Modified the Phobos library to
remove Object.print method, and
to move printf back into std.c.stdio...
(based on earlier Phobos patches from
Thomas Kuehne, updated for DMD 0.118
and other D wishlists from last year)


http://www.algonet.se/~afb/d/dmd-0.118-print.patch
> Removes Object.print method, uses Object.toString() now

http://www.algonet.se/~afb/d/dmd-0.118-printf.patch
> Removes global printf, uses writef or std.c.stdio.printf

To use printf now, first "import std.c.stdio;"
Or do "import std.stdio;" and just use writef?

I did not change the unittest and debugging
code, just the ones that were always on...


Compiles without warnings, and passes Phobos unittest :

http://www.algonet.se/~afb/d/dmd.spec
http://www.algonet.se/~afb/d/dmd-0.118-11.nosrc.rpm

Tested on Linux (Fedora Core 1), not tested on Windows.

--anders

Anders F Björklund wrote:
<snip>
> To use printf now, first "import std.c.stdio;"
> Or do "import std.stdio;" and just use writef?

Do we have a consensus on the idea of making std.stdio or at least writef/writefln implicitly imported?

Just looking at this

-	    printf("Error: %.*s\n", o.toString());
+	    writefln("Error: %s", o.toString());

how about figuring out how to get these going to the right place (i.e. stderr) while we're at it?  (_Is_ this the code that main is implicitly wrapped in, or does that only apply when you start a thread other than the main one?)

Stewart.

-- 
My e-mail is valid but not my primary mailbox.  Please keep replies on the 'group where everyone may benefit.

Stewart Gordon wrote:

> Do we have a consensus on the idea of making std.stdio or at least writef/writefln implicitly imported?

Not that I've heard of. (Theoretically you can still use std.c.stdio)

I think you still need to "import std.stdio;"

> how about figuring out how to get these going to the right place (i.e. stderr) while we're at it?  (_Is_ this the code that main is implicitly wrapped in, or does that only apply when you start a thread other than the main one?)

This was not that patch, but it would be something like fwritefln...

fwritefln(stderr, "Error: %s", o.toString());

--anders

>> how about figuring out how to get these going to the right place (i.e. stderr) while we're at it? 
> 
> This was not that patch, but it would be something like fwritefln...

Here is the exact patch to reroute exceptions to standard error stream:

http://www.algonet.se/~afb/d/dmd-0.118-stderr.patch

(it assumes that the print and printf patches have been applied first)

--anders