auto merge of #15325 : pcwalton/rust/trait-impl-bound-mismatch, r=pnkfelix
with the corresponding trait parameter bounds.
This is a version of the patch in PR #12611 by Florian Hahn, modified to
address Niko's feedback.
It does not address the issue of duplicate type parameter bounds, nor
does it address the issue of implementation-defined methods that contain
*fewer* bounds than the trait, because Niko's review indicates that this
should not be necessary (and indeed I believe it is not). A test has
been added to ensure that this works.
This will be rejected because the implementation requires *more* bounds
than the trait. It can be fixed by either adding the missing bound to
the trait:
trait Foo {
fn bar<T:Baz + Quux>();
// ^~~~
}
impl Foo for Boo {
fn bar<T:Baz + Quux>() { ... } // OK
}
Or by removing the bound from the impl:
trait Foo {
fn bar<T:Baz>();
}
impl Foo for Boo {
fn bar<T:Baz>() { ... } // OK
// ^ remove Quux
}
This patch imports the relevant tests from #2687, as well as the test
case in #5886, which is fixed as well by this patch.
auto merge of #15323 : alexcrichton/rust/no-travis-wait, r=huonw
It's looking like we're still timing out all over the place with travis_wait
because the entire `make -j4 rustc-stage1` command is taking too long. Instead,
achieve roughly the same idea by just having `-Z time-passes` printing
information. We shouldn't have a pass that takes longer than 10 minutes in
isolation.
Jakub Wieczorek [Wed, 25 Jun 2014 17:07:37 +0000 (19:07 +0200)]
Change exhaustiveness analysis to permit multiple constructors per pattern
Slice patterns are different from the rest in that a single slice pattern
does not have a distinct constructor if it contains a variable-length subslice
pattern. For example, the pattern [a, b, ..tail] can match a slice of length 2, 3, 4
and so on.
As a result, the decision tree for exhaustiveness and redundancy analysis should
explore each of those constructors separately to determine if the pattern could be useful
when specialized for any of them.
auto merge of #15286 : alexcrichton/rust/channel-stability, r=cmr
* channel() - #[unstable]. This will likely remain forever
* sync_channel(n: int) - #[unstable with comment]. Concerns have ben raised
about the usage of the term "synchronous channel" because that generally only
applies to the case where n == 0. If n > 0 then these channels are often
referred to as buffered channels.
* Sender::send(), SyncSender::send(), Receiver::recv() - #[experimental]. These
functions directly violate the general guideline of not providing a failing
and non-failing variant. These functions were explicitly selected for being
excused from this guideline, but recent discussions have cast doubt on that
decision. These functions are #[experimental] for now until a decision is made
as they are candidates for removal.
* Sender::send_opt(), SyncSender::send_opt(), Receiver::recv_opt() - #[unstable
with a comment]. If the above no-`_opt` functions are removed, these functions
will be renamed to the non-`_opt` variants.
* SyncSender::try_send(), Receiver::try_recv() - #[unstable with a comment].
These return types of these functions to not follow general conventions. They
are consistent with the rest of the api, but not with the rest of the
libraries. Until their return types are nailed down, these functions are
#[unstable].
* Receiver::iter() - #[unstable]. This will likely remain forever.
* std::com::select - #[experimental]. The functionality is likely to remain in
some form forever, but it is highly unlikely to remain in its current form. It
is unknown how much breakage this will cause if and when the api is
redesigned, so the entire module and its components are all experimental.
* DuplexStream - #[deprecated]. This type is not composable with other channels
in terms of selection or other expected locations. It can also not be used
with ChanWriter and ChanReader, for example. Due to it being only lightly
used, and easily replaced with two channels, this type is being deprecated and
slated for removal.
* Clone for {,Sync}Sender - #[unstable]. This will likely remain forever.
Alex Crichton [Mon, 30 Jun 2014 22:10:57 +0000 (15:10 -0700)]
Apply stability attributes to std::comm
* channel() - #[unstable]. This will likely remain forever
* sync_channel(n: int) - #[unstable with comment]. Concerns have ben raised
about the usage of the term "synchronous channel" because that generally only
applies to the case where n == 0. If n > 0 then these channels are often
referred to as buffered channels.
* Sender::send(), SyncSender::send(), Receiver::recv() - #[experimental]. These
functions directly violate the general guideline of not providing a failing
and non-failing variant. These functions were explicitly selected for being
excused from this guideline, but recent discussions have cast doubt on that
decision. These functions are #[experimental] for now until a decision is made
as they are candidates for removal.
* Sender::send_opt(), SyncSender::send_opt(), Receiver::recv_opt() - #[unstable
with a comment]. If the above no-`_opt` functions are removed, these functions
will be renamed to the non-`_opt` variants.
* SyncSender::try_send(), Receiver::try_recv() - #[unstable with a comment].
These return types of these functions to not follow general conventions. They
are consistent with the rest of the api, but not with the rest of the
libraries. Until their return types are nailed down, these functions are
#[unstable].
* Receiver::iter() - #[unstable]. This will likely remain forever.
* std::com::select - #[experimental]. The functionality is likely to remain in
some form forever, but it is highly unlikely to remain in its current form. It
is unknown how much breakage this will cause if and when the api is
redesigned, so the entire module and its components are all experimental.
* DuplexStream - #[deprecated]. This type is not composable with other channels
in terms of selection or other expected locations. It can also not be used
with ChanWriter and ChanReader, for example. Due to it being only lightly
used, and easily replaced with two channels, this type is being deprecated and
slated for removal.
* Clone for {,Sync}Sender - #[unstable]. This will likely remain forever.
auto merge of #15265 : omasanori/rust/udp, r=alexcrichton
POSIX has recvfrom(2) and sendto(2), but their name seem not to be suitable with Rust. We already renamed getpeername(2) and getsockname(2), so I think it makes sense.
Alternatively, `receive_from` would be fine. However, we have `.recv()` so I chose `recv_from`.
What do you think? If this makes sense, should I provide `recvfrom` and `sendto` deprecated methods just calling new methods for compatibility?
auto merge of #15257 : erickt/rust/hashmap, r=alexcrichton
While `HashMap::new` and `HashMap::with_capacity` were being initialized with a random `SipHasher`, it turns out that `HashMap::from_iter` was just using the default instance of `SipHasher`, which wasn't randomized. This closes that bug, and also inlines some important methods.
auto merge of #15085 : brson/rust/stridx, r=alexcrichton
Being able to index into the bytes of a string encourages
poor UTF-8 hygiene. To get a view of `&[u8]` from either
a `String` or `&str` slice, use the `as_bytes()` method.
Closes #12710.
[breaking-change]
If the diffstat is any indication this shouldn't have a huge impact but it will have some. Most changes in the `str` and `path` module. A lot of the existing usages were in tests where ascii is expected. There are a number of other legit uses where the characters are known to be ascii.
Patrick Walton [Wed, 2 Jul 2014 04:59:16 +0000 (21:59 -0700)]
librustc: Properly compare implementation method type parameter bounds
with the corresponding trait parameter bounds.
This is a version of the patch in PR #12611 by Florian Hahn, modified to
address Niko's feedback.
It does not address the issue of duplicate type parameter bounds, nor
does it address the issue of implementation-defined methods that contain
*fewer* bounds than the trait, because Niko's review indicates that this
should not be necessary (and indeed I believe it is not). A test has
been added to ensure that this works.
This will be rejected because the implementation requires *more* bounds
than the trait. It can be fixed by either adding the missing bound to
the trait:
trait Foo {
fn bar<T:Baz + Quux>();
// ^~~~
}
impl Foo for Boo {
fn bar<T:Baz + Quux>() { ... } // OK
}
Or by removing the bound from the impl:
trait Foo {
fn bar<T:Baz>();
}
impl Foo for Boo {
fn bar<T:Baz>() { ... } // OK
// ^ remove Quux
}
This patch imports the relevant tests from #2687, as well as the test
case in #5886, which is fixed as well by this patch.
Alex Crichton [Wed, 2 Jul 2014 03:19:26 +0000 (20:19 -0700)]
travis: Move from travis_wait to time-passes
It's looking like we're still timing out all over the place with travis_wait
because the entire `make -j4 rustc-stage1` command is taking too long. Instead,
achieve roughly the same idea by just having `-Z time-passes` printing
information. We shouldn't have a pass that takes longer than 10 minutes in
isolation.
Brian Anderson [Fri, 20 Jun 2014 01:22:33 +0000 (18:22 -0700)]
rustc: Remove `&str` indexing from the language.
Being able to index into the bytes of a string encourages
poor UTF-8 hygiene. To get a view of `&[u8]` from either
a `String` or `&str` slice, use the `as_bytes()` method.
auto merge of #15229 : steveklabnik/rust/if, r=cmr
Whew. So much here! Feedback very welcome.
This is the first part where we actually start learning things. I'd like to think I struck a good balance of explaining enough details, without getting too bogged down, and without being confusing... but of course I'd think that. :wink:
As I mention in the commit comment, We probably want to move the guessing game to the rust-lang org, rather than just having it on my GitHub. Or, I could put the code inline. I think it'd be neat to have it as a project, so people can pull it down with Cargo. Until we make that decision, I'll just leave this here.
Steve Klabnik [Fri, 27 Jun 2014 22:51:04 +0000 (18:51 -0400)]
Guide: variable bindings.
Whew! Who knew there was so much to say about variables.
We probably want to move the guessing game to the rust-lang org, rather than
just having it on my GitHub. Or, I could put the code inline. I think it'd be
neat to have it as a project, so people can pull it down with Cargo. Until we
make that decision, I'll just leave this here.
OGINO Masanori [Mon, 30 Jun 2014 03:08:27 +0000 (12:08 +0900)]
Rename recvfrom -> recv_from, sendto -> send_to.
POSIX has recvfrom(2) and sendto(2), but their name seem not to be
suitable with Rust. We already renamed getpeername(2) and
getsockname(2), so I think it makes sense.
Alternatively, `receive_from` would be fine. However, we have `.recv()`
so I chose `recv_from`.
auto merge of #15292 : alxgnon/rust/vim_syntax_cleanup, r=cmr
- Fix a couple mistakes:
- `Ordering` is an enum, not a trait.
- `Container` is now named `Collection`.
- Add missing `CheckedDiv`.
- Remove obsolete `OwnedVector`.
- Reorganize some lines to match [prelude's arrangement](http://doc.rust-lang.org/std/prelude/#reexports), making mistakes easier to spot in the future.
Steve Klabnik [Fri, 27 Jun 2014 21:31:28 +0000 (17:31 -0400)]
Re-arrange TOC.
I'm going to move testing to be right AFTER the guessing game. I wanted it
to be borderline TDD, but I think that, since the first example is just one
file, it might be a bit overkill.
I'm doing this in its own commit to hopefully avoid merge conflicts.
auto merge of #15289 : aturon/rust/libstd-stability, r=alexcrichton
Earlier commits have established a baseline of `experimental` stability
for all crates under the facade (so their contents are considered
experimental within libstd). Since `experimental` is `allow` by
default, we should use the same baseline stability for libstd itself.
This commit adds `experimental` tags to all of the modules defined in
`std`, and `unstable` to `std` itself.
auto merge of #15288 : brson/rust/cleanslice, r=alexcrichton
This does two things:
* Reorganizes the declaration order to be more readable, less random.
* Removes the `slice::traits` module, a public module that does nothing but declare impls of `cmp` traits.
auto merge of #15263 : aturon/rust/rustdoc-stability-index, r=alexcrichton
This commit hooks rustdoc into the stability index infrastructure in two
ways:
1. It looks up stability levels via the index, rather than by manual
attributes.
2. It adds stability level information throughout rustdoc output, rather
than just at the top header. In particular, a stability color (with
mouseover text) appears next to essentially every item that appears
in rustdoc's HTML output.
Along the way, the stability index code has been lightly refactored.
Earlier commits have established a baseline of `experimental` stability
for all crates under the facade (so their contents are considered
experimental within libstd). Since `experimental` is `allow` by
default, we should use the same baseline stability for libstd itself.
This commit adds `experimental` tags to all of the modules defined in
`std`, and `unstable` to `std` itself.
Aaron Turon [Thu, 26 Jun 2014 18:37:39 +0000 (11:37 -0700)]
rustdoc: incorporate stability index throughout
This commit hooks rustdoc into the stability index infrastructure in two
ways:
1. It looks up stability levels via the index, rather than by manual
attributes.
2. It adds stability level information throughout rustdoc output, rather
than just at the top header. In particular, a stability color (with
mouseover text) appears next to essentially every item that appears
in rustdoc's HTML output.
Along the way, the stability index code has been lightly refactored.
auto merge of #15271 : jasonthompson/rust/docs/str, r=huonw
I'm working on adding examples to the API documentation. Should future pull requests include examples for more than one function? Or is this about the right size for a pull request?
bors [Mon, 30 Jun 2014 23:06:35 +0000 (23:06 +0000)]
auto merge of #15247 : smenardpw/rust/patch-1, r=alexcrichton
./hello_world is not recognized on Windows.
We can type either hello_world or hello_world.exe to run the executable. I chose "hello_world.exe", which seems more conventional on Windows.
* Tried to make the code more idiomatic
* Renamed the `wr` field of the `Encoder` and `PrettyEncoder` structs to `writer`
* Replaced some `from_utf8` by `from_utf8_owned` to avoid unnecessary allocations
* Removed unnecessary `unsafe` code
* Added `encode` and `decode` shortcut functions
* Implemented `FromStr` for `Json`
* Implemented `ToJson` for tuples of arity up to 12
* Fixed some details in the documentation
### Questions
* ~~The `encode` shortcut function does the same as the `Encoder::str_encode` function. It seems wrong to me that two functions do exactly the same. Should we deprecate `Encoder::str_encode`?~~
* ~~Do we really want the ToJson trait for tuples? At the moment we have it for (), (A, B), (A, B, C). I would like to remove them.~~
* ~~We are using `String` as key in the `TreeMap` representing a `Json` object. It would be better to use `&str`, but this would require to annotate lots of lifetimes. Is there any easy solution for this?~~
* There is a lot of duplicated code (`PrettyEncoder` copies about 50 lines from `Encoder`). In an OO language this could be solved very elegantly by using inheritance and overriding. What can we do here to reduce the amount of boilerplate?
Adolfo Ochagavía [Sat, 28 Jun 2014 12:34:58 +0000 (14:34 +0200)]
JSON cleanup
* Tried to make the code more idiomatic
* Renamed the `wr` field of the `Encoder` and `PrettyEncoder` structs to `writer`
* Replaced some `from_utf8` by `from_utf8_owned` to avoid unnecessary allocations
* Removed unnecessary `unsafe` code
bors [Mon, 30 Jun 2014 19:26:35 +0000 (19:26 +0000)]
auto merge of #14613 : schmee/rust/utf16-iterator, r=huonw
Closes #14358.
~~The tests are not yet moved to `utf16_iter`, so this probably won't compile. I'm submitting this PR anyway so it can be reviewed and since it was mentioned in #14611.~~ EDIT: Tests now use `utf16_iter`.
This deprecates `.to_utf16`. `x.to_utf16()` should be replaced by either `x.utf16_iter().collect::<Vec<u16>>()` (the type annotation may be optional), or just `x.utf16_iter()` directly, if it can be used in an iterator context.
John Schmidt [Sat, 31 May 2014 11:02:29 +0000 (13:02 +0200)]
Add `utf16_units`
This deprecates `.to_utf16`. `x.to_utf16()` should be replaced by either
`x.utf16_units().collect::<Vec<u16>>()` (the type annotation may be optional), or
just `x.utf16_units()` directly, if it can be used in an iterator context.
bors [Mon, 30 Jun 2014 14:06:31 +0000 (14:06 +0000)]
auto merge of #15267 : ruud-v-a/rust/patch-1, r=huonw
[This commit](https://github.com/rust-lang/rust/commit/25fe2cadb10db1a54cefbd1520708d4397874bc3#diff-04c6e90faac2675aa89e2176d2eec7d8R57) forgot to remove one line.
Erick Tryzelaar [Mon, 23 Jun 2014 23:15:40 +0000 (19:15 -0400)]
std: make sure HashMap from_iter uses random initialization by default
It turns out that HashMap's from_iter implementation was being
initialized without the sip keys being randomized. This adds
a custom default hasher that should avoid this potential vulnerability.
bors [Mon, 30 Jun 2014 07:31:29 +0000 (07:31 +0000)]
auto merge of #15237 : zzmp/rust/feat/markdown-in-crate-documentation, r=huonw
This makes the `in-header`, `markdown-before-content`, and `markdown-after-content` options available to `rustdoc` when generating documentation for any crate.
Before, these options were only available when creating documentation *from* markdown. Now, they are available when generating documentation from source.
This also updates the `rustdoc -h` output to reflect these changes. It does not update the `man rustdoc` page, nor does it update the documentation in [the `rustdoc` manual](http://doc.rust-lang.org/rustdoc.html).
bors [Mon, 30 Jun 2014 05:36:25 +0000 (05:36 +0000)]
auto merge of #15030 : sfackler/rust/partial-cmp, r=huonw
I ended up altering the semantics of Json's PartialOrd implementation.
It used to be the case that Null < Null, but I can't think of any reason
for an ordering other than the default one so I just switched it over to
using the derived implementation.
This also fixes broken `PartialOrd` implementations for `Vec` and
`TreeMap`.
# Note
This isn't ready to merge yet since libcore tests are broken as you end up with 2 versions of `Option`. The rest should be reviewable though.
Steven Fackler [Wed, 18 Jun 2014 06:25:51 +0000 (23:25 -0700)]
Implement RFC#28: Add PartialOrd::partial_cmp
I ended up altering the semantics of Json's PartialOrd implementation.
It used to be the case that Null < Null, but I can't think of any reason
for an ordering other than the default one so I just switched it over to
using the derived implementation.
This also fixes broken `PartialOrd` implementations for `Vec` and
`TreeMap`.
bors [Mon, 30 Jun 2014 03:46:25 +0000 (03:46 +0000)]
auto merge of #15256 : erickt/rust/optimizations, r=alexcrichton
The bug #11084 causes `option::collect` and `result::collect` about twice as slower as it should because llvm is having some trouble optimizing away the scan closure. This gets rid of it so now those functions perform equivalent to a hand written version.
This also adds an impl of `Default` for `Rc` along the way.
bors [Sun, 29 Jun 2014 23:36:43 +0000 (23:36 +0000)]
auto merge of #15245 : sfackler/rust/coretest, r=alexcrichton
Libcore's test infrastructure is complicated by the fact that many lang
items are defined in the crate. The current approach (realcore/realstd
imports) is hacky and hard to work with (tests inside of core::cmp
haven't been run for months!).
Moving tests to a separate crate does mean that they can only test the
public API of libcore, but I don't feel that that is too much of an
issue. The only tests that I had to get rid of were some checking the
various numeric formatters, but those are also exercised through normal
format! calls in other tests.
Steven Fackler [Sat, 28 Jun 2014 20:57:36 +0000 (13:57 -0700)]
Extract tests from libcore to a separate crate
Libcore's test infrastructure is complicated by the fact that many lang
items are defined in the crate. The current approach (realcore/realstd
imports) is hacky and hard to work with (tests inside of core::cmp
haven't been run for months!).
Moving tests to a separate crate does mean that they can only test the
public API of libcore, but I don't feel that that is too much of an
issue. The only tests that I had to get rid of were some checking the
various numeric formatters, but those are also exercised through normal
format! calls in other tests.
bors [Sun, 29 Jun 2014 21:41:45 +0000 (21:41 +0000)]
auto merge of #15252 : alexcrichton/rust/issue-15231, r=pcwalton
When cloning a stream, the data is already guaranteed to be in a consistent
state, so there's no need to perform a zeroing. This prevents segfaults as seen
in #15231
Erick Tryzelaar [Mon, 23 Jun 2014 23:27:54 +0000 (19:27 -0400)]
core: optimize {option,result}::collect
The bug #11084 causes these collect functions to run about
twice as slow as they should because llvm is having trouble
optimizing away the closure for some reason. This patch works
around that performance bug by using a simple adapter iterator
explicitly for capturing if the outer iterator returns an
error.
Alex Crichton [Sun, 29 Jun 2014 16:38:07 +0000 (09:38 -0700)]
rustuv: Don't zero-out data on clones
When cloning a stream, the data is already guaranteed to be in a consistent
state, so there's no need to perform a zeroing. This prevents segfaults as seen
in #15231
Huon Wilson [Sat, 14 Jun 2014 12:46:50 +0000 (22:46 +1000)]
c_str: add `.as_ptr` & `.as_mut_ptr` to replace `.with_[mut_]ref`.
These forms return the pointer directly, rather than the added
indirection, indentation, and inefficiencies of the closures in
`.with_ref` and `.with_mut_ref`. The two closure functions are
deprecated.
Replace
foo(c_str.with_ref(|p| p))
c_str.with_ref(|p| {
foo(p);
bar(p);
})
with
foo(c_str.as_ptr())
let p = c_str.as_ptr();
foo(p);
bar(p);
This change does mean that one has to be careful to avoid writing `let p
= x.to_c_str().as_ptr();` since the `CString` will be freed at the end
of the statement. Previously, `with_ref` was used (and `as_ptr` avoided)
for this reason, but Rust has strongly moved away from closures to more
RAII-style code, and most uses of `.with_ref` where in the form
`.with_ref(|p| p)` anyway, that is, they were exactly `.as_ptr`.
bors [Sat, 28 Jun 2014 22:06:35 +0000 (22:06 +0000)]
auto merge of #15210 : luqmana/rust/windoc, r=alexcrichton
Getting rust to build on windows can be a bit annoying in setting up all the toolchains and whatnot. The whole process is made much easier by using msys2 so let's document that prominently right on the README.