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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
//! Radix-generic, optimized, string-to-integer conversion routines.
//!
//! These routines are highly optimized: they use various optimizations
//! to read multiple digits at-a-time with less multiplication instructions,
//! as well as other optimizations to avoid unnecessary compile-time branching.
//!
//! See [Algorithm.md](/docs/Algorithm.md) for a more detailed description of
//! the algorithm choice here. See [Benchmarks.md](/docs/Benchmarks.md) for
//! recent benchmark data.
//!
//! These allow implementations of partial and complete parsers
//! using a single code-path via macros.
//!
//! This looks relatively, complex, but it's quite simple. Almost all
//! of these branches are resolved at compile-time, so the resulting
//! code is quite small while handling all of the internal complexity.
//!
//! 1. Helpers to process ok and error results for both complete and partial
//! parsers. They have different APIs, and mixing the APIs leads to
//! substantial performance hits.
//! 2. Overflow checking on invalid digits for partial parsers, while just
//! returning invalid digits for complete parsers.
//! 3. A format-aware sign parser.
//! 4. Digit parsing algorithms which explicitly wrap on overflow, for no
//! additional overhead. This has major performance wins for **most**
//! real-world integers, so most valid input will be substantially faster.
//! 5. An algorithm to detect if overflow occurred. This is comprehensive, and
//! short-circuits for common cases.
//! 6. A parsing algorithm for unsigned integers, always producing positive
//! values. This avoids any unnecessary branching.
//! 7. Multi-digit optimizations for larger sizes.
#![doc(hidden)]
use lexical_util::digit::char_to_digit_const;
use lexical_util::error::Error;
use lexical_util::format::NumberFormat;
use lexical_util::iterator::{AsBytes, Bytes, DigitsIter, Iter};
use lexical_util::num::{as_cast, Integer};
use lexical_util::result::Result;
use crate::Options;
// HELPERS
/// Check if we should do multi-digit optimizations
const fn can_try_parse_multidigits<'a, Iter: DigitsIter<'a>, const FORMAT: u128>(_: &Iter) -> bool {
let format = NumberFormat::<FORMAT> {};
Iter::IS_CONTIGUOUS && (cfg!(not(feature = "power-of-two")) || format.mantissa_radix() <= 10)
}
// Get if digits are required for the format.
#[cfg_attr(not(feature = "format"), allow(unused_macros))]
macro_rules! required_digits {
() => {
NumberFormat::<FORMAT>::REQUIRED_INTEGER_DIGITS
|| NumberFormat::<FORMAT>::REQUIRED_MANTISSA_DIGITS
};
}
/// Return an value for a complete parser.
macro_rules! into_ok_complete {
($value:expr, $index:expr, $count:expr) => {{
#[cfg(not(feature = "format"))]
return Ok(as_cast($value));
#[cfg(feature = "format")]
if required_digits!() && $count == 0 {
into_error!(Empty, $index);
} else {
return Ok(as_cast($value));
}
}};
}
/// Return an value and index for a partial parser.
macro_rules! into_ok_partial {
($value:expr, $index:expr, $count:expr) => {{
#[cfg(not(feature = "format"))]
return Ok((as_cast($value), $index));
#[cfg(feature = "format")]
if required_digits!() && $count == 0 {
into_error!(Empty, $index);
} else {
return Ok((as_cast($value), $index));
}
}};
}
/// Return an error for a complete parser upon an invalid digit.
macro_rules! invalid_digit_complete {
($value:expr, $index:expr, $count:expr) => {
// Don't do any overflow checking here: we don't need it.
into_error!(InvalidDigit, $index - 1)
};
}
/// Return a value for a partial parser upon an invalid digit.
/// This checks for numeric overflow, and returns the appropriate error.
macro_rules! invalid_digit_partial {
($value:expr, $index:expr, $count:expr) => {
// NOTE: The value is already positive/negative
into_ok_partial!($value, $index - 1, $count)
};
}
/// Return an error, returning the index and the error.
macro_rules! into_error {
($code:ident, $index:expr) => {{
return Err(Error::$code($index));
}};
}
/// Handle an invalid digit if the format feature is enabled.
///
/// This is because we can have special, non-digit characters near
/// the start or internally. If `$is_end` is set to false, there **MUST**
/// be elements in the underlying slice after the current iterator.
#[cfg(feature = "format")]
macro_rules! fmt_invalid_digit {
(
$value:ident, $iter:ident, $c:expr, $start_index:ident, $invalid_digit:ident, $is_end:expr
) => {{
// NOTE: If we have non-contiguous iterators, we could have a skip character
// here at the boundary. This does not affect safety but it does affect
// correctness.
debug_assert!($iter.is_contiguous() || $is_end);
let base_suffix = NumberFormat::<FORMAT>::BASE_SUFFIX;
let uncased_base_suffix = NumberFormat::<FORMAT>::CASE_SENSITIVE_BASE_SUFFIX;
// Need to check for a base suffix, if so, return a valid value.
// We can't have a base suffix at the first value (need at least
// 1 digit).
if base_suffix != 0 && $iter.cursor() - $start_index > 1 {
let is_suffix = if uncased_base_suffix {
$c == base_suffix
} else {
$c.to_ascii_lowercase() == base_suffix.to_ascii_lowercase()
};
// NOTE: If we're using the `take_n` optimization where it can't
// be the end, then the iterator cannot be done. So, in that case,
// we need to end. `take_n` also can never be used for non-
// contiguous iterators.
if is_suffix && $is_end && $iter.is_buffer_empty() {
// Break out of the loop, we've finished parsing.
break;
} else if !$iter.is_buffer_empty() {
// Haven't finished parsing, so we're going to call
// invalid_digit!. Need to ensure we include the
// base suffix in that.
// SAFETY: safe since the iterator is not empty, as checked
// in `$iter.is_buffer_empty()`. Adding in the check hopefully
// will be elided since it's a known constant.
unsafe { $iter.step_unchecked() };
}
}
// Might have handled our base-prefix here.
$invalid_digit!($value, $iter.cursor(), $iter.current_count())
}};
}
/// Just return an invalid digit
#[cfg(not(feature = "format"))]
macro_rules! fmt_invalid_digit {
(
$value:ident, $iter:ident, $c:expr, $start_index:ident, $invalid_digit:ident, $is_end:expr
) => {{
$invalid_digit!($value, $iter.cursor(), $iter.current_count());
}};
}
/// Parse the sign from the leading digits.
///
/// This routine does the following:
///
/// 1. Parses the sign digit.
/// 2. Handles if positive signs before integers are not allowed.
/// 3. Handles negative signs if the type is unsigned.
/// 4. Handles if the sign is required, but missing.
/// 5. Handles if the iterator is empty, before or after parsing the sign.
/// 6. Handles if the iterator has invalid, leading zeros.
///
/// Returns if the value is negative, or any values detected when
/// validating the input.
#[macro_export]
macro_rules! parse_sign {
(
$byte:ident,
$is_signed:expr,
$no_positive:expr,
$required:expr,
$invalid_positive:ident,
$missing:ident
) => {
// NOTE: read_if optimizes poorly since we then match after
match $byte.integer_iter().first() {
Some(&b'+') if !$no_positive => {
// SAFETY: We have at least 1 item left since we peaked a value
unsafe { $byte.step_unchecked() };
Ok(false)
},
Some(&b'+') if $no_positive => Err(Error::$invalid_positive($byte.cursor())),
Some(&b'-') if $is_signed => {
// SAFETY: We have at least 1 item left since we peaked a value
unsafe { $byte.step_unchecked() };
Ok(true)
},
Some(_) if $required => Err(Error::$missing($byte.cursor())),
_ if $required => Err(Error::$missing($byte.cursor())),
_ => Ok(false),
}
};
}
/// Parse the sign from the leading digits.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn parse_sign<T: Integer, const FORMAT: u128>(byte: &mut Bytes<'_, FORMAT>) -> Result<bool> {
let format = NumberFormat::<FORMAT> {};
parse_sign!(
byte,
T::IS_SIGNED,
format.no_positive_mantissa_sign(),
format.required_mantissa_sign(),
InvalidPositiveSign,
MissingSign
)
}
// FOUR DIGITS
/// Determine if 4 bytes, read raw from bytes, are 4 digits for the radix.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn is_4digits<const FORMAT: u128>(v: u32) -> bool {
let radix = NumberFormat::<{ FORMAT }>::MANTISSA_RADIX;
debug_assert!(radix <= 10);
// We want to have a wrapping add and sub such that only values from the
// range `[0x30, 0x39]` (or narrower for custom radixes) will not
// overflow into the high bit. This means that the value needs to overflow
// into into `0x80` if the digit is 1 above, or `0x46` for the value `0x39`.
// Likewise, we only valid for `[0x30, 0x38]` for radix 8, so we need
// `0x47`.
let add = 0x46 + 10 - radix;
let add = add + (add << 8) + (add << 16) + (add << 24);
// This aims to underflow if anything is below the min digit: if we have any
// values under `0x30`, then this underflows and wraps into the high bit.
let sub = 0x3030_3030;
let a = v.wrapping_add(add);
let b = v.wrapping_sub(sub);
(a | b) & 0x8080_8080 == 0
}
/// Parse 4 bytes read from bytes into 4 digits.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn parse_4digits<const FORMAT: u128>(mut v: u32) -> u32 {
let radix = NumberFormat::<{ FORMAT }>::MANTISSA_RADIX;
debug_assert!(radix <= 10);
// Normalize our digits to the range `[0, 9]`.
v -= 0x3030_3030;
// Scale digits in 0 <= Nn <= 99.
v = (v * radix) + (v >> 8);
// Scale digits in 0 <= Nnnn <= 9999.
v = ((v & 0x0000007f) * radix * radix) + ((v >> 16) & 0x0000007f);
v
}
/// Use a fast-path optimization, where we attempt to parse 4 digits at a time.
/// This reduces the number of multiplications necessary to 2, instead of 4.
///
/// This approach is described in full here:
/// <https://johnnylee-sde.github.io/Fast-numeric-string-to-int/>
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn try_parse_4digits<'a, T, Iter, const FORMAT: u128>(iter: &mut Iter) -> Option<T>
where
T: Integer,
Iter: DigitsIter<'a>,
{
// Can't do fast optimizations with radixes larger than 10, since
// we no longer have a contiguous ASCII block. Likewise, cannot
// use non-contiguous iterators.
debug_assert!(NumberFormat::<{ FORMAT }>::MANTISSA_RADIX <= 10);
debug_assert!(Iter::IS_CONTIGUOUS);
// Read our digits, validate the input, and check from there.
let bytes = u32::from_le(iter.peek_u32()?);
if is_4digits::<FORMAT>(bytes) {
// SAFETY: safe since we have at least 4 bytes in the buffer.
unsafe { iter.step_by_unchecked(4) };
Some(T::as_cast(parse_4digits::<FORMAT>(bytes)))
} else {
None
}
}
// EIGHT DIGITS
/// Determine if 8 bytes, read raw from bytes, are 8 digits for the radix.
/// See `is_4digits` for the algorithm description.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn is_8digits<const FORMAT: u128>(v: u64) -> bool {
let radix = NumberFormat::<{ FORMAT }>::MANTISSA_RADIX;
debug_assert!(radix <= 10);
let add = 0x46 + 10 - radix;
let add = add + (add << 8) + (add << 16) + (add << 24);
let add = (add as u64) | ((add as u64) << 32);
// This aims to underflow if anything is below the min digit: if we have any
// values under `0x30`, then this underflows and wraps into the high bit.
let sub = 0x3030_3030_3030_3030;
let a = v.wrapping_add(add);
let b = v.wrapping_sub(sub);
(a | b) & 0x8080_8080_8080_8080 == 0
}
/// Parse 8 bytes read from bytes into 8 digits.
/// Credit for this goes to @aqrit, which further optimizes the
/// optimization described by Johnny Lee above.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn parse_8digits<const FORMAT: u128>(mut v: u64) -> u64 {
let radix = NumberFormat::<{ FORMAT }>::MANTISSA_RADIX as u64;
debug_assert!(radix <= 10);
// Create our masks. Assume the optimizer will do this at compile time.
// It seems like an optimizing compiler **will** do this, so we
// should be safe.
let radix2 = radix * radix;
let radix4 = radix2 * radix2;
let radix6 = radix2 * radix4;
let mask = 0x0000_00FF_0000_00FFu64;
let mul1 = radix2 + (radix6 << 32);
let mul2 = 1 + (radix4 << 32);
// Normalize our digits to the base.
v -= 0x3030_3030_3030_3030;
// Scale digits in 0 <= Nn <= 99.
v = (v * radix) + (v >> 8);
let v1 = (v & mask).wrapping_mul(mul1);
let v2 = ((v >> 16) & mask).wrapping_mul(mul2);
((v1.wrapping_add(v2) >> 32) as u32) as u64
}
/// Use a fast-path optimization, where we attempt to parse 8 digits at a time.
/// This reduces the number of multiplications necessary to 3, instead of 8.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn try_parse_8digits<'a, T, Iter, const FORMAT: u128>(iter: &mut Iter) -> Option<T>
where
T: Integer,
Iter: DigitsIter<'a>,
{
// Can't do fast optimizations with radixes larger than 10, since
// we no longer have a contiguous ASCII block. Likewise, cannot
// use non-contiguous iterators.
debug_assert!(NumberFormat::<{ FORMAT }>::MANTISSA_RADIX <= 10);
debug_assert!(Iter::IS_CONTIGUOUS);
// Read our digits, validate the input, and check from there.
let bytes = u64::from_le(iter.peek_u64()?);
if is_8digits::<FORMAT>(bytes) {
// SAFETY: safe since we have at least 8 bytes in the buffer.
unsafe { iter.step_by_unchecked(8) };
Some(T::as_cast(parse_8digits::<FORMAT>(bytes)))
} else {
None
}
}
// ONE DIGIT
/// Run a loop where the integer cannot possibly overflow.
///
/// If the len of the str is short compared to the range of the type
/// we are parsing into, then we can be certain that an overflow will not occur.
/// This bound is when `radix.pow(digits.len()) - 1 <= T::MAX` but the condition
/// above is a faster (conservative) approximation of this.
///
/// Consider radix 16 as it has the highest information density per digit and
/// will thus overflow the earliest: `u8::MAX` is `ff` - any str of len 2 is
/// guaranteed to not overflow. `i8::MAX` is `7f` - only a str of len 1 is
/// guaranteed to not overflow.
///
/// This is based off of [core/num](core).
///
/// * `value` - The current parsed value.
/// * `iter` - An iterator over all bytes in the input.
/// * `add_op` - The unchecked add/sub op.
/// * `start_index` - The offset where parsing started.
/// * `invalid_digit` - Behavior when an invalid digit is found.
/// * `is_end` - If iter corresponds to the full input.
///
/// core: <https://doc.rust-lang.org/1.81.0/src/core/num/mod.rs.html#1480>
macro_rules! parse_1digit_unchecked {
(
$value:ident,
$iter:ident,
$add_op:ident,
$start_index:ident,
$invalid_digit:ident,
$is_end:expr
) => {{
// This is a slower parsing algorithm, going 1 digit at a time, but doing it in
// an unchecked loop.
let radix = NumberFormat::<FORMAT>::MANTISSA_RADIX;
while let Some(&c) = $iter.next() {
let digit = match char_to_digit_const(c, radix) {
Some(v) => v,
None => fmt_invalid_digit!($value, $iter, c, $start_index, $invalid_digit, $is_end),
};
// multiply first since compilers are good at optimizing things out and will do
// a fused mul/add We must do this after getting the digit for
// partial parsers
$value = $value.wrapping_mul(as_cast(radix)).$add_op(as_cast(digit));
}
}};
}
/// Run a loop where the integer could overflow.
///
/// This is a standard, unoptimized algorithm. This is based off of
/// [core/num](core)
///
/// * `value` - The current parsed value.
/// * `iter` - An iterator over all bytes in the input.
/// * `add_op` - The checked add/sub op.
/// * `start_index` - The offset where parsing started.
/// * `invalid_digit` - Behavior when an invalid digit is found.
/// * `overflow` - If the error is overflow or underflow.
///
/// core: <https://doc.rust-lang.org/1.81.0/src/core/num/mod.rs.html#1505>
macro_rules! parse_1digit_checked {
(
$value:ident,
$iter:ident,
$add_op:ident,
$start_index:ident,
$invalid_digit:ident,
$overflow:ident
) => {{
// This is a slower parsing algorithm, going 1 digit at a time, but doing it in
// an unchecked loop.
let radix = NumberFormat::<FORMAT>::MANTISSA_RADIX;
while let Some(&c) = $iter.next() {
let digit = match char_to_digit_const(c, radix) {
Some(v) => v,
None => fmt_invalid_digit!($value, $iter, c, $start_index, $invalid_digit, true),
};
// multiply first since compilers are good at optimizing things out and will do
// a fused mul/add
$value =
match $value.checked_mul(as_cast(radix)).and_then(|x| x.$add_op(as_cast(digit))) {
Some(value) => value,
None => into_error!($overflow, $iter.cursor() - 1),
}
}
}};
}
// OVERALL DIGITS
// --------------
/// Run an unchecked loop where digits are processed incrementally.
///
/// If the type size is small or there's not many digits, skip multi-digit
/// optimizations. Otherwise, if the type size is large and we're not manually
/// skipping manual optimizations, then we do this here.
///
/// * `value` - The current parsed value.
/// * `iter` - An iterator over all bytes in the input.
/// * `add_op` - The unchecked add/sub op.
/// * `start_index` - The offset where parsing started.
/// * `invalid_digit` - Behavior when an invalid digit is found.
/// * `no_multi_digit` - If to disable multi-digit optimizations.
/// * `is_end` - If iter corresponds to the full input.
macro_rules! parse_digits_unchecked {
(
$value:ident,
$iter:ident,
$add_op:ident,
$start_index:ident,
$invalid_digit:ident,
$no_multi_digit:expr,
$is_end:expr
) => {{
let can_multi = can_try_parse_multidigits::<_, FORMAT>(&$iter);
let use_multi = can_multi && !$no_multi_digit;
// these cannot overflow. also, we use at most 3 for a 128-bit float and 1 for a
// 64-bit float NOTE: Miri will complain about this if we use radices >=
// 16 but since they won't go into `try_parse_8digits!` or
// `try_parse_4digits` it will be optimized out and the overflow won't
// matter.
let format = NumberFormat::<FORMAT> {};
if use_multi && T::BITS >= 64 && $iter.buffer_length() >= 8 {
// Try our fast, 8-digit at a time optimizations.
let radix8 = T::from_u32(format.radix8());
while let Some(value) = try_parse_8digits::<T, _, FORMAT>(&mut $iter) {
$value = $value.wrapping_mul(radix8).$add_op(value);
}
} else if use_multi && T::BITS == 32 && $iter.buffer_length() >= 4 {
// Try our fast, 8-digit at a time optimizations.
let radix4 = T::from_u32(format.radix4());
while let Some(value) = try_parse_4digits::<T, _, FORMAT>(&mut $iter) {
$value = $value.wrapping_mul(radix4).$add_op(value);
}
}
parse_1digit_unchecked!($value, $iter, $add_op, $start_index, $invalid_digit, $is_end)
}};
}
/// Run checked loop where digits are processed with overflow checking.
///
/// If the type size is small or there's not many digits, skip multi-digit
/// optimizations. Otherwise, if the type size is large and we're not manually
/// skipping manual optimizations, then we do this here.
///
/// * `value` - The current parsed value.
/// * `iter` - An iterator over all bytes in the input.
/// * `add_op` - The checked add/sub op.
/// * `add_op_uc` - The unchecked add/sub op for small digit optimizations.
/// * `start_index` - The offset where parsing started.
/// * `invalid_digit` - Behavior when an invalid digit is found.
/// * `overflow` - If the error is overflow or underflow.
/// * `no_multi_digit` - If to disable multi-digit optimizations.
/// * `overflow_digits` - The number of digits before we need to consider
/// checked ops.
macro_rules! parse_digits_checked {
(
$value:ident,
$iter:ident,
$add_op:ident,
$add_op_uc:ident,
$start_index:ident,
$invalid_digit:ident,
$overflow:ident,
$no_multi_digit:expr,
$overflow_digits:expr
) => {{
// Can use the unchecked for the max_digits here. If we
// have a non-contiguous iterator, we could have a case like
// 123__456, with no consecutive digit sepatators allowed. If
// it's broken between the `_` characters, the integer will be
// seen as valid when it isn't.
if cfg!(not(feature = "format")) || $iter.is_contiguous() {
if let Some(mut small) = $iter.take_n($overflow_digits) {
let mut small_iter = small.integer_iter();
parse_digits_unchecked!(
$value,
small_iter,
$add_op_uc,
$start_index,
$invalid_digit,
$no_multi_digit,
false
);
}
}
// NOTE: all our multi-digit optimizations have been done here: skip this
parse_1digit_checked!($value, $iter, $add_op, $start_index, $invalid_digit, $overflow)
}};
}
// ALGORITHM
/// Generic algorithm for both partial and complete parsers.
///
/// * `invalid_digit` - Behavior on finding an invalid digit.
/// * `into_ok` - Behavior when returning a valid value.
/// * `invalid_digit` - Behavior when an invalid digit is found.
/// * `no_multi_digit` - If to disable multi-digit optimizations.
/// * `is_partial` - If the parser is a partial parser.
#[rustfmt::skip]
macro_rules! algorithm {
($bytes:ident, $into_ok:ident, $invalid_digit:ident, $no_multi_digit:expr) => {{
// WARNING:
// --------
// None of this code can be changed for optimization reasons.
// Do not change it without benchmarking every change.
// 1. You cannot use the NoSkipIterator in the loop,
// you must either return a subslice (indexing)
// or increment outside of the loop.
// Failing to do so leads to numerous more, unnecessary
// conditional move instructions, killing performance.
// 2. Return a 0 or 1 shift, and indexing unchecked outside
// of the loop is slightly faster.
// 3. Partial and complete parsers cannot be efficiently done
// together.
//
// If you try to refactor without carefully monitoring benchmarks or
// assembly generation, please log the number of wasted hours: so
// 16 hours so far.
// With `step_by_unchecked`, this is sufficiently optimized.
// Removes conditional paths, to, which simplifies maintenance.
// The skip version of the iterator automatically coalesces to
// the noskip iterator.
let mut byte = $bytes.bytes::<FORMAT>();
let radix = NumberFormat::<FORMAT>::MANTISSA_RADIX;
let is_negative = parse_sign::<T, FORMAT>(&mut byte)?;
let mut iter = byte.integer_iter();
if iter.is_buffer_empty() {
// Our default format **ALWAYS** requires significant digits, however,
// we can have cases where we don
#[cfg(not(feature = "format"))]
into_error!(Empty, iter.cursor());
#[cfg(feature = "format")]
if required_digits!() {
into_error!(Empty, iter.cursor());
} else {
$into_ok!(T::ZERO, iter.cursor(), 0)
}
}
// Feature-gate a lot of format-only code here to simplify analysis with our branching
// We only want to skip the zeros if have either require a base prefix or we don't
// allow integer leading zeros, since the skip is expensive
#[allow(unused_variables, unused_mut)]
let mut start_index = iter.cursor();
#[cfg_attr(not(feature = "format"), allow(unused_variables))]
let format = NumberFormat::<FORMAT> {};
#[cfg(feature = "format")]
if format.has_base_prefix() || format.no_integer_leading_zeros() {
// Skip any leading zeros. We want to do our check if it can't possibly overflow after.
// For skipping digit-based formats, this approximation is a way over estimate.
// NOTE: Skipping zeros is **EXPENSIVE* so we skip that without our format feature
let zeros = iter.skip_zeros();
start_index += zeros;
// Now, check to see if we have a valid base prefix.
let mut is_prefix = false;
let base_prefix = format.base_prefix();
if base_prefix != 0 && zeros == 1 {
// Check to see if the next character is the base prefix.
// We must have a format like `0x`, `0d`, `0o`. Note:
if iter.read_if_value(base_prefix, format.case_sensitive_base_prefix()).is_some() {
is_prefix = true;
if iter.is_buffer_empty() {
into_error!(Empty, iter.cursor());
} else {
start_index += 1;
}
}
}
// If we have a format that doesn't accept leading zeros,
// check if the next value is invalid. It's invalid if the
// first is 0, and the next is not a valid digit.
if !is_prefix && format.no_integer_leading_zeros() && zeros != 0 {
// Cannot have a base prefix and no leading zeros.
let index = iter.cursor() - zeros;
if zeros > 1 {
into_error!(InvalidLeadingZeros, index);
}
// NOTE: Zeros has to be 0 here, so our index == 1 or 2 (depending on sign)
match iter.peek().map(|&c| char_to_digit_const(c, format.radix())) {
// Valid digit, we have an invalid value.
Some(Some(_)) => into_error!(InvalidLeadingZeros, index),
// Have a non-digit character that follows.
Some(None) => $invalid_digit!(<T>::ZERO, iter.cursor() + 1, iter.current_count()),
// No digits following, has to be ok
None => $into_ok!(<T>::ZERO, index, iter.current_count()),
};
}
}
// shorter strings cannot possibly overflow so a great optimization
let overflow_digits = T::overflow_digits(radix);
let cannot_overflow = iter.as_slice().len() <= overflow_digits;
// NOTE:
// Don't add optimizations for 128-bit integers.
// 128-bit multiplication is rather efficient, it's only division
// that's very slow. Any shortcut optimizations increasing branching,
// and even if parsing a 64-bit integer is marginally faster, it
// culminates in **way** slower performance overall for simple
// integers, and no improvement for large integers.
let mut value = T::ZERO;
if cannot_overflow && is_negative {
parse_digits_unchecked!(value, iter, wrapping_sub, start_index, $invalid_digit, $no_multi_digit, true);
} if cannot_overflow {
parse_digits_unchecked!(value, iter, wrapping_add, start_index, $invalid_digit, $no_multi_digit, true);
} else if is_negative {
parse_digits_checked!(value, iter, checked_sub, wrapping_sub, start_index, $invalid_digit, Underflow, $no_multi_digit, overflow_digits);
} else {
parse_digits_checked!(value, iter, checked_add, wrapping_add, start_index, $invalid_digit, Overflow, $no_multi_digit, overflow_digits);
}
$into_ok!(value, iter.buffer_length(), iter.current_count())
}};
}
/// Algorithm for the complete parser.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn algorithm_complete<T, const FORMAT: u128>(bytes: &[u8], options: &Options) -> Result<T>
where
T: Integer,
{
algorithm!(bytes, into_ok_complete, invalid_digit_complete, options.get_no_multi_digit())
}
/// Algorithm for the partial parser.
#[cfg_attr(not(feature = "compact"), inline(always))]
pub fn algorithm_partial<T, const FORMAT: u128>(
bytes: &[u8],
options: &Options,
) -> Result<(T, usize)>
where
T: Integer,
{
algorithm!(bytes, into_ok_partial, invalid_digit_partial, options.get_no_multi_digit())
}