$1@digitaldaemon.com...
> > How can I fix this?  Will someone be willing to point me on the right direction?
> The implicit conversion from D array to pointer has been removed
in 0.177.
> All you need to do is add ".ptr" to convert from arrays to
pointers now.
> Though since you have no way of knowing whether or not the passed-
in strings
> are zero-terminated, you should use toStringz whenever passing D
strings to
> C functions.
> return toString(getenv(toStringz(pSymbol)));


thanks.

== Quote from jicman (jicman@mail.com)'s article
> > are zero-terminated, you should use toStringz whenever passing D
> strings to
> > C functions.
> > return toString(getenv(toStringz(pSymbol)));
> thanks.

in non-english Windows, for my big sorrow,
imho, you need use
std.windows.charset.toMBSz() to pass char[] from D to C function,
and
std.windows.charset.fromMBSz() to pass char[] from C function to D,

in general case in _every char* in every call_ :(
how i bored with utf8 while program small windows console utilitis :
(

"invalid UTF8 sequence" on every time then i forget about fromMBSz()
it is very tiresome!

novice2 wrote:
<snip>
> in non-english Windows, for my big sorrow,

The language of your version of Windows has no effect whatsoever on what follows.

> imho, you need use
> std.windows.charset.toMBSz() to pass char[] from D to C function,
> and
> std.windows.charset.fromMBSz() to pass char[] from C function to D,

Not quite.  C functions can, as they please, work in any character encoding or not care at all about encoding.

But if you're interfacing the OS, as is the case here, then you will need to make sure the data you're transmitting is in the right encoding.

> in general case in _every char* in every call_ :(
> how i bored with utf8 while program small windows console utilitis :
> (
<snip>

Check out smjg.libs.util.console in my utility library:

http://pr.stewartsplace.org.uk/d/sutil/

Stewart.

== Quote from Stewart Gordon (smjg_1998@yahoo.com)'s article
> Not quite.  C functions can, as they please, work in any character encoding or not care at all about encoding.

You are right. But.. i was not saw C libraries, working with utf8.
I just wanted to warn this topic author, that if we have C function
prototype, using char* params, it is very very low probability,
that this char* means utf8 string. In Windows at least.
And i never saw utf8 functions in Windows API.

He will have "invalid utf8 sequence" if will not convert to utf8
 strings, that returned from C function.

And i wanted to complain, that in D it is bored to interface with Windows :)

> Check out smjg.libs.util.console in my utility library: http://pr.stewartsplace.org.uk/d/sutil/

Thank you!