src/libstd/ffi/c_str.rs

   1 // Copyright 2012 The Rust Project Developers. See the COPYRIGHT
   2 // file at the top-level directory of this distribution and at
   3 // http://rust-lang.org/COPYRIGHT.
   4 //
   5 // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
   6 // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
   7 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
   8 // option. This file may not be copied, modified, or distributed
   9 // except according to those terms.
  10
  11 use borrow::{Cow, ToOwned};
  12 use boxed::Box;
  13 use clone::Clone;
  14 use convert::{Into, From};
  15 use cmp::{PartialEq, Eq, PartialOrd, Ord, Ordering};
  16 use error::Error;
  17 use fmt;
  18 use io;
  19 use iter::Iterator;
  20 use libc;
  21 use mem;
  22 use ops::Deref;
  23 use option::Option::{self, Some, None};
  24 use result::Result::{self, Ok, Err};
  25 use slice;
  26 use str;
  27 use string::String;
  28 use vec::Vec;
  29
  30 /// A type representing an owned C-compatible string
  31 ///
  32 /// This type serves the primary purpose of being able to safely generate a
  33 /// C-compatible string from a Rust byte slice or vector. An instance of this
  34 /// type is a static guarantee that the underlying bytes contain no interior 0
  35 /// bytes and the final byte is 0.
  36 ///
  37 /// A `CString` is created from either a byte slice or a byte vector. After
  38 /// being created, a `CString` predominately inherits all of its methods from
  39 /// the `Deref` implementation to `[libc::c_char]`. Note that the underlying
  40 /// array is represented as an array of `libc::c_char` as opposed to `u8`. A
  41 /// `u8` slice can be obtained with the `as_bytes` method.  Slices produced from
  42 /// a `CString` do *not* contain the trailing nul terminator unless otherwise
  43 /// specified.
  44 ///
  45 /// # Examples
  46 ///
  47 /// ```no_run
  48 /// # #![feature(libc)]
  49 /// # extern crate libc;
  50 /// # fn main() {
  51 /// use std::ffi::CString;
  52 /// use libc;
  53 ///
  54 /// extern {
  55 ///     fn my_printer(s: *const libc::c_char);
  56 /// }
  57 ///
  58 /// let c_to_print = CString::new("Hello, world!").unwrap();
  59 /// unsafe {
  60 ///     my_printer(c_to_print.as_ptr());
  61 /// }
  62 /// # }
  63 /// ```
  64 #[derive(PartialEq, PartialOrd, Eq, Ord, Hash)]
  65 #[stable(feature = "rust1", since = "1.0.0")]
  66 pub struct CString {
  67     inner: Box<[u8]>,
  68 }
  69
  70 /// Representation of a borrowed C string.
  71 ///
  72 /// This dynamically sized type is only safely constructed via a borrowed
  73 /// version of an instance of `CString`. This type can be constructed from a raw
  74 /// C string as well and represents a C string borrowed from another location.
  75 ///
  76 /// Note that this structure is **not** `repr(C)` and is not recommended to be
  77 /// placed in the signatures of FFI functions. Instead safe wrappers of FFI
  78 /// functions may leverage the unsafe `from_ptr` constructor to provide a safe
  79 /// interface to other consumers.
  80 ///
  81 /// # Examples
  82 ///
  83 /// Inspecting a foreign C string
  84 ///
  85 /// ```no_run
  86 /// # #![feature(libc)]
  87 /// extern crate libc;
  88 /// use std::ffi::CStr;
  89 ///
  90 /// extern { fn my_string() -> *const libc::c_char; }
  91 ///
  92 /// fn main() {
  93 ///     unsafe {
  94 ///         let slice = CStr::from_ptr(my_string());
  95 ///         println!("string length: {}", slice.to_bytes().len());
  96 ///     }
  97 /// }
  98 /// ```
  99 ///
 100 /// Passing a Rust-originating C string
 101 ///
 102 /// ```no_run
 103 /// # #![feature(libc)]
 104 /// extern crate libc;
 105 /// use std::ffi::{CString, CStr};
 106 ///
 107 /// fn work(data: &CStr) {
 108 ///     extern { fn work_with(data: *const libc::c_char); }
 109 ///
 110 ///     unsafe { work_with(data.as_ptr()) }
 111 /// }
 112 ///
 113 /// fn main() {
 114 ///     let s = CString::new("data data data data").unwrap();
 115 ///     work(&s);
 116 /// }
 117 /// ```
 118 ///
 119 /// Converting a foreign C string into a Rust `String`
 120 ///
 121 /// ```no_run
 122 /// # #![feature(libc,cstr_to_str)]
 123 /// extern crate libc;
 124 /// use std::ffi::CStr;
 125 ///
 126 /// extern { fn my_string() -> *const libc::c_char; }
 127 ///
 128 /// fn my_string_safe() -> String {
 129 ///     unsafe {
 130 ///         CStr::from_ptr(my_string()).to_string_lossy().into_owned()
 131 ///     }
 132 /// }
 133 ///
 134 /// fn main() {
 135 ///     println!("string: {}", my_string_safe());
 136 /// }
 137 /// ```
 138 #[derive(Hash)]
 139 #[stable(feature = "rust1", since = "1.0.0")]
 140 pub struct CStr {
 141     // FIXME: this should not be represented with a DST slice but rather with
 142     //        just a raw `libc::c_char` along with some form of marker to make
 143     //        this an unsized type. Essentially `sizeof(&CStr)` should be the
 144     //        same as `sizeof(&c_char)` but `CStr` should be an unsized type.
 145     inner: [libc::c_char]
 146 }
 147
 148 /// An error returned from `CString::new` to indicate that a nul byte was found
 149 /// in the vector provided.
 150 #[derive(Clone, PartialEq, Debug)]
 151 #[stable(feature = "rust1", since = "1.0.0")]
 152 pub struct NulError(usize, Vec<u8>);
 153
 154 impl CString {
 155     /// Creates a new C-compatible string from a container of bytes.
 156     ///
 157     /// This method will consume the provided data and use the underlying bytes
 158     /// to construct a new string, ensuring that there is a trailing 0 byte.
 159     ///
 160     /// # Examples
 161     ///
 162     /// ```no_run
 163     /// # #![feature(libc)]
 164     /// extern crate libc;
 165     /// use std::ffi::CString;
 166     ///
 167     /// extern { fn puts(s: *const libc::c_char); }
 168     ///
 169     /// fn main() {
 170     ///     let to_print = CString::new("Hello!").unwrap();
 171     ///     unsafe {
 172     ///         puts(to_print.as_ptr());
 173     ///     }
 174     /// }
 175     /// ```
 176     ///
 177     /// # Errors
 178     ///
 179     /// This function will return an error if the bytes yielded contain an
 180     /// internal 0 byte. The error returned will contain the bytes as well as
 181     /// the position of the nul byte.
 182     #[stable(feature = "rust1", since = "1.0.0")]
 183     pub fn new<T: Into<Vec<u8>>>(t: T) -> Result<CString, NulError> {
 184         let bytes = t.into();
 185         match bytes.iter().position(|x| *x == 0) {
 186             Some(i) => Err(NulError(i, bytes)),
 187             None => Ok(unsafe { CString::from_vec_unchecked(bytes) }),
 188         }
 189     }
 190
 191     /// Creates a C-compatible string from a byte vector without checking for
 192     /// interior 0 bytes.
 193     ///
 194     /// This method is equivalent to `new` except that no runtime assertion
 195     /// is made that `v` contains no 0 bytes, and it requires an actual
 196     /// byte vector, not anything that can be converted to one with Into.
 197     #[stable(feature = "rust1", since = "1.0.0")]
 198     pub unsafe fn from_vec_unchecked(mut v: Vec<u8>) -> CString {
 199         v.push(0);
 200         CString { inner: v.into_boxed_slice() }
 201     }
 202
 203     /// Retakes ownership of a CString that was transferred to C.
 204     ///
 205     /// The only appropriate argument is a pointer obtained by calling
 206     /// `into_ptr`. The length of the string will be recalculated
 207     /// using the pointer.
 208     #[unstable(feature = "cstr_memory", reason = "recently added")]
 209     // NB: may want to be called from_raw, needs to consider CStr::from_ptr,
 210     //     Box::from_raw (or whatever it's currently called), and
 211     //     slice::from_raw_parts
 212     pub unsafe fn from_ptr(ptr: *const libc::c_char) -> CString {
 213         let len = libc::strlen(ptr) + 1; // Including the NUL byte
 214         let slice = slice::from_raw_parts(ptr, len as usize);
 215         CString { inner: mem::transmute(slice) }
 216     }
 217
 218     /// Transfers ownership of the string to a C caller.
 219     ///
 220     /// The pointer must be returned to Rust and reconstituted using
 221     /// `from_ptr` to be properly deallocated. Specifically, one
 222     /// should *not* use the standard C `free` function to deallocate
 223     /// this string.
 224     ///
 225     /// Failure to call `from_ptr` will lead to a memory leak.
 226     #[unstable(feature = "cstr_memory", reason = "recently added")]
 227     // NB: may want to be called into_raw, see comments on from_ptr
 228     pub fn into_ptr(self) -> *const libc::c_char {
 229         // It is important that the bytes be sized to fit - we need
 230         // the capacity to be determinable from the string length, and
 231         // shrinking to fit is the only way to be sure.
 232         Box::into_raw(self.inner) as *const libc::c_char
 233     }
 234
 235     /// Returns the contents of this `CString` as a slice of bytes.
 236     ///
 237     /// The returned slice does **not** contain the trailing nul separator and
 238     /// it is guaranteed to not have any interior nul bytes.
 239     #[stable(feature = "rust1", since = "1.0.0")]
 240     pub fn as_bytes(&self) -> &[u8] {
 241         &self.inner[..self.inner.len() - 1]
 242     }
 243
 244     /// Equivalent to the `as_bytes` function except that the returned slice
 245     /// includes the trailing nul byte.
 246     #[stable(feature = "rust1", since = "1.0.0")]
 247     pub fn as_bytes_with_nul(&self) -> &[u8] {
 248         &self.inner
 249     }
 250 }
 251
 252 #[stable(feature = "rust1", since = "1.0.0")]
 253 impl Clone for CString {
 254     fn clone(&self) -> Self {
 255         CString { inner: self.inner.to_owned().into_boxed_slice() }
 256     }
 257 }
 258
 259 #[stable(feature = "rust1", since = "1.0.0")]
 260 impl Deref for CString {
 261     type Target = CStr;
 262
 263     fn deref(&self) -> &CStr {
 264         unsafe { mem::transmute(self.as_bytes_with_nul()) }
 265     }
 266 }
 267
 268 #[stable(feature = "rust1", since = "1.0.0")]
 269 impl fmt::Debug for CString {
 270     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
 271         fmt::Debug::fmt(&String::from_utf8_lossy(self.as_bytes()), f)
 272     }
 273 }
 274
 275 impl NulError {
 276     /// Returns the position of the nul byte in the slice that was provided to
 277     /// `CString::new`.
 278     #[stable(feature = "rust1", since = "1.0.0")]
 279     pub fn nul_position(&self) -> usize { self.0 }
 280
 281     /// Consumes this error, returning the underlying vector of bytes which
 282     /// generated the error in the first place.
 283     #[stable(feature = "rust1", since = "1.0.0")]
 284     pub fn into_vec(self) -> Vec<u8> { self.1 }
 285 }
 286
 287 #[stable(feature = "rust1", since = "1.0.0")]
 288 impl Error for NulError {
 289     fn description(&self) -> &str { "nul byte found in data" }
 290 }
 291
 292 #[stable(feature = "rust1", since = "1.0.0")]
 293 impl fmt::Display for NulError {
 294     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
 295         write!(f, "nul byte found in provided data at position: {}", self.0)
 296     }
 297 }
 298
 299 #[stable(feature = "rust1", since = "1.0.0")]
 300 impl From<NulError> for io::Error {
 301     fn from(_: NulError) -> io::Error {
 302         io::Error::new(io::ErrorKind::InvalidInput,
 303                        "data provided contains a nul byte")
 304     }
 305 }
 306
 307 impl CStr {
 308     /// Casts a raw C string to a safe C string wrapper.
 309     ///
 310     /// This function will cast the provided `ptr` to the `CStr` wrapper which
 311     /// allows inspection and interoperation of non-owned C strings. This method
 312     /// is unsafe for a number of reasons:
 313     ///
 314     /// * There is no guarantee to the validity of `ptr`
 315     /// * The returned lifetime is not guaranteed to be the actual lifetime of
 316     ///   `ptr`
 317     /// * There is no guarantee that the memory pointed to by `ptr` contains a
 318     ///   valid nul terminator byte at the end of the string.
 319     ///
 320     /// > **Note**: This operation is intended to be a 0-cost cast but it is
 321     /// > currently implemented with an up-front calculation of the length of
 322     /// > the string. This is not guaranteed to always be the case.
 323     ///
 324     /// # Examples
 325     ///
 326     /// ```no_run
 327     /// # #![feature(libc)]
 328     /// # extern crate libc;
 329     /// # fn main() {
 330     /// use std::ffi::CStr;
 331     /// use std::str;
 332     /// use libc;
 333     ///
 334     /// extern {
 335     ///     fn my_string() -> *const libc::c_char;
 336     /// }
 337     ///
 338     /// unsafe {
 339     ///     let slice = CStr::from_ptr(my_string());
 340     ///     println!("string returned: {}",
 341     ///              str::from_utf8(slice.to_bytes()).unwrap());
 342     /// }
 343     /// # }
 344     /// ```
 345     #[stable(feature = "rust1", since = "1.0.0")]
 346     pub unsafe fn from_ptr<'a>(ptr: *const libc::c_char) -> &'a CStr {
 347         let len = libc::strlen(ptr);
 348         mem::transmute(slice::from_raw_parts(ptr, len as usize + 1))
 349     }
 350
 351     /// Returns the inner pointer to this C string.
 352     ///
 353     /// The returned pointer will be valid for as long as `self` is and points
 354     /// to a contiguous region of memory terminated with a 0 byte to represent
 355     /// the end of the string.
 356     #[stable(feature = "rust1", since = "1.0.0")]
 357     pub fn as_ptr(&self) -> *const libc::c_char {
 358         self.inner.as_ptr()
 359     }
 360
 361     /// Converts this C string to a byte slice.
 362     ///
 363     /// This function will calculate the length of this string (which normally
 364     /// requires a linear amount of work to be done) and then return the
 365     /// resulting slice of `u8` elements.
 366     ///
 367     /// The returned slice will **not** contain the trailing nul that this C
 368     /// string has.
 369     ///
 370     /// > **Note**: This method is currently implemented as a 0-cost cast, but
 371     /// > it is planned to alter its definition in the future to perform the
 372     /// > length calculation whenever this method is called.
 373     #[stable(feature = "rust1", since = "1.0.0")]
 374     pub fn to_bytes(&self) -> &[u8] {
 375         let bytes = self.to_bytes_with_nul();
 376         &bytes[..bytes.len() - 1]
 377     }
 378
 379     /// Converts this C string to a byte slice containing the trailing 0 byte.
 380     ///
 381     /// This function is the equivalent of `to_bytes` except that it will retain
 382     /// the trailing nul instead of chopping it off.
 383     ///
 384     /// > **Note**: This method is currently implemented as a 0-cost cast, but
 385     /// > it is planned to alter its definition in the future to perform the
 386     /// > length calculation whenever this method is called.
 387     #[stable(feature = "rust1", since = "1.0.0")]
 388     pub fn to_bytes_with_nul(&self) -> &[u8] {
 389         unsafe { mem::transmute::<&[libc::c_char], &[u8]>(&self.inner) }
 390     }
 391
 392     /// Yields a `&str` slice if the `CStr` contains valid UTF-8.
 393     ///
 394     /// This function will calculate the length of this string and check for
 395     /// UTF-8 validity, and then return the `&str` if it's valid.
 396     ///
 397     /// > **Note**: This method is currently implemented to check for validity
 398     /// > after a 0-cost cast, but it is planned to alter its definition in the
 399     /// > future to perform the length calculation in addition to the UTF-8
 400     /// > check whenever this method is called.
 401     #[unstable(feature = "cstr_to_str", reason = "recently added")]
 402     pub fn to_str(&self) -> Result<&str, str::Utf8Error> {
 403         // NB: When CStr is changed to perform the length check in .to_bytes() instead of in
 404         // from_ptr(), it may be worth considering if this should be rewritten to do the UTF-8
 405         // check inline with the length calculation instead of doing it afterwards.
 406         str::from_utf8(self.to_bytes())
 407     }
 408
 409     /// Converts a `CStr` into a `Cow<str>`.
 410     ///
 411     /// This function will calculate the length of this string (which normally
 412     /// requires a linear amount of work to be done) and then return the
 413     /// resulting slice as a `Cow<str>`, replacing any invalid UTF-8 sequences
 414     /// with `U+FFFD REPLACEMENT CHARACTER`.
 415     ///
 416     /// > **Note**: This method is currently implemented to check for validity
 417     /// > after a 0-cost cast, but it is planned to alter its definition in the
 418     /// > future to perform the length calculation in addition to the UTF-8
 419     /// > check whenever this method is called.
 420     #[unstable(feature = "cstr_to_str", reason = "recently added")]
 421     pub fn to_string_lossy(&self) -> Cow<str> {
 422         String::from_utf8_lossy(self.to_bytes())
 423     }
 424 }
 425
 426 #[stable(feature = "rust1", since = "1.0.0")]
 427 impl PartialEq for CStr {
 428     fn eq(&self, other: &CStr) -> bool {
 429         self.to_bytes().eq(other.to_bytes())
 430     }
 431 }
 432 #[stable(feature = "rust1", since = "1.0.0")]
 433 impl Eq for CStr {}
 434 #[stable(feature = "rust1", since = "1.0.0")]
 435 impl PartialOrd for CStr {
 436     fn partial_cmp(&self, other: &CStr) -> Option<Ordering> {
 437         self.to_bytes().partial_cmp(&other.to_bytes())
 438     }
 439 }
 440 #[stable(feature = "rust1", since = "1.0.0")]
 441 impl Ord for CStr {
 442     fn cmp(&self, other: &CStr) -> Ordering {
 443         self.to_bytes().cmp(&other.to_bytes())
 444     }
 445 }
 446
 447 #[cfg(test)]
 448 mod tests {
 449     use prelude::v1::*;
 450     use super::*;
 451     use libc;
 452     use borrow::Cow::{Borrowed, Owned};
 453
 454     #[test]
 455     fn c_to_rust() {
 456         let data = b"123\0";
 457         let ptr = data.as_ptr() as *const libc::c_char;
 458         unsafe {
 459             assert_eq!(CStr::from_ptr(ptr).to_bytes(), b"123");
 460             assert_eq!(CStr::from_ptr(ptr).to_bytes_with_nul(), b"123\0");
 461         }
 462     }
 463
 464     #[test]
 465     fn simple() {
 466         let s = CString::new("1234").unwrap();
 467         assert_eq!(s.as_bytes(), b"1234");
 468         assert_eq!(s.as_bytes_with_nul(), b"1234\0");
 469     }
 470
 471     #[test]
 472     fn build_with_zero1() {
 473         assert!(CString::new(&b"\0"[..]).is_err());
 474     }
 475     #[test]
 476     fn build_with_zero2() {
 477         assert!(CString::new(vec![0]).is_err());
 478     }
 479
 480     #[test]
 481     fn build_with_zero3() {
 482         unsafe {
 483             let s = CString::from_vec_unchecked(vec![0]);
 484             assert_eq!(s.as_bytes(), b"\0");
 485         }
 486     }
 487
 488     #[test]
 489     fn formatted() {
 490         let s = CString::new(&b"12"[..]).unwrap();
 491         assert_eq!(format!("{:?}", s), "\"12\"");
 492     }
 493
 494     #[test]
 495     fn borrowed() {
 496         unsafe {
 497             let s = CStr::from_ptr(b"12\0".as_ptr() as *const _);
 498             assert_eq!(s.to_bytes(), b"12");
 499             assert_eq!(s.to_bytes_with_nul(), b"12\0");
 500         }
 501     }
 502
 503     #[test]
 504     fn to_str() {
 505         let data = b"123\xE2\x80\xA6\0";
 506         let ptr = data.as_ptr() as *const libc::c_char;
 507         unsafe {
 508             assert_eq!(CStr::from_ptr(ptr).to_str(), Ok("123…"));
 509             assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Borrowed("123…"));
 510         }
 511         let data = b"123\xE2\0";
 512         let ptr = data.as_ptr() as *const libc::c_char;
 513         unsafe {
 514             assert!(CStr::from_ptr(ptr).to_str().is_err());
 515             assert_eq!(CStr::from_ptr(ptr).to_string_lossy(), Owned::<str>(format!("123\u{FFFD}")));
 516         }
 517     }
 518 }