| 1 |
/* |
| 2 |
* Copyright (c) 2023. |
| 3 |
* |
| 4 |
* This software is free software; |
| 5 |
* |
| 6 |
* You can redistribute it or modify it under terms of the MIT, Apache License or Zlib license |
| 7 |
*/ |
| 8 |
|
| 9 |
use core::cmp::min; |
| 10 |
|
| 11 |
use crate::bytestream::traits::ZReaderTrait; |
| 12 |
|
| 13 |
const ERROR_MSG: &str = "No more bytes"; |
| 14 |
|
| 15 |
/// An encapsulation of a byte stream reader |
| 16 |
/// |
| 17 |
/// This provides an interface similar to [std::io::Cursor] but |
| 18 |
/// it provides fine grained options for reading different integer data types from |
| 19 |
/// the underlying buffer. |
| 20 |
/// |
| 21 |
/// There are two variants mainly error and non error variants, |
| 22 |
/// the error variants are useful for cases where you need bytes |
| 23 |
/// from the underlying stream, and cannot do with zero result. |
| 24 |
/// the non error variants are useful when you may have proved data already exists |
| 25 |
/// eg by using [`has`] method or you are okay with returning zero if the underlying |
| 26 |
/// buffer has been completely read. |
| 27 |
/// |
| 28 |
/// [std::io::Cursor]: https://doc.rust-lang.org/std/io/struct.Cursor.html |
| 29 |
/// [`has`]: Self::has |
| 30 |
pub struct ZByteReader<T: ZReaderTrait> { |
| 31 |
/// Data stream |
| 32 |
stream: T, |
| 33 |
position: usize |
| 34 |
} |
| 35 |
|
| 36 |
enum Mode { |
| 37 |
// Big endian |
| 38 |
BE, |
| 39 |
// Little Endian |
| 40 |
LE |
| 41 |
} |
| 42 |
#[cfg(feature = "std")] |
| 43 |
impl<T: ZReaderTrait> std::io::Read for ZByteReader<T> { |
| 44 |
fn read(&mut self, buf: &mut [u8]) -> std::io::Result<usize> { |
| 45 |
Ok(self.read(buf).unwrap()) |
| 46 |
} |
| 47 |
} |
| 48 |
|
| 49 |
impl<T: ZReaderTrait> ZByteReader<T> { |
| 50 |
/// Create a new instance of the byte stream |
| 51 |
/// |
| 52 |
/// Bytes will be read from the start of `buf`. |
| 53 |
/// |
| 54 |
/// `buf` is expected to live as long as this and |
| 55 |
/// all references to it live |
| 56 |
/// |
| 57 |
/// # Returns |
| 58 |
/// A byte reader which will pull bits from bye |
| 59 |
pub const fn new(buf: T) -> ZByteReader<T> { |
| 60 |
ZByteReader { |
| 61 |
stream: buf, |
| 62 |
position: 0 |
| 63 |
} |
| 64 |
} |
| 65 |
/// Destroy this reader returning |
| 66 |
/// the underlying source of the bytes |
| 67 |
/// from which we were decoding |
| 68 |
pub fn consume(self) -> T { |
| 69 |
self.stream |
| 70 |
} |
| 71 |
/// Skip `num` bytes ahead of the stream. |
| 72 |
/// |
| 73 |
/// This bumps up the internal cursor wit a wrapping addition |
| 74 |
/// The bytes between current position and `num` will be skipped |
| 75 |
/// |
| 76 |
/// # Arguments |
| 77 |
/// `num`: How many bytes to skip |
| 78 |
/// |
| 79 |
/// # Note |
| 80 |
/// This does not consider length of the buffer, so skipping more bytes |
| 81 |
/// than possible and then reading bytes will return an error if using error variants |
| 82 |
/// or zero if using non-error variants |
| 83 |
/// |
| 84 |
/// # Example |
| 85 |
/// ``` |
| 86 |
/// use zune_core::bytestream::ZByteReader; |
| 87 |
/// let zero_to_hundred:Vec<u8> = (0..100).collect(); |
| 88 |
/// let mut stream = ZByteReader::new(&zero_to_hundred); |
| 89 |
/// // skip 37 bytes |
| 90 |
/// stream.skip(37); |
| 91 |
/// |
| 92 |
/// assert_eq!(stream.get_u8(),37); |
| 93 |
/// ``` |
| 94 |
/// |
| 95 |
/// See [`rewind`](ZByteReader::rewind) for moving the internal cursor back |
| 96 |
pub fn skip(&mut self, num: usize) { |
| 97 |
// Can this overflow ?? |
| 98 |
self.position = self.position.wrapping_add(num); |
| 99 |
} |
| 100 |
/// Undo a buffer read by moving the position pointer `num` |
| 101 |
/// bytes behind. |
| 102 |
/// |
| 103 |
/// This operation will saturate at zero |
| 104 |
pub fn rewind(&mut self, num: usize) { |
| 105 |
self.position = self.position.saturating_sub(num); |
| 106 |
} |
| 107 |
|
| 108 |
/// Return whether the underlying buffer |
| 109 |
/// has `num` bytes available for reading |
| 110 |
/// |
| 111 |
/// # Example |
| 112 |
/// |
| 113 |
/// ``` |
| 114 |
/// use zune_core::bytestream::ZByteReader; |
| 115 |
/// let data = [0_u8;120]; |
| 116 |
/// let reader = ZByteReader::new(data.as_slice()); |
| 117 |
/// assert!(reader.has(3)); |
| 118 |
/// assert!(!reader.has(121)); |
| 119 |
/// ``` |
| 120 |
#[inline] |
| 121 |
pub fn has(&self, num: usize) -> bool { |
| 122 |
self.position.saturating_add(num) <= self.stream.get_len() |
| 123 |
} |
| 124 |
/// Get number of bytes available in the stream |
| 125 |
#[inline] |
| 126 |
pub fn get_bytes_left(&self) -> usize { |
| 127 |
// Must be saturating to prevent underflow |
| 128 |
self.stream.get_len().saturating_sub(self.position) |
| 129 |
} |
| 130 |
/// Get length of the underlying buffer. |
| 131 |
/// |
| 132 |
/// To get the number of bytes left in the buffer, |
| 133 |
/// use [remaining] method |
| 134 |
/// |
| 135 |
/// [remaining]: Self::remaining |
| 136 |
#[inline] |
| 137 |
pub fn len(&self) -> usize { |
| 138 |
self.stream.get_len() |
| 139 |
} |
| 140 |
/// Return true if the underlying buffer stream is empty |
| 141 |
#[inline] |
| 142 |
pub fn is_empty(&self) -> bool { |
| 143 |
self.stream.get_len() == 0 |
| 144 |
} |
| 145 |
/// Get current position of the buffer. |
| 146 |
#[inline] |
| 147 |
pub const fn get_position(&self) -> usize { |
| 148 |
self.position |
| 149 |
} |
| 150 |
/// Return true whether or not we read to the end of the |
| 151 |
/// buffer and have no more bytes left. |
| 152 |
#[inline] |
| 153 |
pub fn eof(&self) -> bool { |
| 154 |
self.position >= self.len() |
| 155 |
} |
| 156 |
/// Get number of bytes unread inside this |
| 157 |
/// stream. |
| 158 |
/// |
| 159 |
/// To get the length of the underlying stream, |
| 160 |
/// use [len] method |
| 161 |
/// |
| 162 |
/// [len]: Self::len() |
| 163 |
#[inline] |
| 164 |
pub fn remaining(&self) -> usize { |
| 165 |
self.stream.get_len().saturating_sub(self.position) |
| 166 |
} |
| 167 |
/// Get a part of the bytestream as a reference. |
| 168 |
/// |
| 169 |
/// This increments the position to point past the bytestream |
| 170 |
/// if position+num is in bounds |
| 171 |
pub fn get(&mut self, num: usize) -> Result<&[u8], &'static str> { |
| 172 |
match self.stream.get_slice(self.position..self.position + num) { |
| 173 |
Some(bytes) => { |
| 174 |
self.position += num; |
| 175 |
Ok(bytes) |
| 176 |
} |
| 177 |
None => Err(ERROR_MSG) |
| 178 |
} |
| 179 |
} |
| 180 |
/// Look ahead position bytes and return a reference |
| 181 |
/// to num_bytes from that position, or an error if the |
| 182 |
/// peek would be out of bounds. |
| 183 |
/// |
| 184 |
/// This doesn't increment the position, bytes would have to be discarded |
| 185 |
/// at a later point. |
| 186 |
#[inline] |
| 187 |
pub fn peek_at(&self, position: usize, num_bytes: usize) -> Result<&[u8], &'static str> { |
| 188 |
let start = self.position + position; |
| 189 |
let end = self.position + position + num_bytes; |
| 190 |
|
| 191 |
match self.stream.get_slice(start..end) { |
| 192 |
Some(bytes) => Ok(bytes), |
| 193 |
None => Err(ERROR_MSG) |
| 194 |
} |
| 195 |
} |
| 196 |
/// Get a fixed amount of bytes or return an error if we cant |
| 197 |
/// satisfy the read |
| 198 |
/// |
| 199 |
/// This should be combined with [`has`] since if there are no |
| 200 |
/// more bytes you get an error. |
| 201 |
/// |
| 202 |
/// But it's useful for cases where you expect bytes but they are not present |
| 203 |
/// |
| 204 |
/// For the zero variant see, [`get_fixed_bytes_or_zero`] |
| 205 |
/// |
| 206 |
/// # Example |
| 207 |
/// ```rust |
| 208 |
/// use zune_core::bytestream::ZByteReader; |
| 209 |
/// let mut stream = ZByteReader::new([0x0,0x5,0x3,0x2].as_slice()); |
| 210 |
/// let first_bytes = stream.get_fixed_bytes_or_err::<10>(); // not enough bytes |
| 211 |
/// assert!(first_bytes.is_err()); |
| 212 |
/// ``` |
| 213 |
/// |
| 214 |
/// [`has`]:Self::has |
| 215 |
/// [`get_fixed_bytes_or_zero`]: Self::get_fixed_bytes_or_zero |
| 216 |
#[inline] |
| 217 |
pub fn get_fixed_bytes_or_err<const N: usize>(&mut self) -> Result<[u8; N], &'static str> { |
| 218 |
let mut byte_store: [u8; N] = [0; N]; |
| 219 |
|
| 220 |
match self.stream.get_slice(self.position..self.position + N) { |
| 221 |
Some(bytes) => { |
| 222 |
self.position += N; |
| 223 |
byte_store.copy_from_slice(bytes); |
| 224 |
|
| 225 |
Ok(byte_store) |
| 226 |
} |
| 227 |
None => Err(ERROR_MSG) |
| 228 |
} |
| 229 |
} |
| 230 |
|
| 231 |
/// Get a fixed amount of bytes or return a zero array size |
| 232 |
/// if we can't satisfy the read |
| 233 |
/// |
| 234 |
/// This should be combined with [`has`] since if there are no |
| 235 |
/// more bytes you get a zero initialized array |
| 236 |
/// |
| 237 |
/// For the error variant see, [`get_fixed_bytes_or_err`] |
| 238 |
/// |
| 239 |
/// # Example |
| 240 |
/// ```rust |
| 241 |
/// use zune_core::bytestream::ZByteReader; |
| 242 |
/// let mut stream = ZByteReader::new([0x0,0x5,0x3,0x2].as_slice()); |
| 243 |
/// let first_bytes = stream.get_fixed_bytes_or_zero::<2>(); |
| 244 |
/// assert_eq!(first_bytes,[0x0,0x5]); |
| 245 |
/// ``` |
| 246 |
/// |
| 247 |
/// [`has`]:Self::has |
| 248 |
/// [`get_fixed_bytes_or_err`]: Self::get_fixed_bytes_or_err |
| 249 |
#[inline] |
| 250 |
pub fn get_fixed_bytes_or_zero<const N: usize>(&mut self) -> [u8; N] { |
| 251 |
let mut byte_store: [u8; N] = [0; N]; |
| 252 |
|
| 253 |
match self.stream.get_slice(self.position..self.position + N) { |
| 254 |
Some(bytes) => { |
| 255 |
self.position += N; |
| 256 |
byte_store.copy_from_slice(bytes); |
| 257 |
|
| 258 |
byte_store |
| 259 |
} |
| 260 |
None => byte_store |
| 261 |
} |
| 262 |
} |
| 263 |
#[inline] |
| 264 |
/// Skip bytes until a condition becomes false or the stream runs out of bytes |
| 265 |
/// |
| 266 |
/// # Example |
| 267 |
/// |
| 268 |
/// ```rust |
| 269 |
/// use zune_core::bytestream::ZByteReader; |
| 270 |
/// let mut stream = ZByteReader::new([0;10].as_slice()); |
| 271 |
/// stream.skip_until_false(|x| x.is_ascii()) // skip until we meet a non ascii character |
| 272 |
/// ``` |
| 273 |
pub fn skip_until_false<F: Fn(u8) -> bool>(&mut self, func: F) { |
| 274 |
// iterate until we have no more bytes |
| 275 |
while !self.eof() { |
| 276 |
// get a byte from stream |
| 277 |
let byte = self.get_u8(); |
| 278 |
|
| 279 |
if !(func)(byte) { |
| 280 |
// function returned false meaning we stop skipping |
| 281 |
self.rewind(1); |
| 282 |
break; |
| 283 |
} |
| 284 |
} |
| 285 |
} |
| 286 |
/// Return the remaining unread bytes in this byte reader |
| 287 |
pub fn remaining_bytes(&self) -> &[u8] { |
| 288 |
debug_assert!(self.position <= self.len()); |
| 289 |
self.stream.get_slice(self.position..self.len()).unwrap() |
| 290 |
} |
| 291 |
|
| 292 |
pub fn read(&mut self, buf: &mut [u8]) -> Result<usize, &'static str> { |
| 293 |
let buf_length = buf.len(); |
| 294 |
let start = self.position; |
| 295 |
let end = min(self.len(), self.position + buf_length); |
| 296 |
let diff = end - start; |
| 297 |
|
| 298 |
buf[0..diff].copy_from_slice(self.stream.get_slice(start..end).unwrap()); |
| 299 |
|
| 300 |
self.skip(diff); |
| 301 |
|
| 302 |
Ok(diff) |
| 303 |
} |
| 304 |
|
| 305 |
/// Read enough bytes to fill in |
| 306 |
pub fn read_exact(&mut self, buf: &mut [u8]) -> Result<(), &'static str> { |
| 307 |
let size = self.read(buf)?; |
| 308 |
|
| 309 |
if size != buf.len() { |
| 310 |
return Err("Could not read into the whole buffer"); |
| 311 |
} |
| 312 |
Ok(()) |
| 313 |
} |
| 314 |
|
| 315 |
/// Set the cursor position |
| 316 |
/// |
| 317 |
/// After this, all reads will proceed from the position as an anchor |
| 318 |
/// point |
| 319 |
pub fn set_position(&mut self, position: usize) { |
| 320 |
self.position = position; |
| 321 |
} |
| 322 |
} |
| 323 |
|
| 324 |
macro_rules! get_single_type { |
| 325 |
($name:tt,$name2:tt,$name3:tt,$name4:tt,$name5:tt,$name6:tt,$int_type:tt) => { |
| 326 |
impl<T:ZReaderTrait> ZByteReader<T> |
| 327 |
{ |
| 328 |
#[inline(always)] |
| 329 |
fn $name(&mut self, mode: Mode) -> $int_type |
| 330 |
{ |
| 331 |
const SIZE_OF_VAL: usize = core::mem::size_of::<$int_type>(); |
| 332 |
|
| 333 |
let mut space = [0; SIZE_OF_VAL]; |
| 334 |
|
| 335 |
match self.stream.get_slice(self.position..self.position + SIZE_OF_VAL) |
| 336 |
{ |
| 337 |
Some(position) => |
| 338 |
{ |
| 339 |
space.copy_from_slice(position); |
| 340 |
self.position += SIZE_OF_VAL; |
| 341 |
|
| 342 |
match mode |
| 343 |
{ |
| 344 |
Mode::LE => $int_type::from_le_bytes(space), |
| 345 |
Mode::BE => $int_type::from_be_bytes(space), |
| 346 |
} |
| 347 |
} |
| 348 |
None => 0, |
| 349 |
} |
| 350 |
} |
| 351 |
|
| 352 |
#[inline(always)] |
| 353 |
fn $name2(&mut self, mode: Mode) -> Result<$int_type, &'static str> |
| 354 |
{ |
| 355 |
const SIZE_OF_VAL: usize = core::mem::size_of::<$int_type>(); |
| 356 |
|
| 357 |
let mut space = [0; SIZE_OF_VAL]; |
| 358 |
|
| 359 |
match self.stream.get_slice(self.position..self.position + SIZE_OF_VAL) |
| 360 |
{ |
| 361 |
Some(position) => |
| 362 |
{ |
| 363 |
space.copy_from_slice(position); |
| 364 |
self.position += SIZE_OF_VAL; |
| 365 |
|
| 366 |
match mode |
| 367 |
{ |
| 368 |
Mode::LE => Ok($int_type::from_le_bytes(space)), |
| 369 |
Mode::BE => Ok($int_type::from_be_bytes(space)), |
| 370 |
} |
| 371 |
} |
| 372 |
None => Err(ERROR_MSG), |
| 373 |
} |
| 374 |
} |
| 375 |
#[doc=concat!("Read ",stringify!($int_type)," as a big endian integer")] |
| 376 |
#[doc=concat!("Returning an error if the underlying buffer cannot support a ",stringify!($int_type)," read.")] |
| 377 |
#[inline] |
| 378 |
pub fn $name3(&mut self) -> Result<$int_type, &'static str> |
| 379 |
{ |
| 380 |
self.$name2(Mode::BE) |
| 381 |
} |
| 382 |
|
| 383 |
#[doc=concat!("Read ",stringify!($int_type)," as a little endian integer")] |
| 384 |
#[doc=concat!("Returning an error if the underlying buffer cannot support a ",stringify!($int_type)," read.")] |
| 385 |
#[inline] |
| 386 |
pub fn $name4(&mut self) -> Result<$int_type, &'static str> |
| 387 |
{ |
| 388 |
self.$name2(Mode::LE) |
| 389 |
} |
| 390 |
#[doc=concat!("Read ",stringify!($int_type)," as a big endian integer")] |
| 391 |
#[doc=concat!("Returning 0 if the underlying buffer does not have enough bytes for a ",stringify!($int_type)," read.")] |
| 392 |
#[inline(always)] |
| 393 |
pub fn $name5(&mut self) -> $int_type |
| 394 |
{ |
| 395 |
self.$name(Mode::BE) |
| 396 |
} |
| 397 |
#[doc=concat!("Read ",stringify!($int_type)," as a little endian integer")] |
| 398 |
#[doc=concat!("Returning 0 if the underlying buffer does not have enough bytes for a ",stringify!($int_type)," read.")] |
| 399 |
#[inline(always)] |
| 400 |
pub fn $name6(&mut self) -> $int_type |
| 401 |
{ |
| 402 |
self.$name(Mode::LE) |
| 403 |
} |
| 404 |
} |
| 405 |
}; |
| 406 |
} |
| 407 |
// U8 implementation |
| 408 |
// The benefit of our own unrolled u8 impl instead of macros is that this is sometimes used in some |
| 409 |
// impls and is called multiple times, e.g jpeg during huffman decoding. |
| 410 |
// we can make some functions leaner like get_u8 is branchless |
| 411 |
impl<T> ZByteReader<T> |
| 412 |
where |
| 413 |
T: ZReaderTrait |
| 414 |
{ |
| 415 |
/// Retrieve a byte from the underlying stream |
| 416 |
/// returning 0 if there are no more bytes available |
| 417 |
/// |
| 418 |
/// This means 0 might indicate a bit or an end of stream, but |
| 419 |
/// this is useful for some scenarios where one needs a byte. |
| 420 |
/// |
| 421 |
/// For the panicking one, see [`get_u8_err`] |
| 422 |
/// |
| 423 |
/// [`get_u8_err`]: Self::get_u8_err |
| 424 |
#[inline(always)] |
| 425 |
pub fn get_u8(&mut self) -> u8 { |
| 426 |
let byte = *self.stream.get_byte(self.position).unwrap_or(&0); |
| 427 |
|
| 428 |
self.position += usize::from(self.position < self.len()); |
| 429 |
byte |
| 430 |
} |
| 431 |
|
| 432 |
/// Retrieve a byte from the underlying stream |
| 433 |
/// returning an error if there are no more bytes available |
| 434 |
/// |
| 435 |
/// For the non panicking one, see [`get_u8`] |
| 436 |
/// |
| 437 |
/// [`get_u8`]: Self::get_u8 |
| 438 |
#[inline(always)] |
| 439 |
pub fn get_u8_err(&mut self) -> Result<u8, &'static str> { |
| 440 |
match self.stream.get_byte(self.position) { |
| 441 |
Some(byte) => { |
| 442 |
self.position += 1; |
| 443 |
Ok(*byte) |
| 444 |
} |
| 445 |
None => Err(ERROR_MSG) |
| 446 |
} |
| 447 |
} |
| 448 |
} |
| 449 |
|
| 450 |
// u16,u32,u64 -> macros |
| 451 |
get_single_type!( |
| 452 |
get_u16_inner_or_default, |
| 453 |
get_u16_inner_or_die, |
| 454 |
get_u16_be_err, |
| 455 |
get_u16_le_err, |
| 456 |
get_u16_be, |
| 457 |
get_u16_le, |
| 458 |
u16 |
| 459 |
); |
| 460 |
get_single_type!( |
| 461 |
get_u32_inner_or_default, |
| 462 |
get_u32_inner_or_die, |
| 463 |
get_u32_be_err, |
| 464 |
get_u32_le_err, |
| 465 |
get_u32_be, |
| 466 |
get_u32_le, |
| 467 |
u32 |
| 468 |
); |
| 469 |
get_single_type!( |
| 470 |
get_u64_inner_or_default, |
| 471 |
get_u64_inner_or_die, |
| 472 |
get_u64_be_err, |
| 473 |
get_u64_le_err, |
| 474 |
get_u64_be, |
| 475 |
get_u64_le, |
| 476 |
u64 |
| 477 |
); |