Rust wrapper for libctru
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

1885 lines
58 KiB

// 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.
//! Traits, helpers, and type definitions for core I/O functionality.
//!
//! The `std::io` module contains a number of common things you'll need
//! when doing input and output. The most core part of this module is
//! the [`Read`][read] and [`Write`][write] traits, which provide the
//! most general interface for reading and writing input and output.
//!
//! [read]: trait.Read.html
//! [write]: trait.Write.html
//!
//! # Read and Write
//!
//! Because they are traits, `Read` and `Write` are implemented by a number
//! of other types, and you can implement them for your types too. As such,
//! you'll see a few different types of I/O throughout the documentation in
//! this module: `File`s, `TcpStream`s, and sometimes even `Vec<T>`s. For
//! example, `Read` adds a `read()` method, which we can use on `File`s:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//! use std::fs::File;
//!
//! # fn foo() -> io::Result<()> {
//! let mut f = try!(File::open("foo.txt"));
//! let mut buffer = [0; 10];
//!
//! // read up to 10 bytes
//! try!(f.read(&mut buffer));
//!
//! println!("The bytes: {:?}", buffer);
//! # Ok(())
//! # }
//! ```
//!
//! `Read` and `Write` are so important, implementors of the two traits have a
//! nickname: readers and writers. So you'll sometimes see 'a reader' instead
//! of 'a type that implements the `Read` trait'. Much easier!
//!
//! ## Seek and BufRead
//!
//! Beyond that, there are two important traits that are provided: [`Seek`][seek]
//! and [`BufRead`][bufread]. Both of these build on top of a reader to control
//! how the reading happens. `Seek` lets you control where the next byte is
//! coming from:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//! use std::io::SeekFrom;
//! use std::fs::File;
//!
//! # fn foo() -> io::Result<()> {
//! let mut f = try!(File::open("foo.txt"));
//! let mut buffer = [0; 10];
//!
//! // skip to the last 10 bytes of the file
//! try!(f.seek(SeekFrom::End(-10)));
//!
//! // read up to 10 bytes
//! try!(f.read(&mut buffer));
//!
//! println!("The bytes: {:?}", buffer);
//! # Ok(())
//! # }
//! ```
//!
//! [seek]: trait.Seek.html
//! [bufread]: trait.BufRead.html
//!
//! `BufRead` uses an internal buffer to provide a number of other ways to read, but
//! to show it off, we'll need to talk about buffers in general. Keep reading!
//!
//! ## BufReader and BufWriter
//!
//! Byte-based interfaces are unwieldy and can be inefficient, as we'd need to be
//! making near-constant calls to the operating system. To help with this,
//! `std::io` comes with two structs, `BufReader` and `BufWriter`, which wrap
//! readers and writers. The wrapper uses a buffer, reducing the number of
//! calls and providing nicer methods for accessing exactly what you want.
//!
//! For example, `BufReader` works with the `BufRead` trait to add extra
//! methods to any reader:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//! use std::io::BufReader;
//! use std::fs::File;
//!
//! # fn foo() -> io::Result<()> {
//! let f = try!(File::open("foo.txt"));
//! let mut reader = BufReader::new(f);
//! let mut buffer = String::new();
//!
//! // read a line into buffer
//! try!(reader.read_line(&mut buffer));
//!
//! println!("{}", buffer);
//! # Ok(())
//! # }
//! ```
//!
//! `BufWriter` doesn't add any new ways of writing; it just buffers every call
//! to [`write()`][write()]:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//! use std::io::BufWriter;
//! use std::fs::File;
//!
//! # fn foo() -> io::Result<()> {
//! let f = try!(File::create("foo.txt"));
//! {
//! let mut writer = BufWriter::new(f);
//!
//! // write a byte to the buffer
//! try!(writer.write(&[42]));
//!
//! } // the buffer is flushed once writer goes out of scope
//!
//! # Ok(())
//! # }
//! ```
//!
//! [write()]: trait.Write.html#tymethod.write
//!
//! ## Standard input and output
//!
//! A very common source of input is standard input:
//!
//! ```
//! use std::io;
//!
//! # fn foo() -> io::Result<()> {
//! let mut input = String::new();
//!
//! try!(io::stdin().read_line(&mut input));
//!
//! println!("You typed: {}", input.trim());
//! # Ok(())
//! # }
//! ```
//!
//! And a very common source of output is standard output:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//!
//! # fn foo() -> io::Result<()> {
//! try!(io::stdout().write(&[42]));
//! # Ok(())
//! # }
//! ```
//!
//! Of course, using `io::stdout()` directly is less common than something like
//! `println!`.
//!
//! ## Iterator types
//!
//! A large number of the structures provided by `std::io` are for various
//! ways of iterating over I/O. For example, `Lines` is used to split over
//! lines:
//!
//! ```
//! use std::io;
//! use std::io::prelude::*;
//! use std::io::BufReader;
//! use std::fs::File;
//!
//! # fn foo() -> io::Result<()> {
//! let f = try!(File::open("foo.txt"));
//! let reader = BufReader::new(f);
//!
//! for line in reader.lines() {
//! println!("{}", try!(line));
//! }
//!
//! # Ok(())
//! # }
//! ```
//!
//! ## Functions
//!
//! There are a number of [functions][functions-list] that offer access to various
//! features. For example, we can use three of these functions to copy everything
//! from standard input to standard output:
//!
//! ```
//! use std::io;
//!
//! # fn foo() -> io::Result<()> {
//! try!(io::copy(&mut io::stdin(), &mut io::stdout()));
//! # Ok(())
//! # }
//! ```
//!
//! [functions-list]: #functions-1
//!
//! ## io::Result
//!
//! Last, but certainly not least, is [`io::Result`][result]. This type is used
//! as the return type of many `std::io` functions that can cause an error, and
//! can be returned from your own functions as well. Many of the examples in this
//! module use the [`try!`][try] macro:
//!
//! ```
//! use std::io;
//!
//! fn read_input() -> io::Result<()> {
//! let mut input = String::new();
//!
//! try!(io::stdin().read_line(&mut input));
//!
//! println!("You typed: {}", input.trim());
//!
//! Ok(())
//! }
//! ```
//!
//! The return type of `read_input()`, `io::Result<()>`, is a very common type
//! for functions which don't have a 'real' return value, but do want to return
//! errors if they happen. In this case, the only purpose of this function is
//! to read the line and print it, so we use `()`.
//!
//! [result]: type.Result.html
//! [try]: ../macro.try.html
//!
//! ## Platform-specific behavior
//!
//! Many I/O functions throughout the standard library are documented to indicate
//! what various library or syscalls they are delegated to. This is done to help
//! applications both understand what's happening under the hood as well as investigate
//! any possibly unclear semantics. Note, however, that this is informative, not a binding
//! contract. The implementation of many of these functions are subject to change over
//! time and may call fewer or more syscalls/library functions.
use cmp;
use rustc_unicode::str as core_str;
use error as std_error;
use fmt;
use result;
use str;
use memchr;
pub use self::buffered::{BufReader, BufWriter, LineWriter};
pub use self::buffered::IntoInnerError;
pub use self::cursor::Cursor;
pub use self::error::{Result, Error, ErrorKind};
pub use self::util::{copy, sink, Sink, empty, Empty, repeat, Repeat};
pub use self::print::{STDOUT, _print};
//pub use self::stdio::{stdin, stdout, stderr, _print, Stdin, Stdout, Stderr};
//pub use self::stdio::{StdoutLock, StderrLock, StdinLock};
#[doc(no_inline, hidden)]
//pub use self::stdio::{set_panic, set_print};
pub mod prelude;
mod buffered;
mod cursor;
mod error;
mod impls;
mod util;
mod print;
//mod lazy;
//mod stdio;
const DEFAULT_BUF_SIZE: usize = 8 * 1024;
// A few methods below (read_to_string, read_line) will append data into a
// `String` buffer, but we need to be pretty careful when doing this. The
// implementation will just call `.as_mut_vec()` and then delegate to a
// byte-oriented reading method, but we must ensure that when returning we never
// leave `buf` in a state such that it contains invalid UTF-8 in its bounds.
//
// To this end, we use an RAII guard (to protect against panics) which updates
// the length of the string when it is dropped. This guard initially truncates
// the string to the prior length and only after we've validated that the
// new contents are valid UTF-8 do we allow it to set a longer length.
//
// The unsafety in this function is twofold:
//
// 1. We're looking at the raw bytes of `buf`, so we take on the burden of UTF-8
// checks.
// 2. We're passing a raw buffer to the function `f`, and it is expected that
// the function only *appends* bytes to the buffer. We'll get undefined
// behavior if existing bytes are overwritten to have non-UTF-8 data.
fn append_to_string<F>(buf: &mut String, f: F) -> Result<usize>
where F: FnOnce(&mut Vec<u8>) -> Result<usize>
{
struct Guard<'a> { s: &'a mut Vec<u8>, len: usize }
impl<'a> Drop for Guard<'a> {
fn drop(&mut self) {
unsafe { self.s.set_len(self.len); }
}
}
unsafe {
let mut g = Guard { len: buf.len(), s: buf.as_mut_vec() };
let ret = f(g.s);
if str::from_utf8(&g.s[g.len..]).is_err() {
ret.and_then(|_| {
Err(Error::new(ErrorKind::InvalidData,
"stream did not contain valid UTF-8"))
})
} else {
g.len = g.s.len();
ret
}
}
}
// This uses an adaptive system to extend the vector when it fills. We want to
// avoid paying to allocate and zero a huge chunk of memory if the reader only
// has 4 bytes while still making large reads if the reader does have a ton
// of data to return. Simply tacking on an extra DEFAULT_BUF_SIZE space every
// time is 4,500 times (!) slower than this if the reader has a very small
// amount of data to return.
fn read_to_end<R: Read + ?Sized>(r: &mut R, buf: &mut Vec<u8>) -> Result<usize> {
let start_len = buf.len();
let mut len = start_len;
let mut new_write_size = 16;
let ret;
loop {
if len == buf.len() {
if new_write_size < DEFAULT_BUF_SIZE {
new_write_size *= 2;
}
buf.resize(len + new_write_size, 0);
}
match r.read(&mut buf[len..]) {
Ok(0) => {
ret = Ok(len - start_len);
break;
}
Ok(n) => len += n,
Err(ref e) if e.kind() == ErrorKind::Interrupted => {}
Err(e) => {
ret = Err(e);
break;
}
}
}
buf.truncate(len);
ret
}
/// The `Read` trait allows for reading bytes from a source.
///
/// Implementors of the `Read` trait are sometimes called 'readers'.
///
/// Readers are defined by one required method, `read()`. Each call to `read`
/// will attempt to pull bytes from this source into a provided buffer. A
/// number of other methods are implemented in terms of `read()`, giving
/// implementors a number of ways to read bytes while only needing to implement
/// a single method.
///
/// Readers are intended to be composable with one another. Many implementors
/// throughout `std::io` take and provide types which implement the `Read`
/// trait.
///
/// Please note that each call to `read` may involve a system call, and
/// therefore, using something that implements [`BufRead`][bufread], such as
/// [`BufReader`][bufreader], will be more efficient.
///
/// [bufread]: trait.BufRead.html
/// [bufreader]: struct.BufReader.html
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = [0; 10];
///
/// // read up to 10 bytes
/// try!(f.read(&mut buffer));
///
/// let mut buffer = vec![0; 10];
/// // read the whole file
/// try!(f.read_to_end(&mut buffer));
///
/// // read into a String, so that you don't need to do the conversion.
/// let mut buffer = String::new();
/// try!(f.read_to_string(&mut buffer));
///
/// // and more! See the other methods for more details.
/// # Ok(())
/// # }
/// ```
pub trait Read {
/// Pull some bytes from this source into the specified buffer, returning
/// how many bytes were read.
///
/// This function does not provide any guarantees about whether it blocks
/// waiting for data, but if an object needs to block for a read but cannot
/// it will typically signal this via an `Err` return value.
///
/// If the return value of this method is `Ok(n)`, then it must be
/// guaranteed that `0 <= n <= buf.len()`. A nonzero `n` value indicates
/// that the buffer `buf` has been filled in with `n` bytes of data from this
/// source. If `n` is `0`, then it can indicate one of two scenarios:
///
/// 1. This reader has reached its "end of file" and will likely no longer
/// be able to produce bytes. Note that this does not mean that the
/// reader will *always* no longer be able to produce bytes.
/// 2. The buffer specified was 0 bytes in length.
///
/// No guarantees are provided about the contents of `buf` when this
/// function is called, implementations cannot rely on any property of the
/// contents of `buf` being true. It is recommended that implementations
/// only write data to `buf` instead of reading its contents.
///
/// # Errors
///
/// If this function encounters any form of I/O or other error, an error
/// variant will be returned. If an error is returned then it must be
/// guaranteed that no bytes were read.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = [0; 10];
///
/// // read 10 bytes
/// try!(f.read(&mut buffer[..]));
/// # Ok(())
/// # }
/// ```
fn read(&mut self, buf: &mut [u8]) -> Result<usize>;
/// Read all bytes until EOF in this source, placing them into `buf`.
///
/// All bytes read from this source will be appended to the specified buffer
/// `buf`. This function will continuously call `read` to append more data to
/// `buf` until `read` returns either `Ok(0)` or an error of
/// non-`ErrorKind::Interrupted` kind.
///
/// If successful, this function will return the total number of bytes read.
///
/// # Errors
///
/// If this function encounters an error of the kind
/// `ErrorKind::Interrupted` then the error is ignored and the operation
/// will continue.
///
/// If any other read error is encountered then this function immediately
/// returns. Any bytes which have already been read will be appended to
/// `buf`.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = Vec::new();
///
/// // read the whole file
/// try!(f.read_to_end(&mut buffer));
/// # Ok(())
/// # }
/// ```
fn read_to_end(&mut self, buf: &mut Vec<u8>) -> Result<usize> {
read_to_end(self, buf)
}
/// Read all bytes until EOF in this source, placing them into `buf`.
///
/// If successful, this function returns the number of bytes which were read
/// and appended to `buf`.
///
/// # Errors
///
/// If the data in this stream is *not* valid UTF-8 then an error is
/// returned and `buf` is unchanged.
///
/// See [`read_to_end()`][readtoend] for other error semantics.
///
/// [readtoend]: #method.read_to_end
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = String::new();
///
/// try!(f.read_to_string(&mut buffer));
/// # Ok(())
/// # }
/// ```
fn read_to_string(&mut self, buf: &mut String) -> Result<usize> {
// Note that we do *not* call `.read_to_end()` here. We are passing
// `&mut Vec<u8>` (the raw contents of `buf`) into the `read_to_end`
// method to fill it up. An arbitrary implementation could overwrite the
// entire contents of the vector, not just append to it (which is what
// we are expecting).
//
// To prevent extraneously checking the UTF-8-ness of the entire buffer
// we pass it to our hardcoded `read_to_end` implementation which we
// know is guaranteed to only read data into the end of the buffer.
append_to_string(buf, |b| read_to_end(self, b))
}
/// Read the exact number of bytes required to fill `buf`.
///
/// This function reads as many bytes as necessary to completely fill the
/// specified buffer `buf`.
///
/// No guarantees are provided about the contents of `buf` when this
/// function is called, implementations cannot rely on any property of the
/// contents of `buf` being true. It is recommended that implementations
/// only write data to `buf` instead of reading its contents.
///
/// # Errors
///
/// If this function encounters an error of the kind
/// `ErrorKind::Interrupted` then the error is ignored and the operation
/// will continue.
///
/// If this function encounters an "end of file" before completely filling
/// the buffer, it returns an error of the kind `ErrorKind::UnexpectedEof`.
/// The contents of `buf` are unspecified in this case.
///
/// If any other read error is encountered then this function immediately
/// returns. The contents of `buf` are unspecified in this case.
///
/// If this function returns an error, it is unspecified how many bytes it
/// has read, but it will never read more than would be necessary to
/// completely fill the buffer.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = [0; 10];
///
/// // read exactly 10 bytes
/// try!(f.read_exact(&mut buffer));
/// # Ok(())
/// # }
/// ```
fn read_exact(&mut self, mut buf: &mut [u8]) -> Result<()> {
while !buf.is_empty() {
match self.read(buf) {
Ok(0) => break,
Ok(n) => { let tmp = buf; buf = &mut tmp[n..]; }
Err(ref e) if e.kind() == ErrorKind::Interrupted => {}
Err(e) => return Err(e),
}
}
if !buf.is_empty() {
Err(Error::new(ErrorKind::UnexpectedEof,
"failed to fill whole buffer"))
} else {
Ok(())
}
}
/// Creates a "by reference" adaptor for this instance of `Read`.
///
/// The returned adaptor also implements `Read` and will simply borrow this
/// current reader.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::Read;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = Vec::new();
/// let mut other_buffer = Vec::new();
///
/// {
/// let reference = f.by_ref();
///
/// // read at most 5 bytes
/// try!(reference.take(5).read_to_end(&mut buffer));
///
/// } // drop our &mut reference so we can use f again
///
/// // original file still usable, read the rest
/// try!(f.read_to_end(&mut other_buffer));
/// # Ok(())
/// # }
/// ```
fn by_ref(&mut self) -> &mut Self where Self: Sized { self }
/// Transforms this `Read` instance to an `Iterator` over its bytes.
///
/// The returned type implements `Iterator` where the `Item` is `Result<u8,
/// R::Err>`. The yielded item is `Ok` if a byte was successfully read and
/// `Err` otherwise for I/O errors. EOF is mapped to returning `None` from
/// this iterator.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
///
/// for byte in f.bytes() {
/// println!("{}", byte.unwrap());
/// }
/// # Ok(())
/// # }
/// ```
fn bytes(self) -> Bytes<Self> where Self: Sized {
Bytes { inner: self }
}
/// Transforms this `Read` instance to an `Iterator` over `char`s.
///
/// This adaptor will attempt to interpret this reader as a UTF-8 encoded
/// sequence of characters. The returned iterator will return `None` once
/// EOF is reached for this reader. Otherwise each element yielded will be a
/// `Result<char, E>` where `E` may contain information about what I/O error
/// occurred or where decoding failed.
///
/// Currently this adaptor will discard intermediate data read, and should
/// be avoided if this is not desired.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// #![feature(io)]
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
///
/// for c in f.chars() {
/// println!("{}", c.unwrap());
/// }
/// # Ok(())
/// # }
/// ```
fn chars(self) -> Chars<Self> where Self: Sized {
Chars { inner: self }
}
/// Creates an adaptor which will chain this stream with another.
///
/// The returned `Read` instance will first read all bytes from this object
/// until EOF is encountered. Afterwards the output is equivalent to the
/// output of `next`.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f1 = try!(File::open("foo.txt"));
/// let mut f2 = try!(File::open("bar.txt"));
///
/// let mut handle = f1.chain(f2);
/// let mut buffer = String::new();
///
/// // read the value into a String. We could use any Read method here,
/// // this is just one example.
/// try!(handle.read_to_string(&mut buffer));
/// # Ok(())
/// # }
/// ```
fn chain<R: Read>(self, next: R) -> Chain<Self, R> where Self: Sized {
Chain { first: self, second: next, done_first: false }
}
/// Creates an adaptor which will read at most `limit` bytes from it.
///
/// This function returns a new instance of `Read` which will read at most
/// `limit` bytes, after which it will always return EOF (`Ok(0)`). Any
/// read errors will not count towards the number of bytes read and future
/// calls to `read` may succeed.
///
/// # Examples
///
/// [`File`][file]s implement `Read`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
/// let mut buffer = [0; 5];
///
/// // read at most five bytes
/// let mut handle = f.take(5);
///
/// try!(handle.read(&mut buffer));
/// # Ok(())
/// # }
/// ```
fn take(self, limit: u64) -> Take<Self> where Self: Sized {
Take { inner: self, limit: limit }
}
}
/// A trait for objects which are byte-oriented sinks.
///
/// Implementors of the `Write` trait are sometimes called 'writers'.
///
/// Writers are defined by two required methods, `write()` and `flush()`:
///
/// * The `write()` method will attempt to write some data into the object,
/// returning how many bytes were successfully written.
///
/// * The `flush()` method is useful for adaptors and explicit buffers
/// themselves for ensuring that all buffered data has been pushed out to the
/// 'true sink'.
///
/// Writers are intended to be composable with one another. Many implementors
/// throughout `std::io` take and provide types which implement the `Write`
/// trait.
///
/// # Examples
///
/// ```
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = try!(File::create("foo.txt"));
///
/// try!(buffer.write(b"some bytes"));
/// # Ok(())
/// # }
/// ```
pub trait Write {
/// Write a buffer into this object, returning how many bytes were written.
///
/// This function will attempt to write the entire contents of `buf`, but
/// the entire write may not succeed, or the write may also generate an
/// error. A call to `write` represents *at most one* attempt to write to
/// any wrapped object.
///
/// Calls to `write` are not guaranteed to block waiting for data to be
/// written, and a write which would otherwise block can be indicated through
/// an `Err` variant.
///
/// If the return value is `Ok(n)` then it must be guaranteed that
/// `0 <= n <= buf.len()`. A return value of `0` typically means that the
/// underlying object is no longer able to accept bytes and will likely not
/// be able to in the future as well, or that the buffer provided is empty.
///
/// # Errors
///
/// Each call to `write` may generate an I/O error indicating that the
/// operation could not be completed. If an error is returned then no bytes
/// in the buffer were written to this writer.
///
/// It is **not** considered an error if the entire buffer could not be
/// written to this writer.
///
/// # Examples
///
/// ```
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = try!(File::create("foo.txt"));
///
/// try!(buffer.write(b"some bytes"));
/// # Ok(())
/// # }
/// ```
fn write(&mut self, buf: &[u8]) -> Result<usize>;
/// Flush this output stream, ensuring that all intermediately buffered
/// contents reach their destination.
///
/// # Errors
///
/// It is considered an error if not all bytes could be written due to
/// I/O errors or EOF being reached.
///
/// # Examples
///
/// ```
/// use std::io::prelude::*;
/// use std::io::BufWriter;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = BufWriter::new(try!(File::create("foo.txt")));
///
/// try!(buffer.write(b"some bytes"));
/// try!(buffer.flush());
/// # Ok(())
/// # }
/// ```
fn flush(&mut self) -> Result<()>;
/// Attempts to write an entire buffer into this write.
///
/// This method will continuously call `write` while there is more data to
/// write. This method will not return until the entire buffer has been
/// successfully written or an error occurs. The first error generated from
/// this method will be returned.
///
/// # Errors
///
/// This function will return the first error that `write` returns.
///
/// # Examples
///
/// ```
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = try!(File::create("foo.txt"));
///
/// try!(buffer.write_all(b"some bytes"));
/// # Ok(())
/// # }
/// ```
fn write_all(&mut self, mut buf: &[u8]) -> Result<()> {
while !buf.is_empty() {
match self.write(buf) {
Ok(0) => return Err(Error::new(ErrorKind::WriteZero,
"failed to write whole buffer")),
Ok(n) => buf = &buf[n..],
Err(ref e) if e.kind() == ErrorKind::Interrupted => {}
Err(e) => return Err(e),
}
}
Ok(())
}
/// Writes a formatted string into this writer, returning any error
/// encountered.
///
/// This method is primarily used to interface with the
/// [`format_args!`][formatargs] macro, but it is rare that this should
/// explicitly be called. The [`write!`][write] macro should be favored to
/// invoke this method instead.
///
/// [formatargs]: ../macro.format_args.html
/// [write]: ../macro.write.html
///
/// This function internally uses the [`write_all`][writeall] method on
/// this trait and hence will continuously write data so long as no errors
/// are received. This also means that partial writes are not indicated in
/// this signature.
///
/// [writeall]: #method.write_all
///
/// # Errors
///
/// This function will return any I/O error reported while formatting.
///
/// # Examples
///
/// ```
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = try!(File::create("foo.txt"));
///
/// // this call
/// try!(write!(buffer, "{:.*}", 2, 1.234567));
/// // turns into this:
/// try!(buffer.write_fmt(format_args!("{:.*}", 2, 1.234567)));
/// # Ok(())
/// # }
/// ```
fn write_fmt(&mut self, fmt: fmt::Arguments) -> Result<()> {
// Create a shim which translates a Write to a fmt::Write and saves
// off I/O errors. instead of discarding them
struct Adaptor<'a, T: ?Sized + 'a> {
inner: &'a mut T,
error: Result<()>,
}
impl<'a, T: Write + ?Sized> fmt::Write for Adaptor<'a, T> {
fn write_str(&mut self, s: &str) -> fmt::Result {
match self.inner.write_all(s.as_bytes()) {
Ok(()) => Ok(()),
Err(e) => {
self.error = Err(e);
Err(fmt::Error)
}
}
}
}
let mut output = Adaptor { inner: self, error: Ok(()) };
match fmt::write(&mut output, fmt) {
Ok(()) => Ok(()),
Err(..) => {
// check if the error came from the underlying `Write` or not
if output.error.is_err() {
output.error
} else {
Err(Error::new(ErrorKind::Other, "formatter error"))
}
}
}
}
/// Creates a "by reference" adaptor for this instance of `Write`.
///
/// The returned adaptor also implements `Write` and will simply borrow this
/// current writer.
///
/// # Examples
///
/// ```
/// use std::io::Write;
/// use std::fs::File;
///
/// # fn foo() -> std::io::Result<()> {
/// let mut buffer = try!(File::create("foo.txt"));
///
/// let reference = buffer.by_ref();
///
/// // we can use reference just like our original buffer
/// try!(reference.write_all(b"some bytes"));
/// # Ok(())
/// # }
/// ```
fn by_ref(&mut self) -> &mut Self where Self: Sized { self }
}
/// The `Seek` trait provides a cursor which can be moved within a stream of
/// bytes.
///
/// The stream typically has a fixed size, allowing seeking relative to either
/// end or the current offset.
///
/// # Examples
///
/// [`File`][file]s implement `Seek`:
///
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
/// use std::io::SeekFrom;
///
/// # fn foo() -> io::Result<()> {
/// let mut f = try!(File::open("foo.txt"));
///
/// // move the cursor 42 bytes from the start of the file
/// try!(f.seek(SeekFrom::Start(42)));
/// # Ok(())
/// # }
/// ```
pub trait Seek {
/// Seek to an offset, in bytes, in a stream.
///
/// A seek beyond the end of a stream is allowed, but implementation
/// defined.
///
/// If the seek operation completed successfully,
/// this method returns the new position from the start of the stream.
/// That position can be used later with [`SeekFrom::Start`].
///
/// # Errors
///
/// Seeking to a negative offset is considered an error.
///
/// [`SeekFrom::Start`]: enum.SeekFrom.html#variant.Start
fn seek(&mut self, pos: SeekFrom) -> Result<u64>;
}
/// Enumeration of possible methods to seek within an I/O object.
///
/// It is used by the [`Seek`] trait.
///
/// [`Seek`]: trait.Seek.html
#[derive(Copy, PartialEq, Eq, Clone, Debug)]
pub enum SeekFrom {
/// Set the offset to the provided number of bytes.
Start(u64),
/// Set the offset to the size of this object plus the specified number of
/// bytes.
///
/// It is possible to seek beyond the end of an object, but it's an error to
/// seek before byte 0.
End(i64),
/// Set the offset to the current position plus the specified number of
/// bytes.
///
/// It is possible to seek beyond the end of an object, but it's an error to
/// seek before byte 0.
Current(i64),
}
fn read_until<R: BufRead + ?Sized>(r: &mut R, delim: u8, buf: &mut Vec<u8>)
-> Result<usize> {
let mut read = 0;
loop {
let (done, used) = {
let available = match r.fill_buf() {
Ok(n) => n,
Err(ref e) if e.kind() == ErrorKind::Interrupted => continue,
Err(e) => return Err(e)
};
match memchr::memchr(delim, available) {
Some(i) => {
buf.extend_from_slice(&available[..i + 1]);
(true, i + 1)
}
None => {
buf.extend_from_slice(available);
(false, available.len())
}
}
};
r.consume(used);
read += used;
if done || used == 0 {
return Ok(read);
}
}
}
/// A `BufRead` is a type of `Read`er which has an internal buffer, allowing it
/// to perform extra ways of reading.
///
/// For example, reading line-by-line is inefficient without using a buffer, so
/// if you want to read by line, you'll need `BufRead`, which includes a
/// [`read_line()`][readline] method as well as a [`lines()`][lines] iterator.
///
/// [readline]: #method.read_line
/// [lines]: #method.lines
///
/// # Examples
///
/// A locked standard input implements `BufRead`:
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// let stdin = io::stdin();
/// for line in stdin.lock().lines() {
/// println!("{}", line.unwrap());
/// }
/// ```
///
/// If you have something that implements `Read`, you can use the [`BufReader`
/// type][bufreader] to turn it into a `BufRead`.
///
/// For example, [`File`][file] implements `Read`, but not `BufRead`.
/// `BufReader` to the rescue!
///
/// [bufreader]: struct.BufReader.html
/// [file]: ../fs/struct.File.html
///
/// ```
/// use std::io::{self, BufReader};
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let f = try!(File::open("foo.txt"));
/// let f = BufReader::new(f);
///
/// for line in f.lines() {
/// println!("{}", line.unwrap());
/// }
///
/// # Ok(())
/// # }
/// ```
///
pub trait BufRead: Read {
/// Fills the internal buffer of this object, returning the buffer contents.
///
/// This function is a lower-level call. It needs to be paired with the
/// [`consume`][consume] method to function properly. When calling this
/// method, none of the contents will be "read" in the sense that later
/// calling `read` may return the same contents. As such, `consume` must be
/// called with the number of bytes that are consumed from this buffer to
/// ensure that the bytes are never returned twice.
///
/// [consume]: #tymethod.consume
///
/// An empty buffer returned indicates that the stream has reached EOF.
///
/// # Errors
///
/// This function will return an I/O error if the underlying reader was
/// read, but returned an error.
///
/// # Examples
///
/// A locked standard input implements `BufRead`:
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// let stdin = io::stdin();
/// let mut stdin = stdin.lock();
///
/// // we can't have two `&mut` references to `stdin`, so use a block
/// // to end the borrow early.
/// let length = {
/// let buffer = stdin.fill_buf().unwrap();
///
/// // work with buffer
/// println!("{:?}", buffer);
///
/// buffer.len()
/// };
///
/// // ensure the bytes we worked with aren't returned again later
/// stdin.consume(length);
/// ```
fn fill_buf(&mut self) -> Result<&[u8]>;
/// Tells this buffer that `amt` bytes have been consumed from the buffer,
/// so they should no longer be returned in calls to `read`.
///
/// This function is a lower-level call. It needs to be paired with the
/// [`fill_buf`][fillbuf] method to function properly. This function does
/// not perform any I/O, it simply informs this object that some amount of
/// its buffer, returned from `fill_buf`, has been consumed and should no
/// longer be returned. As such, this function may do odd things if
/// `fill_buf` isn't called before calling it.
///
/// [fillbuf]: #tymethod.fill_buf
///
/// The `amt` must be `<=` the number of bytes in the buffer returned by
/// `fill_buf`.
///
/// # Examples
///
/// Since `consume()` is meant to be used with [`fill_buf()`][fillbuf],
/// that method's example includes an example of `consume()`.
fn consume(&mut self, amt: usize);
/// Read all bytes into `buf` until the delimiter `byte` is reached.
///
/// This function will read bytes from the underlying stream until the
/// delimiter or EOF is found. Once found, all bytes up to, and including,
/// the delimiter (if found) will be appended to `buf`.
///
/// If this reader is currently at EOF then this function will not modify
/// `buf` and will return `Ok(n)` where `n` is the number of bytes which
/// were read.
///
/// # Errors
///
/// This function will ignore all instances of `ErrorKind::Interrupted` and
/// will otherwise return any errors returned by `fill_buf`.
///
/// If an I/O error is encountered then all bytes read so far will be
/// present in `buf` and its length will have been adjusted appropriately.
///
/// # Examples
///
/// A locked standard input implements `BufRead`. In this example, we'll
/// read from standard input until we see an `a` byte.
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// fn foo() -> io::Result<()> {
/// let stdin = io::stdin();
/// let mut stdin = stdin.lock();
/// let mut buffer = Vec::new();
///
/// try!(stdin.read_until(b'a', &mut buffer));
///
/// println!("{:?}", buffer);
/// # Ok(())
/// # }
/// ```
fn read_until(&mut self, byte: u8, buf: &mut Vec<u8>) -> Result<usize> {
read_until(self, byte, buf)
}
/// Read all bytes until a newline (the 0xA byte) is reached, and append
/// them to the provided buffer.
///
/// This function will read bytes from the underlying stream until the
/// newline delimiter (the 0xA byte) or EOF is found. Once found, all bytes
/// up to, and including, the delimiter (if found) will be appended to
/// `buf`.
///
/// If this reader is currently at EOF then this function will not modify
/// `buf` and will return `Ok(n)` where `n` is the number of bytes which
/// were read.
///
/// # Errors
///
/// This function has the same error semantics as `read_until` and will also
/// return an error if the read bytes are not valid UTF-8. If an I/O error
/// is encountered then `buf` may contain some bytes already read in the
/// event that all data read so far was valid UTF-8.
///
/// # Examples
///
/// A locked standard input implements `BufRead`. In this example, we'll
/// read all of the lines from standard input. If we were to do this in
/// an actual project, the [`lines()`][lines] method would be easier, of
/// course.
///
/// [lines]: #method.lines
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// let stdin = io::stdin();
/// let mut stdin = stdin.lock();
/// let mut buffer = String::new();
///
/// while stdin.read_line(&mut buffer).unwrap() > 0 {
/// // work with buffer
/// println!("{:?}", buffer);
///
/// buffer.clear();
/// }
/// ```
fn read_line(&mut self, buf: &mut String) -> Result<usize> {
// Note that we are not calling the `.read_until` method here, but
// rather our hardcoded implementation. For more details as to why, see
// the comments in `read_to_end`.
append_to_string(buf, |b| read_until(self, b'\n', b))
}
/// Returns an iterator over the contents of this reader split on the byte
/// `byte`.
///
/// The iterator returned from this function will return instances of
/// `io::Result<Vec<u8>>`. Each vector returned will *not* have the
/// delimiter byte at the end.
///
/// This function will yield errors whenever `read_until` would have also
/// yielded an error.
///
/// # Examples
///
/// A locked standard input implements `BufRead`. In this example, we'll
/// read some input from standard input, splitting on commas.
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// let stdin = io::stdin();
///
/// for content in stdin.lock().split(b',') {
/// println!("{:?}", content.unwrap());
/// }
/// ```
fn split(self, byte: u8) -> Split<Self> where Self: Sized {
Split { buf: self, delim: byte }
}
/// Returns an iterator over the lines of this reader.
///
/// The iterator returned from this function will yield instances of
/// `io::Result<String>`. Each string returned will *not* have a newline
/// byte (the 0xA byte) or CRLF (0xD, 0xA bytes) at the end.
///
/// # Examples
///
/// A locked standard input implements `BufRead`:
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
///
/// let stdin = io::stdin();
///
/// for line in stdin.lock().lines() {
/// println!("{}", line.unwrap());
/// }
/// ```
fn lines(self) -> Lines<Self> where Self: Sized {
Lines { buf: self }
}
}
/// Adaptor to chain together two readers.
///
/// This struct is generally created by calling [`chain()`][chain] on a reader.
/// Please see the documentation of `chain()` for more details.
///
/// [chain]: trait.Read.html#method.chain
pub struct Chain<T, U> {
first: T,
second: U,
done_first: bool,
}
impl<T: Read, U: Read> Read for Chain<T, U> {
fn read(&mut self, buf: &mut [u8]) -> Result<usize> {
if !self.done_first {
match self.first.read(buf)? {
0 => { self.done_first = true; }
n => return Ok(n),
}
}
self.second.read(buf)
}
}
impl<T: BufRead, U: BufRead> BufRead for Chain<T, U> {
fn fill_buf(&mut self) -> Result<&[u8]> {
if !self.done_first {
match self.first.fill_buf()? {
buf if buf.len() == 0 => { self.done_first = true; }
buf => return Ok(buf),
}
}
self.second.fill_buf()
}
fn consume(&mut self, amt: usize) {
if !self.done_first {
self.first.consume(amt)
} else {
self.second.consume(amt)
}
}
}
/// Reader adaptor which limits the bytes read from an underlying reader.
///
/// This struct is generally created by calling [`take()`][take] on a reader.
/// Please see the documentation of `take()` for more details.
///
/// [take]: trait.Read.html#method.take
pub struct Take<T> {
inner: T,
limit: u64,
}
impl<T> Take<T> {
/// Returns the number of bytes that can be read before this instance will
/// return EOF.
///
/// # Note
///
/// This instance may reach EOF after reading fewer bytes than indicated by
/// this method if the underlying `Read` instance reaches EOF.
///
/// # Examples
///
/// ```
/// use std::io;
/// use std::io::prelude::*;
/// use std::fs::File;
///
/// # fn foo() -> io::Result<()> {
/// let f = try!(File::open("foo.txt"));
///
/// // read at most five bytes
/// let handle = f.take(5);
///
/// println!("limit: {}", handle.limit());
/// # Ok(())
/// # }
/// ```
pub fn limit(&self) -> u64 { self.limit }
}
impl<T: Read> Read for Take<T> {
fn read(&mut self, buf: &mut [u8]) -> Result<usize> {
// Don't call into inner reader at all at EOF because it may still block
if self.limit == 0 {
return Ok(0);
}
let max = cmp::min(buf.len() as u64, self.limit) as usize;
let n = self.inner.read(&mut buf[..max])?;
self.limit -= n as u64;
Ok(n)
}
}
impl<T: BufRead> BufRead for Take<T> {
fn fill_buf(&mut self) -> Result<&[u8]> {
// Don't call into inner reader at all at EOF because it may still block
if self.limit == 0 {
return Ok(&[]);
}
let buf = self.inner.fill_buf()?;
let cap = cmp::min(buf.len() as u64, self.limit) as usize;
Ok(&buf[..cap])
}
fn consume(&mut self, amt: usize) {
// Don't let callers reset the limit by passing an overlarge value
let amt = cmp::min(amt as u64, self.limit) as usize;
self.limit -= amt as u64;
self.inner.consume(amt);
}
}
fn read_one_byte(reader: &mut Read) -> Option<Result<u8>> {
let mut buf = [0];
loop {
return match reader.read(&mut buf) {
Ok(0) => None,
Ok(..) => Some(Ok(buf[0])),
Err(ref e) if e.kind() == ErrorKind::Interrupted => continue,
Err(e) => Some(Err(e)),
};
}
}
/// An iterator over `u8` values of a reader.
///
/// This struct is generally created by calling [`bytes()`][bytes] on a reader.
/// Please see the documentation of `bytes()` for more details.
///
/// [bytes]: trait.Read.html#method.bytes
pub struct Bytes<R> {
inner: R,
}
impl<R: Read> Iterator for Bytes<R> {
type Item = Result<u8>;
fn next(&mut self) -> Option<Result<u8>> {
read_one_byte(&mut self.inner)
}
}
/// An iterator over the `char`s of a reader.
///
/// This struct is generally created by calling [`chars()`][chars] on a reader.
/// Please see the documentation of `chars()` for more details.
///
/// [chars]: trait.Read.html#method.chars
pub struct Chars<R> {
inner: R,
}
/// An enumeration of possible errors that can be generated from the `Chars`
/// adapter.
#[derive(Debug)]
pub enum CharsError {
/// Variant representing that the underlying stream was read successfully
/// but it did not contain valid utf8 data.
NotUtf8,
/// Variant representing that an I/O error occurred.
Other(Error),
}
impl<R: Read> Iterator for Chars<R> {
type Item = result::Result<char, CharsError>;
fn next(&mut self) -> Option<result::Result<char, CharsError>> {
let first_byte = match read_one_byte(&mut self.inner) {
None => return None,
Some(Ok(b)) => b,
Some(Err(e)) => return Some(Err(CharsError::Other(e))),
};
let width = core_str::utf8_char_width(first_byte);
if width == 1 { return Some(Ok(first_byte as char)) }
if width == 0 { return Some(Err(CharsError::NotUtf8)) }
let mut buf = [first_byte, 0, 0, 0];
{
let mut start = 1;
while start < width {
match self.inner.read(&mut buf[start..width]) {
Ok(0) => return Some(Err(CharsError::NotUtf8)),
Ok(n) => start += n,
Err(ref e) if e.kind() == ErrorKind::Interrupted => continue,
Err(e) => return Some(Err(CharsError::Other(e))),
}
}
}
Some(match str::from_utf8(&buf[..width]).ok() {
Some(s) => Ok(s.chars().next().unwrap()),
None => Err(CharsError::NotUtf8),
})
}
}
impl std_error::Error for CharsError {
fn description(&self) -> &str {
match *self {
CharsError::NotUtf8 => "invalid utf8 encoding",
CharsError::Other(ref e) => std_error::Error::description(e),
}
}
fn cause(&self) -> Option<&std_error::Error> {
match *self {
CharsError::NotUtf8 => None,
CharsError::Other(ref e) => e.cause(),
}
}
}
impl fmt::Display for CharsError {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
match *self {
CharsError::NotUtf8 => {
"byte stream did not contain valid utf8".fmt(f)
}
CharsError::Other(ref e) => e.fmt(f),
}
}
}
/// An iterator over the contents of an instance of `BufRead` split on a
/// particular byte.
///
/// This struct is generally created by calling [`split()`][split] on a
/// `BufRead`. Please see the documentation of `split()` for more details.
///
/// [split]: trait.BufRead.html#method.split
pub struct Split<B> {
buf: B,
delim: u8,
}
impl<B: BufRead> Iterator for Split<B> {
type Item = Result<Vec<u8>>;
fn next(&mut self) -> Option<Result<Vec<u8>>> {
let mut buf = Vec::new();
match self.buf.read_until(self.delim, &mut buf) {
Ok(0) => None,
Ok(_n) => {
if buf[buf.len() - 1] == self.delim {
buf.pop();
}
Some(Ok(buf))
}
Err(e) => Some(Err(e))
}
}
}
/// An iterator over the lines of an instance of `BufRead`.
///
/// This struct is generally created by calling [`lines()`][lines] on a
/// `BufRead`. Please see the documentation of `lines()` for more details.
///
/// [lines]: trait.BufRead.html#method.lines
pub struct Lines<B> {
buf: B,
}
impl<B: BufRead> Iterator for Lines<B> {
type Item = Result<String>;
fn next(&mut self) -> Option<Result<String>> {
let mut buf = String::new();
match self.buf.read_line(&mut buf) {
Ok(0) => None,
Ok(_n) => {
if buf.ends_with("\n") {
buf.pop();
if buf.ends_with("\r") {
buf.pop();
}
}
Some(Ok(buf))
}
Err(e) => Some(Err(e))
}
}
}
#[cfg(test)]
mod tests {
use io::prelude::*;
use io;
use super::Cursor;
use super::repeat;
use test;
use collections::{Vec, String};
use collections::string::ToString;
#[test]
fn read_until() {
let mut buf = Cursor::new(&b"12"[..]);
let mut v = Vec::new();
assert_eq!(buf.read_until(b'3', &mut v).unwrap(), 2);
assert_eq!(v, b"12");
let mut buf = Cursor::new(&b"1233"[..]);
let mut v = Vec::new();
assert_eq!(buf.read_until(b'3', &mut v).unwrap(), 3);
assert_eq!(v, b"123");
v.truncate(0);
assert_eq!(buf.read_until(b'3', &mut v).unwrap(), 1);
assert_eq!(v, b"3");
v.truncate(0);
assert_eq!(buf.read_until(b'3', &mut v).unwrap(), 0);
assert_eq!(v, []);
}
#[test]
fn split() {
let buf = Cursor::new(&b"12"[..]);
let mut s = buf.split(b'3');
assert_eq!(s.next().unwrap().unwrap(), vec![b'1', b'2']);
assert!(s.next().is_none());
let buf = Cursor::new(&b"1233"[..]);
let mut s = buf.split(b'3');
assert_eq!(s.next().unwrap().unwrap(), vec![b'1', b'2']);
assert_eq!(s.next().unwrap().unwrap(), vec![]);
assert!(s.next().is_none());
}
#[test]
fn read_line() {
let mut buf = Cursor::new(&b"12"[..]);
let mut v = String::new();
assert_eq!(buf.read_line(&mut v).unwrap(), 2);
assert_eq!(v, "12");
let mut buf = Cursor::new(&b"12\n\n"[..]);
let mut v = String::new();
assert_eq!(buf.read_line(&mut v).unwrap(), 3);
assert_eq!(v, "12\n");
v.truncate(0);
assert_eq!(buf.read_line(&mut v).unwrap(), 1);
assert_eq!(v, "\n");
v.truncate(0);
assert_eq!(buf.read_line(&mut v).unwrap(), 0);
assert_eq!(v, "");
}
#[test]
fn lines() {
let buf = Cursor::new(&b"12\r"[..]);
let mut s = buf.lines();
assert_eq!(s.next().unwrap().unwrap(), "12\r".to_string());
assert!(s.next().is_none());
let buf = Cursor::new(&b"12\r\n\n"[..]);
let mut s = buf.lines();
assert_eq!(s.next().unwrap().unwrap(), "12".to_string());
assert_eq!(s.next().unwrap().unwrap(), "".to_string());
assert!(s.next().is_none());
}
#[test]
fn read_to_end() {
let mut c = Cursor::new(&b""[..]);
let mut v = Vec::new();
assert_eq!(c.read_to_end(&mut v).unwrap(), 0);
assert_eq!(v, []);
let mut c = Cursor::new(&b"1"[..]);
let mut v = Vec::new();
assert_eq!(c.read_to_end(&mut v).unwrap(), 1);
assert_eq!(v, b"1");
let cap = 1024 * 1024;
let data = (0..cap).map(|i| (i / 3) as u8).collect::<Vec<_>>();
let mut v = Vec::new();
let (a, b) = data.split_at(data.len() / 2);
assert_eq!(Cursor::new(a).read_to_end(&mut v).unwrap(), a.len());
assert_eq!(Cursor::new(b).read_to_end(&mut v).unwrap(), b.len());
assert_eq!(v, data);
}
#[test]
fn read_to_string() {
let mut c = Cursor::new(&b""[..]);
let mut v = String::new();
assert_eq!(c.read_to_string(&mut v).unwrap(), 0);
assert_eq!(v, "");
let mut c = Cursor::new(&b"1"[..]);
let mut v = String::new();
assert_eq!(c.read_to_string(&mut v).unwrap(), 1);
assert_eq!(v, "1");
let mut c = Cursor::new(&b"\xff"[..]);
let mut v = String::new();
assert!(c.read_to_string(&mut v).is_err());
}
#[test]
fn read_exact() {
let mut buf = [0; 4];
let mut c = Cursor::new(&b""[..]);
assert_eq!(c.read_exact(&mut buf).unwrap_err().kind(),
io::ErrorKind::UnexpectedEof);
let mut c = Cursor::new(&b"123"[..]).chain(Cursor::new(&b"456789"[..]));
c.read_exact(&mut buf).unwrap();
assert_eq!(&buf, b"1234");
c.read_exact(&mut buf).unwrap();
assert_eq!(&buf, b"5678");
assert_eq!(c.read_exact(&mut buf).unwrap_err().kind(),
io::ErrorKind::UnexpectedEof);
}
#[test]
fn read_exact_slice() {
let mut buf = [0; 4];
let mut c = &b""[..];
assert_eq!(c.read_exact(&mut buf).unwrap_err().kind(),
io::ErrorKind::UnexpectedEof);
let mut c = &b"123"[..];
assert_eq!(c.read_exact(&mut buf).unwrap_err().kind(),
io::ErrorKind::UnexpectedEof);
// make sure the optimized (early returning) method is being used
assert_eq!(&buf, &[0; 4]);
let mut c = &b"1234"[..];
c.read_exact(&mut buf).unwrap();
assert_eq!(&buf, b"1234");
let mut c = &b"56789"[..];
c.read_exact(&mut buf).unwrap();
assert_eq!(&buf, b"5678");
assert_eq!(c, b"9");
}
#[test]
fn take_eof() {
struct R;
impl Read for R {
fn read(&mut self, _: &mut [u8]) -> io::Result<usize> {
Err(io::Error::new(io::ErrorKind::Other, ""))
}
}
impl BufRead for R {
fn fill_buf(&mut self) -> io::Result<&[u8]> {
Err(io::Error::new(io::ErrorKind::Other, ""))
}
fn consume(&mut self, _amt: usize) { }
}
let mut buf = [0; 1];
assert_eq!(0, R.take(0).read(&mut buf).unwrap());
assert_eq!(b"", R.take(0).fill_buf().unwrap());
}
fn cmp_bufread<Br1: BufRead, Br2: BufRead>(mut br1: Br1, mut br2: Br2, exp: &[u8]) {
let mut cat = Vec::new();
loop {
let consume = {
let buf1 = br1.fill_buf().unwrap();
let buf2 = br2.fill_buf().unwrap();
let minlen = if buf1.len() < buf2.len() { buf1.len() } else { buf2.len() };
assert_eq!(buf1[..minlen], buf2[..minlen]);
cat.extend_from_slice(&buf1[..minlen]);
minlen
};
if consume == 0 {
break;
}
br1.consume(consume);
br2.consume(consume);
}
assert_eq!(br1.fill_buf().unwrap().len(), 0);
assert_eq!(br2.fill_buf().unwrap().len(), 0);
assert_eq!(&cat[..], &exp[..])
}
#[test]
fn chain_bufread() {
let testdata = b"ABCDEFGHIJKL";
let chain1 = (&testdata[..3]).chain(&testdata[3..6])
.chain(&testdata[6..9])
.chain(&testdata[9..]);
let chain2 = (&testdata[..4]).chain(&testdata[4..8])
.chain(&testdata[8..]);
cmp_bufread(chain1, chain2, &testdata[..]);
}
#[bench]
fn bench_read_to_end(b: &mut test::Bencher) {
b.iter(|| {
let mut lr = repeat(1).take(10000000);
let mut vec = Vec::with_capacity(1024);
super::read_to_end(&mut lr, &mut vec)
});
}
}