1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
//! Utilities for decoding from and encoding into bytes. //! //! This module defines zero-copy (de)serialization traits, [`ToBytes`] and [`FromBytes`], as well //! as the helper structs [`ByteWriter`] and [`ByteReader`], which wrap a `&mut [u8]` or `&[u8]` //! and offer useful utilities to read and write values. //! //! All types that end up getting transmitted over the air will want to implement [`ToBytes`] and //! [`FromBytes`]. This includes the raw PDUs sent and received on advertising and data channels, //! as well as messages used by a high-level protocol transferred over L2CAP. //! //! Also defined in this module is the [`BytesOr`] type, which can be used to store objects and //! slices of objects either as a direct reference or as a `&[u8]` that is lazily decoded. //! //! [`ToBytes`]: trait.ToBytes.html //! [`FromBytes`]: trait.FromBytes.html //! [`ByteWriter`]: struct.ByteWriter.html //! [`ByteReader`]: struct.ByteReader.html //! [`BytesOr`]: struct.BytesOr.html use { crate::Error, byteorder::{ByteOrder, LittleEndian}, core::{fmt, iter, mem}, }; /// Reference to a `T`, or to a byte slice that can be decoded as a `T`. /// /// # Motivation /// /// Many packets can contain dynamically-sized lists of objects. These packets all need to implement /// [`ToBytes`] and [`FromBytes`]. For [`FromBytes`], it is impossible to go from `&[u8]` to `&[T]`. /// /// A workaround is to just store the `&[u8]` and decode `T`s only when necessary. However, this /// isn't very type-safe and also makes it difficult to create the type when you have a list of /// `T`s, but can't easily get a `&[u8]` (such as when creating a packet to be sent out). You'd have /// to define your own byte buffer and serialize the `T`s into it, which is problematic due to the /// potentially unknown size requirement and lifetime management. /// /// A workaround around the workaround would be to use 2 types for the same packet: One storing a /// `&[u8]` and implementing [`FromBytes`] which can only do *deserialization*, and one storing a /// `&[T]` and implementing [`ToBytes`], which can only do *serialization*. This has the obvious /// drawback of essentially duplicating all packet definitions. /// /// Rubble's solution for this is `BytesOr`: It can store either an `&[u8]` or a `&T` (where `T` /// might be a slice), and always implements [`ToBytes`] and [`FromBytes`] if `T` does. Methods /// allowing access to the stored `T` (or the elements in the `&[T]` slice) will either directly /// return the value, or decode it using its [`FromBytes`] implementation. /// /// When encoding a `T`, [`BytesOr::from_ref`] can be used to store a `&T` in a `BytesOr`, which can /// then be turned into bytes via [`ToBytes`]. When decoding data, [`FromBytes`] can be used to /// create a `BytesOr` from bytes. /// /// This type can also be used in structures when storing a `T` directly is not desirable due to /// size concerns: It could be inside a rarely-encountered variant or would blow up the total size /// of the containing enum). The size of `BytesOr` is currently 2 `usize`s plus a discriminant byte, /// but could potentially be (unsafely) reduced further, should that be required. /// /// [`ToBytes`]: trait.ToBytes.html /// [`FromBytes`]: trait.FromBytes.html /// [`BytesOr::from_ref`]: #method.from_ref pub struct BytesOr<'a, T: ?Sized>(Inner<'a, T>); impl<'a, T: ?Sized> From<&'a T> for BytesOr<'a, T> { fn from(r: &'a T) -> Self { BytesOr(Inner::Or(r)) } } enum Inner<'a, T: ?Sized> { Bytes(&'a [u8]), Or(&'a T), } impl<'a, T: ?Sized> Clone for Inner<'a, T> { fn clone(&self) -> Self { match self { Inner::Bytes(b) => Inner::Bytes(b), Inner::Or(t) => Inner::Or(t), } } } impl<'a, T: ?Sized> Clone for BytesOr<'a, T> { fn clone(&self) -> Self { BytesOr(self.0) } } impl<'a, T: ?Sized> Copy for BytesOr<'a, T> {} impl<'a, T: ?Sized> Copy for Inner<'a, T> {} impl<'a, T: fmt::Debug + FromBytes<'a> + Copy> fmt::Debug for BytesOr<'a, T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { self.read().fmt(f) } } impl<'a, T: fmt::Debug + FromBytes<'a> + Copy> fmt::Debug for BytesOr<'a, [T]> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { f.debug_list().entries(self.iter()).finish() } } impl<'a, T: ?Sized> BytesOr<'a, T> { /// Creates a `BytesOr` that holds on to a `T` via reference. /// /// For creating a `BytesOr` that references a byte slice, the [`FromBytes`] impl(s) can be /// used. /// /// [`FromBytes`]: trait.FromBytes.html pub fn from_ref(value: &'a T) -> Self { BytesOr(Inner::Or(value)) } } /// Creates a `BytesOr` that stores bytes that can be decoded to a `T`. /// /// This will check that `bytes` can indeed be decoded as a `T` using its [`FromBytes`] /// implementation, and returns an error if not. /// /// The [`ByteReader`] will be advanced to point past the decoded `T` if the conversion succeeds. /// /// [`FromBytes`]: trait.FromBytes.html /// [`ByteReader`]: struct.ByteReader.html impl<'a, T: FromBytes<'a>> FromBytes<'a> for BytesOr<'a, T> { fn from_bytes(bytes: &mut ByteReader<'a>) -> Result<Self, Error> { let raw = bytes.as_raw_bytes(); T::from_bytes(bytes)?; let used = raw.len() - bytes.bytes_left(); Ok(BytesOr(Inner::Bytes(&raw[..used]))) } } /// Creates a `BytesOr` that stores bytes that can be decoded to a sequence of `T`s. /// /// This will check that `bytes` can indeed be decoded as a sequence of `T`s, and returns an error /// if not. Note that this will read *as many `T`s as possible* until the [`ByteReader`] is at its /// end of input. Any trailing data after the list of `T`s will result in an error. /// /// The [`ByteReader`] will be advanced to point past the decoded list of `T`s if the conversion /// succeeds. In that case, it will be at EOF and no more data can be read. /// /// [`ByteReader`]: struct.ByteReader.html impl<'a, T: FromBytes<'a>> FromBytes<'a> for BytesOr<'a, [T]> { fn from_bytes(bytes: &mut ByteReader<'a>) -> Result<Self, Error> { let raw = bytes.as_raw_bytes(); while !bytes.is_empty() { T::from_bytes(bytes)?; } Ok(BytesOr(Inner::Bytes(raw))) } } impl<'a, T: ToBytes + ?Sized> ToBytes for BytesOr<'a, T> { fn to_bytes(&self, buffer: &mut ByteWriter) -> Result<(), Error> { match self.0 { Inner::Bytes(b) => buffer.write_slice(b), Inner::Or(t) => t.to_bytes(buffer), } } } impl<'a, T: Copy + FromBytes<'a>> BytesOr<'a, T> { /// Reads the `T`, possibly by parsing the stored bytes. /// /// If `self` already stores a reference to a `T`, the `T` will just be copied out. If `self` /// stores a byte slice, the `T` will be parsed using its [`FromBytes`] implementation. /// /// [`FromBytes`]: trait.FromBytes.html pub fn read(&self) -> T { match self.0 { Inner::Bytes(b) => { let mut bytes = ByteReader::new(b); let t = T::from_bytes(&mut bytes).unwrap(); assert!(bytes.is_empty()); t } Inner::Or(t) => *t, } } } impl<'a, T: Copy + FromBytes<'a>> BytesOr<'a, T> { /// Returns an iterator over all `T`s stored in `self` (which is just one `T` in this case). /// /// This method exists to mirror its twin implemented for `BytesOr<'a, [T]>`. pub fn iter(&self) -> impl Iterator<Item = T> + 'a { iter::once(self.read()) } } impl<'a, T: Copy + FromBytes<'a>> BytesOr<'a, [T]> { /// Returns an iterator over all `T`s stored in `self`. /// /// The iterator will copy or decode `T`s out of `self`. pub fn iter(&self) -> impl Iterator<Item = T> + 'a { IterBytesOr { inner: *self } } } /// An iterator over values stored in a `BytesOr`. struct IterBytesOr<'a, T> { inner: BytesOr<'a, [T]>, } impl<'a, T: Copy + FromBytes<'a>> Iterator for IterBytesOr<'a, T> { type Item = T; fn next(&mut self) -> Option<Self::Item> { match &mut self.inner.0 { Inner::Bytes(b) => { if b.is_empty() { None } else { // Read a `T` and overwrite our `b` with the left-over data let mut reader = ByteReader::new(*b); let t = T::from_bytes(&mut reader).unwrap(); *b = reader.into_rest(); Some(t) } } Inner::Or(slice) => { let (first, rest) = slice.split_first()?; *slice = rest; Some(*first) } } } } /// Wrapper around a byte slice that can be used to encode data into bytes. /// /// All `write_*` methods on this type will return `Error::Eof` when the underlying buffer slice is /// full. pub struct ByteWriter<'a>(&'a mut [u8]); impl<'a> ByteWriter<'a> { /// Creates a writer that will write to `buf`. pub fn new(buf: &'a mut [u8]) -> Self { ByteWriter(buf) } /// Consumes `self` and returns the part of the contained buffer that has not yet been written /// to. pub fn into_rest(self) -> &'a mut [u8] { self.0 } /// Returns the raw buffer this `ByteWriter` would write to. /// /// Combined with `skip`, this method allows advanced operations on the underlying byte buffer. pub fn rest(&mut self) -> &mut [u8] { self.0 } /// Skips the given number of bytes in the output data without writing anything there. /// /// This is a potentially dangerous operation that should only be used when necessary (eg. when /// the skipped data will be filled in by other code). If the skipped bytes are *not* written, /// they will probably contain garbage data from an earlier use of the underlying buffer. pub fn skip(&mut self, bytes: usize) -> Result<(), Error> { if self.space_left() < bytes { Err(Error::Eof) } else { let this = mem::replace(&mut self.0, &mut []); self.0 = &mut this[bytes..]; Ok(()) } } /// Creates and returns another `ByteWriter` that can write to the next `len` Bytes in the /// buffer. /// /// `self` will be modified to point after the split-off bytes. /// /// Note that if the created `ByteWriter` is not used, the bytes will contain whatever contents /// they had before creating `self` (ie. most likely garbage data left over from earlier use). /// If you are really sure you want that, `skip` is a more explicit way of accomplishing that. #[must_use] pub fn split_off(&mut self, len: usize) -> Result<Self, Error> { if self.space_left() < len { Err(Error::Eof) } else { let this = mem::replace(&mut self.0, &mut []); let (head, tail) = this.split_at_mut(len); self.0 = tail; Ok(ByteWriter::new(head)) } } /// Splits off the next byte in the buffer. /// /// The writer will be advanced to point to the rest of the underlying buffer. /// /// This allows filling in the value of the byte later, after writing more data. /// /// For a similar, but more flexible operation, see [`split_off`]. /// /// [`split_off`]: #method.split_off pub fn split_next_mut(&mut self) -> Option<&'a mut u8> { let this = mem::replace(&mut self.0, &mut []); // Slight contortion to please the borrow checker: if this.is_empty() { self.0 = this; None } else { let (first, rest) = this.split_first_mut().unwrap(); self.0 = rest; Some(first) } } /// Returns the number of bytes that can be written to `self` until it is full. pub fn space_left(&self) -> usize { self.0.len() } /// Writes all bytes from `other` to `self`. /// /// Returns `Error::Eof` when `self` does not have enough space left to fit `other`. In that /// case, `self` will not be modified. pub fn write_slice(&mut self, other: &[u8]) -> Result<(), Error> { if self.space_left() < other.len() { Err(Error::Eof) } else { self.0[..other.len()].copy_from_slice(other); let this = mem::replace(&mut self.0, &mut []); self.0 = &mut this[other.len()..]; Ok(()) } } /// Writes a single byte to `self`. /// /// Returns `Error::Eof` when no space is left. pub fn write_u8(&mut self, byte: u8) -> Result<(), Error> { let first = self.split_next_mut().ok_or(Error::Eof)?; *first = byte; Ok(()) } /// Writes a `u16` to `self`, using Little Endian byte order. /// /// If `self` does not have enough space left, an error will be returned and no bytes will be /// written to `self`. pub fn write_u16_le(&mut self, value: u16) -> Result<(), Error> { let mut bytes = [0; 2]; LittleEndian::write_u16(&mut bytes, value); self.write_slice(&bytes) } /// Writes a `u32` to `self`, using Little Endian byte order. /// /// If `self` does not have enough space left, an error will be returned and no bytes will be /// written to `self`. pub fn write_u32_le(&mut self, value: u32) -> Result<(), Error> { let mut bytes = [0; 4]; LittleEndian::write_u32(&mut bytes, value); self.write_slice(&bytes) } /// Writes a `u64` to `self`, using Little Endian byte order. /// /// If `self` does not have enough space left, an error will be returned and no bytes will be /// written to `self`. pub fn write_u64_le(&mut self, value: u64) -> Result<(), Error> { let mut bytes = [0; 8]; LittleEndian::write_u64(&mut bytes, value); self.write_slice(&bytes) } } /// Allows reading values from a borrowed byte slice. pub struct ByteReader<'a>(&'a [u8]); impl<'a> ByteReader<'a> { /// Creates a new `ByteReader` that will read from the given byte slice. pub fn new(bytes: &'a [u8]) -> Self { ByteReader(bytes) } /// Returns a reference to the raw bytes in `self`, without advancing `self` or reading any /// data. pub fn as_raw_bytes(&self) -> &'a [u8] { self.0 } /// Consumes `self` and returns the part of the contained buffer that has not yet been read /// from. pub fn into_rest(self) -> &'a [u8] { self.0 } /// Skips the given number of bytes in the input data without inspecting them. /// /// This is a potentially dangerous operation that should only be used when the bytes really do /// not matter. pub fn skip(&mut self, bytes: usize) -> Result<(), Error> { if self.bytes_left() < bytes { Err(Error::Eof) } else { self.0 = &self.0[bytes..]; Ok(()) } } /// Creates and returns another `ByteReader` that will read from the next `len` Bytes in the /// buffer. /// /// `self` will be modified to point after the split-off bytes, and will continue reading from /// there. /// /// Note that if the created `ByteReader` is not used, the bytes will be ignored. If you are /// really sure you want that, `skip` is a more explicit way of accomplishing that. #[must_use] pub fn split_off(&mut self, len: usize) -> Result<Self, Error> { if self.bytes_left() < len { Err(Error::Eof) } else { let (head, tail) = (&self.0[..len], &self.0[len..]); self.0 = tail; Ok(ByteReader::new(head)) } } /// Returns the number of bytes that can still be read from `self`. pub fn bytes_left(&self) -> usize { self.0.len() } /// Returns whether `self` is at the end of the underlying buffer (EOF). /// /// If this returns `true`, no data can be read from `self` anymore. pub fn is_empty(&self) -> bool { self.0.is_empty() } /// Reads a byte slice of length `len` from `self`. /// /// If `self` contains less than `len` bytes, `Error::Eof` will be returned and `self` will not /// be modified. pub fn read_slice(&mut self, len: usize) -> Result<&'a [u8], Error> { if self.bytes_left() < len { Err(Error::Eof) } else { let slice = &self.0[..len]; self.0 = &self.0[len..]; Ok(slice) } } /// Reads a byte-array-like type `S` from `self`. /// /// `S` must implement `Default` and `AsMut<[u8]>`, which allows using small arrays up to 32 /// bytes as well as datastructures from `alloc` (eg. `Box<[u8]>` or `Vec<u8>`). pub fn read_array<S>(&mut self) -> Result<S, Error> where S: Default + AsMut<[u8]>, { let mut buf = S::default(); let slice = buf.as_mut(); if self.bytes_left() < slice.len() { return Err(Error::Eof); } slice.copy_from_slice(&self.0[..slice.len()]); self.0 = &self.0[slice.len()..]; Ok(buf) } /// Reads the remaining bytes from `self`. pub fn read_rest(&mut self) -> &'a [u8] { let rest = self.0; self.0 = &[]; rest } /// Reads a single byte from `self`. /// /// Returns `Error::Eof` when `self` is empty. pub fn read_u8(&mut self) -> Result<u8, Error> { Ok(self.read_array::<[u8; 1]>()?[0]) } /// Reads a `u16` from `self`, using Little Endian byte order. pub fn read_u16_le(&mut self) -> Result<u16, Error> { let arr = self.read_array::<[u8; 2]>()?; Ok(LittleEndian::read_u16(&arr)) } /// Reads a `u32` from `self`, using Little Endian byte order. pub fn read_u32_le(&mut self) -> Result<u32, Error> { let arr = self.read_array::<[u8; 4]>()?; Ok(LittleEndian::read_u32(&arr)) } /// Reads a `u64` from `self`, using Little Endian byte order. pub fn read_u64_le(&mut self) -> Result<u64, Error> { let arr = self.read_array::<[u8; 8]>()?; Ok(LittleEndian::read_u64(&arr)) } } /// Trait for encoding a value into a byte buffer. pub trait ToBytes { /// Converts `self` to bytes and writes them into `writer`, advancing `writer` to point past the /// encoded value. /// /// If `writer` does not contain enough space, an error will be returned and the state of the /// buffer is unspecified (eg. `self` may be partially written into `writer`). fn to_bytes(&self, writer: &mut ByteWriter) -> Result<(), Error>; } /// Trait for decoding values from a byte slice. pub trait FromBytes<'a>: Sized { /// Decode a `Self` from a byte slice, advancing `bytes` to point past the data that was read. /// /// If `bytes` contains data not valid for the target type, or contains an insufficient number /// of bytes, an error will be returned and the state of `bytes` is unspecified (it can point to /// arbitrary data). fn from_bytes(bytes: &mut ByteReader<'a>) -> Result<Self, Error>; } impl<T: ToBytes> ToBytes for [T] { fn to_bytes(&self, writer: &mut ByteWriter) -> Result<(), Error> { for t in self { t.to_bytes(writer)?; } Ok(()) } } impl<'a> ToBytes for &'a [u8] { fn to_bytes(&self, writer: &mut ByteWriter) -> Result<(), Error> { writer.write_slice(*self) } } impl<'a> FromBytes<'a> for &'a [u8] { fn from_bytes(bytes: &mut ByteReader<'a>) -> Result<Self, Error> { Ok(bytes.read_rest()) } } impl<'a> FromBytes<'a> for u8 { fn from_bytes(bytes: &mut ByteReader<'a>) -> Result<Self, Error> { bytes.read_u8() } }