Remove these in favor of the two traits themselves and the wrapper
function std::from_str::from_str.

Add the function std::num::from_str_radix in the corresponding role for
the FromStrRadix trait.

With `from_str` in the prelude, and `from_str_radix` in `std::num`, the feature is unfied under the type annotation of these functions instead of using the modules-named-as-types (std::uint and others):

What was before:

    let n = std::uint::from_str("1");
    let m = std::i32::from_str_radix("10", 16);

is now:

    let n = from_str::<uint>("1");
    let m = std::num::from_str_radix::<i32>("10", 16);

Remove these in favor of the two traits themselves and the wrapper
function std::from_str::from_str.

Add the function std::num::from_str_radix in the corresponding role for
the FromStrRadix trait.

This follows from the discussion in #9012.

* All macros are now defined in terms of `format_args!` allowing for removal of a good bit of code in the syntax extension
* The syntax extension is now in a more aptly-named file, `format.rs`
* Documentation was added for the `format!`-related macros.

The same fix as before is still relevant, I just forgot to update the
expand_stmt macro expansion site. The tests for format!() suffice as tests for
this change.

This renames the syntax-extension file to format from ifmt, and it also reduces
the amount of complexity inside by defining all other macros in terms of
format_args!

r? @metajack - This solves the problem you were having earlier with things like ```extern mod geom = "rust-geom";``` (or something like that).

The cases they test are well covered in the `std::iter` tests.

For some reason, I thought it wasn't necessary to write the package_id
attribute (which rustc's filesearch checks when searching for a package)
if the package ID had a single component (like "foo") as opposed to multiple
components (like "foo/bar/quux"). This meant that
`extern mod quux = "an-awesome-library";` didn't work, even if an-awesome-library
existed in the RUST_PATH.

Fixed it.

Replaced by `std::iter::range_step`

Use the iterator version instead of the old uint::/int::range_step
functions.

My focus was on getting these to be correct in all cases by handling overflow properly. I'll clean them up and work on the performance later.

SharedPort implementation was missing in std::comm. Since this module
also wraps SharedChan, it makes sense to have SharedPort defined there
as well.

Allows `std::rt::io::timer::sleep(1000)` rather than `std::rt::io::timer::Timer::new().unwrap().sleep(1000)`.

SharedPort implementation was missing in std::comm. Since this module
also wraps SharedChan, it makes sense to have SharedPort defined there
as well.

Closes #5794

Somehow this was missed!

cc #4819

I'm getting the three `make check` failures mentioned in issue #9127, which I also get building master.

Work a bit towards #9157 "Remove Either". These instances don't need to use Either and are better expressed in other ways (removing allocations and simplifying types).

An expression such as `bottom == not_bottom` or `not_bottom == bottom`
already compiled, but this fixes the case where both sides are `bottom`.

It was broken on win32 because of header inconsistency.

Closes #9123

Remove references to local_data::Key and only mention the macro for how to
construct new keys into local data.

This is a workaround for #9155. Currently, any uses of BufferedStream
outside of libstd ICE.