提交 8871c17b 编写于 作者: B bors

Auto merge of #24781 - bluss:vec-drain-range, r=alexcrichton

Implement Vec::drain(\<range type\>) from rust-lang/rfcs#574, tracking issue #23055.

This is a big step forward for vector usability. This is an introduction of an API for removing a range of *m* consecutive elements from a vector, as efficently as possible.

New features:

- Introduce trait `std::collections::range::RangeArgument` implemented by all four built-in range types.
- Change `Vec::drain()` to use `Vec::drain<R: RangeArgument>(R)`

Implementation notes:

- Use @Gankro's idea for memory safety: Use `set_len` on the source vector when creating the iterator, to make sure that the part of the vector that will be modified is unreachable. Fix up things in Drain's destructor — but even if it doesn't run, we don't expose any moved-out-from slots of the vector.
- This `.drain<R>(R)` very close to how it is specified in the RFC.
- Introduced as unstable
- Drain reuses the slice iterator — copying and pasting the same iterator pointer arithmetic again felt very bad
- The `usize` index as a range argument in the RFC is not included. The ranges trait would have to change to accomodate it.

Please help me with:

- Name and location of the new ranges trait.
- Design of the ranges trait
- Understanding Niko's comments about variance (Note: for a long time I was using a straight up &mut Vec in the iterator, but I changed this to permit reusing the slice iterator).

Previous PR and discussion: #23071
......@@ -546,7 +546,7 @@ pub fn is_empty(&self) -> bool { self.len() == 0 }
#[unstable(feature = "collections",
reason = "matches collection reform specification, waiting for dust to settle")]
pub fn drain(&mut self) -> Drain<T> {
Drain { iter: self.data.drain() }
Drain { iter: self.data.drain(..) }
}
/// Drops all items from the binary heap.
......
......@@ -42,7 +42,8 @@
#![feature(slice_patterns)]
#![feature(debug_builders)]
#![feature(utf8_error)]
#![cfg_attr(test, feature(rand, rustc_private, test, hash, collections))]
#![cfg_attr(test, feature(rand, rustc_private, test, hash, collections,
collections_drain, collections_range))]
#![cfg_attr(test, allow(deprecated))] // rand
#![feature(no_std)]
......@@ -82,6 +83,7 @@
pub mod enum_set;
pub mod fmt;
pub mod linked_list;
pub mod range;
pub mod slice;
pub mod str;
pub mod string;
......
// Copyright 2015 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
#![unstable(feature = "collections_range", reason = "was just added")]
//! Range syntax.
use core::option::Option::{self, None, Some};
use core::ops::{RangeFull, Range, RangeTo, RangeFrom};
/// **RangeArgument** is implemented by Rust's built-in range types, produced
/// by range syntax like `..`, `a..`, `..b` or `c..d`.
pub trait RangeArgument<T> {
/// Start index (inclusive)
///
/// Return start value if present, else `None`.
fn start(&self) -> Option<&T> { None }
/// End index (exclusive)
///
/// Return end value if present, else `None`.
fn end(&self) -> Option<&T> { None }
}
impl<T> RangeArgument<T> for RangeFull {}
impl<T> RangeArgument<T> for RangeFrom<T> {
fn start(&self) -> Option<&T> { Some(&self.start) }
}
impl<T> RangeArgument<T> for RangeTo<T> {
fn end(&self) -> Option<&T> { Some(&self.end) }
}
impl<T> RangeArgument<T> for Range<T> {
fn start(&self) -> Option<&T> { Some(&self.start) }
fn end(&self) -> Option<&T> { Some(&self.end) }
}
......@@ -69,6 +69,8 @@
use borrow::{Cow, IntoCow};
use super::range::RangeArgument;
// FIXME- fix places which assume the max vector allowed has memory usize::MAX.
static MAX_MEMORY_SIZE: usize = isize::MAX as usize;
......@@ -714,36 +716,61 @@ pub fn append(&mut self, other: &mut Self) {
unsafe { other.set_len(0); }
}
/// Creates a draining iterator that clears the `Vec` and iterates over
/// the removed items from start to end.
/// Create a draining iterator that removes the specified range in the vector
/// and yields the removed items from start to end. The element range is
/// removed even if the iterator is not consumed until the end.
///
/// Note: It is unspecified how many elements are removed from the vector,
/// if the `Drain` value is leaked.
///
/// # Panics
///
/// Panics if the starting point is greater than the end point or if
/// the end point is greater than the length of the vector.
///
/// # Examples
///
/// ```
/// # #![feature(collections)]
/// let mut v = vec!["a".to_string(), "b".to_string()];
/// for s in v.drain() {
/// // s has type String, not &String
/// println!("{}", s);
/// }
/// assert!(v.is_empty());
/// # #![feature(collections_drain, collections_range)]
///
/// // Draining using `..` clears the whole vector.
/// let mut v = vec![1, 2, 3];
/// let u: Vec<_> = v.drain(..).collect();
/// assert_eq!(v, &[]);
/// assert_eq!(u, &[1, 2, 3]);
/// ```
#[inline]
#[unstable(feature = "collections",
reason = "matches collection reform specification, waiting for dust to settle")]
pub fn drain(&mut self) -> Drain<T> {
#[unstable(feature = "collections_drain",
reason = "recently added, matches RFC")]
pub fn drain<R>(&mut self, range: R) -> Drain<T> where R: RangeArgument<usize> {
// Memory safety
//
// When the Drain is first created, it shortens the length of
// the source vector to make sure no uninitalized or moved-from elements
// are accessible at all if the Drain's destructor never gets to run.
//
// Drain will ptr::read out the values to remove.
// When finished, remaining tail of the vec is copied back to cover
// the hole, and the vector length is restored to the new length.
//
let len = self.len();
let start = *range.start().unwrap_or(&0);
let end = *range.end().unwrap_or(&len);
assert!(start <= end);
assert!(end <= len);
unsafe {
let begin = *self.ptr as *const T;
let end = if mem::size_of::<T>() == 0 {
(*self.ptr as usize + self.len()) as *const T
} else {
(*self.ptr).offset(self.len() as isize) as *const T
};
self.set_len(0);
// set self.vec length's to start, to be safe in case Drain is leaked
self.set_len(start);
// Use the borrow in the IterMut to indicate borrowing behavior of the
// whole Drain iterator (like &mut T).
let range_slice = slice::from_raw_parts_mut(
self.as_mut_ptr().offset(start as isize),
end - start);
Drain {
ptr: begin,
end: end,
marker: PhantomData,
tail_start: end,
tail_len: len - end,
iter: range_slice.iter_mut(),
vec: self as *mut _,
}
}
}
......@@ -1795,14 +1822,16 @@ fn drop(&mut self) {
}
}
/// An iterator that drains a vector.
#[unsafe_no_drop_flag]
#[unstable(feature = "collections",
reason = "recently added as part of collections reform 2")]
pub struct Drain<'a, T:'a> {
ptr: *const T,
end: *const T,
marker: PhantomData<&'a T>,
/// A draining iterator for `Vec<T>`.
#[unstable(feature = "collections_drain", reason = "recently added")]
pub struct Drain<'a, T: 'a> {
/// Index of tail to preserve
tail_start: usize,
/// Length of tail
tail_len: usize,
/// Current remaining range to remove
iter: slice::IterMut<'a, T>,
vec: *mut Vec<T>,
}
unsafe impl<'a, T: Sync> Sync for Drain<'a, T> {}
......@@ -1814,34 +1843,15 @@ impl<'a, T> Iterator for Drain<'a, T> {
#[inline]
fn next(&mut self) -> Option<T> {
unsafe {
if self.ptr == self.end {
None
} else {
if mem::size_of::<T>() == 0 {
// purposefully don't use 'ptr.offset' because for
// vectors with 0-size elements this would return the
// same pointer.
self.ptr = mem::transmute(self.ptr as usize + 1);
// Use a non-null pointer value
Some(ptr::read(EMPTY as *mut T))
} else {
let old = self.ptr;
self.ptr = self.ptr.offset(1);
Some(ptr::read(old))
}
self.iter.next().map(|elt|
unsafe {
ptr::read(elt as *const _)
}
}
)
}
#[inline]
fn size_hint(&self) -> (usize, Option<usize>) {
let diff = (self.end as usize) - (self.ptr as usize);
let size = mem::size_of::<T>();
let exact = diff / (if size == 0 {1} else {size});
(exact, Some(exact))
self.iter.size_hint()
}
}
......@@ -1849,41 +1859,40 @@ fn size_hint(&self) -> (usize, Option<usize>) {
impl<'a, T> DoubleEndedIterator for Drain<'a, T> {
#[inline]
fn next_back(&mut self) -> Option<T> {
unsafe {
if self.end == self.ptr {
None
} else {
if mem::size_of::<T>() == 0 {
// See above for why 'ptr.offset' isn't used
self.end = mem::transmute(self.end as usize - 1);
// Use a non-null pointer value
Some(ptr::read(EMPTY as *mut T))
} else {
self.end = self.end.offset(-1);
Some(ptr::read(self.end))
}
self.iter.next_back().map(|elt|
unsafe {
ptr::read(elt as *const _)
}
}
)
}
}
#[stable(feature = "rust1", since = "1.0.0")]
impl<'a, T> ExactSizeIterator for Drain<'a, T> {}
#[unsafe_destructor]
#[stable(feature = "rust1", since = "1.0.0")]
impl<'a, T> Drop for Drain<'a, T> {
fn drop(&mut self) {
// self.ptr == self.end == mem::POST_DROP_USIZE if drop has already been called,
// so we can use #[unsafe_no_drop_flag].
// exhaust self first
while let Some(_) = self.next() { }
// destroy the remaining elements
for _x in self.by_ref() {}
if self.tail_len > 0 {
unsafe {
let source_vec = &mut *self.vec;
// memmove back untouched tail, update to new length
let start = source_vec.len();
let tail = self.tail_start;
let src = source_vec.as_ptr().offset(tail as isize);
let dst = source_vec.as_mut_ptr().offset(start as isize);
ptr::copy(src, dst, self.tail_len);
source_vec.set_len(start + self.tail_len);
}
}
}
}
#[stable(feature = "rust1", since = "1.0.0")]
impl<'a, T> ExactSizeIterator for Drain<'a, T> {}
////////////////////////////////////////////////////////////////////////////////
// Conversion from &[T] to &Vec<T>
////////////////////////////////////////////////////////////////////////////////
......
......@@ -418,7 +418,7 @@ fn filter<A>((i, v): (usize, Option<A>)) -> Option<(usize, A)> {
}
let filter: fn((usize, Option<V>)) -> Option<(usize, V)> = filter; // coerce to fn ptr
Drain { iter: self.v.drain().enumerate().filter_map(filter) }
Drain { iter: self.v.drain(..).enumerate().filter_map(filter) }
}
/// Returns the number of elements in the map.
......
......@@ -10,6 +10,7 @@
#![feature(box_syntax)]
#![feature(collections)]
#![feature(collections_drain)]
#![feature(core)]
#![feature(hash)]
#![feature(rand)]
......
......@@ -460,7 +460,7 @@ fn test_move_items_zero_sized() {
fn test_drain_items() {
let mut vec = vec![1, 2, 3];
let mut vec2 = vec![];
for i in vec.drain() {
for i in vec.drain(..) {
vec2.push(i);
}
assert_eq!(vec, []);
......@@ -471,7 +471,7 @@ fn test_drain_items() {
fn test_drain_items_reverse() {
let mut vec = vec![1, 2, 3];
let mut vec2 = vec![];
for i in vec.drain().rev() {
for i in vec.drain(..).rev() {
vec2.push(i);
}
assert_eq!(vec, []);
......@@ -482,13 +482,43 @@ fn test_drain_items_reverse() {
fn test_drain_items_zero_sized() {
let mut vec = vec![(), (), ()];
let mut vec2 = vec![];
for i in vec.drain() {
for i in vec.drain(..) {
vec2.push(i);
}
assert_eq!(vec, []);
assert_eq!(vec2, [(), (), ()]);
}
#[test]
#[should_panic]
fn test_drain_out_of_bounds() {
let mut v = vec![1, 2, 3, 4, 5];
v.drain(5..6);
}
#[test]
fn test_drain_range() {
let mut v = vec![1, 2, 3, 4, 5];
for _ in v.drain(4..) {
}
assert_eq!(v, &[1, 2, 3, 4]);
let mut v: Vec<_> = (1..6).map(|x| x.to_string()).collect();
for _ in v.drain(1..4) {
}
assert_eq!(v, &[1.to_string(), 5.to_string()]);
let mut v: Vec<_> = (1..6).map(|x| x.to_string()).collect();
for _ in v.drain(1..4).rev() {
}
assert_eq!(v, &[1.to_string(), 5.to_string()]);
let mut v: Vec<_> = vec![(); 5];
for _ in v.drain(1..4).rev() {
}
assert_eq!(v, &[(), ()]);
}
#[test]
fn test_into_boxed_slice() {
let xs = vec![1, 2, 3];
......
......@@ -12,6 +12,7 @@
#![allow(unused_mut)]
#![feature(collections)]
#![feature(collections_drain)]
extern crate collections;
......@@ -96,5 +97,6 @@ fn from_usize(v: usize) -> Foo {
all_sync_send!(VecMap::<usize>::new(), iter, iter_mut, drain, into_iter, keys, values);
all_sync_send!(Vec::<usize>::new(), into_iter, drain);
all_sync_send!(Vec::<usize>::new(), into_iter);
is_sync_send!(Vec::<usize>::new(), drain(..));
}
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册