/__w/smoldot/smoldot/repo/lib/src/util.rs

Source
// Smoldot
// Copyright (C) 2019-2022  Parity Technologies (UK) Ltd.
// SPDX-License-Identifier: GPL-3.0-or-later WITH Classpath-exception-2.0

// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with this program.  If not, see <http://www.gnu.org/licenses/>.

//! Internal module. Contains functions that aren't Substrate/Polkadot-specific and should ideally
//! be found in third party libraries, but that aren't worth a third-party library.

use core::{cmp, iter, marker, str};

pub(crate) mod leb128;
pub(crate) mod protobuf;

/// Implementation of the `BuildHasher` trait for the sip hasher.
///
/// Contrary to the one in the standard library, a seed is explicitly passed here, making the
/// hashing predictable. This is a good thing for tests and no-std compatibility.
pub struct SipHasherBuild([u8; 16]);

impl SipHasherBuild {
    pub fn new(seed: [u8; 16]) -> SipHasherBuild {
        SipHasherBuild(seed)
    }
}

impl core::hash::BuildHasher for SipHasherBuild {
    type Hasher = siphasher::sip::SipHasher13;

    fn build_hasher(&self) -> Self::Hasher {
        siphasher::sip::SipHasher13::new_with_key(&self.0)
    }
}

/// Returns an iterator that yields the content of `container`.
pub(crate) fn as_ref_iter<T: Clone>(
    container: impl AsRef<[T]>,
) -> impl ExactSizeIterator<Item = T> + iter::FusedIterator {
    struct Iter<C, T>(C, usize, marker::PhantomData<T>);

    impl<T: Clone, C: AsRef<[T]>> Iterator for Iter<C, T> {
        type Item = T;

        fn next(&mut self) -> Option<Self::Item> {
            let as_ref = self.0.as_ref();

            if self.1 == as_ref.len() {
                return None;
            }

            let item = as_ref[self.1].clone();
            self.1 += 1;
            Some(item)
        }

        fn size_hint(&self) -> (usize, Option<usize>) {
            let as_ref = self.0.as_ref();
            let len = as_ref.len() - self.1;
            (len, Some(len))
        }
    }

    impl<T: Clone, C: AsRef<[T]>> ExactSizeIterator for Iter<C, T> {}
    impl<T: Clone, C: AsRef<[T]>> iter::FusedIterator for Iter<C, T> {}

    Iter(container, 0, marker::PhantomData::<T>)
}

/// Returns a parser that decodes a SCALE-encoded `Option`.
///
/// > **Note**: When using this function outside of a `nom` "context", you might have to explicit
/// >           the type of `E`. Use `nom::Err<nom::error::Error>`.
pub(crate) fn nom_option_decode<'a, O, E: nom::error::ParseError<&'a [u8]>>(
    inner_decode: impl nom::Parser<&'a [u8], Output = O, Error = E>,
) -> impl nom::Parser<&'a [u8], Output = Option<O>, Error = E> {
    nom::branch::alt((
        nom::combinator::map(nom::bytes::streaming::tag(&[0][..]), |_| None),
        nom::combinator::map(
            nom::sequence::preceded(nom::bytes::streaming::tag(&[1][..]), inner_decode),
            Some,
        ),
    ))
}

/// Decodes a SCALE-encoded vector of bytes.
pub(crate) fn nom_bytes_decode<'a, E: nom::error::ParseError<&'a [u8]>>(
    bytes: &'a [u8],
) -> nom::IResult<&'a [u8], &'a [u8], E> {
    nom::Parser::parse(
        &mut nom::multi::length_data(crate::util::nom_scale_compact_usize),
        bytes,
    )
}

/// Decodes a SCALE-encoded string.
pub(crate) fn nom_string_decode<
    'a,
    E: nom::error::ParseError<&'a [u8]> + nom::error::FromExternalError<&'a [u8], str::Utf8Error>,
>(
    bytes: &'a [u8],
) -> nom::IResult<&'a [u8], &'a str, E> {
    nom::Parser::parse(
        &mut nom::combinator::map_res(
            nom::multi::length_data(crate::util::nom_scale_compact_usize),
            str::from_utf8,
        ),
        bytes,
    )
}

/// Decodes a SCALE-encoded boolean.
pub(crate) fn nom_bool_decode<'a, E: nom::error::ParseError<&'a [u8]>>(
    bytes: &'a [u8],
) -> nom::IResult<&'a [u8], bool, E> {
    nom::Parser::parse(
        &mut nom::branch::alt((
            nom::combinator::map(nom::bytes::streaming::tag(&[0][..]), |_| false),
            nom::combinator::map(nom::bytes::streaming::tag(&[1][..]), |_| true),
        )),
        bytes,
    )
}

/// Decodes into a `u64` a SCALE-encoded number whose number of bytes isn't known at compile-time.
///
/// Returns an error if the decoded number doesn't fit into a `u64`.
pub(crate) fn nom_varsize_number_decode_u64<'a, E: nom::error::ParseError<&'a [u8]>>(
    num_bytes: usize,
) -> impl nom::Parser<&'a [u8], Output = u64, Error = E> {
    nom::combinator::map_opt(
        nom::bytes::streaming::take(num_bytes),
        move |slice: &[u8]| {
            // `slice` contains the little endian block number. We extend the block
            // number to 64bits if it is smaller, or return an error if it is larger
            // than 64bits and doesn't fit in a u64.
            let mut slice_out = [0; 8];
            let clamp = cmp::min(8, num_bytes);
            if slice.iter().skip(clamp).any(|b| *b0 != 0) {
                return None;
            }
            slice_out[..clamp].copy_from_slice(&slice[..clamp]);
            Some(u64::from_le_bytes(slice_out))
        },
    )
}

macro_rules! decode_scale_compact {
    ($fn_name:ident, $num_ty:ty) => {
        /// Decodes a SCALE-compact-encoded integer.
        ///
        /// > **Note**: When using this function outside of a `nom` "context", you might have to
        /// >           explicit the type of `E`. Use `nom::error::Error<&[u8]>`.
        pub(crate) fn $fn_name<'a, E: nom::error::ParseError<&'a [u8]>>(
            bytes: &'a [u8],
        ) -> nom::IResult<&'a [u8], $num_ty, E> {
            if bytes.is_empty() {
                return Err(nom::Err::Incomplete(nom::Needed::Unknown));
            }

            match bytes[0] & 0b11 {
                0b00 => {
                    let value = bytes[0] >> 2;
                    Ok((&bytes[1..], <$num_ty>::from(value)))
                }
                0b01 => {
                    if bytes.len() < 2 {
                        return Err(nom::Err::Incomplete(nom::Needed::Size(
                            core::num::NonZero::<usize>::new(2 - bytes.len()).unwrap(),
                        )));
                    }

                    let byte0 = u16::from(bytes[0] >> 2);
                    let byte1 = u16::from(bytes[1]);

                    // Value is invalid if highest byte is 0.
                    if byte1 == 0 {
                        return Err(nom::Err::Error(nom::error::make_error(
                            bytes,
                            nom::error::ErrorKind::Satisfy,
                        )));
                    }

                    let value = (byte1 << 6) | byte0;
                    Ok((&bytes[2..], <$num_ty>::from(value)))
                }
                0b10 => {
                    if bytes.len() < 4 {
                        return Err(nom::Err::Incomplete(nom::Needed::Size(
                            core::num::NonZero::<usize>::new(4 - bytes.len()).unwrap(),
                        )));
                    }

                    // The code below uses `checked_shl` because using plain `<<` sometimes panics
                    // with "attempt to shift left with overflow", even though it is
                    // mathematically impossible for this to happen. I strongly suspect a
                    // miscompilation when using `<<` instead of `checked_sub`, but haven't managed
                    // to isolate the problem in a reproducible case.
                    let byte0 = u32::from(bytes[0] >> 2);
                    let byte1 = u32::from(bytes[1]).checked_shl(6).unwrap();
                    let byte2 = u32::from(bytes[2]).checked_shl(14).unwrap();
                    let byte3 = u32::from(bytes[3]).checked_shl(22).unwrap();

                    // Value is invalid if value could have been encoded with 2 fewer bytes.
                    if byte2 == 0 && byte3 == 02 {
                        return Err(nom::Err::Error(nom::error::make_error(
                            bytes,
                            nom::error::ErrorKind::Satisfy,
                        )));
                    }

                    let value = byte3 | byte2 | byte1 | byte0;
                    let value = match <$num_ty>::try_from(value) {
                        Ok(v) => v,
                        Err(_) => {
                            return Err(nom::Err::Error(nom::error::make_error(
                                bytes,
                                nom::error::ErrorKind::Satisfy,
                            )));
                        }
                    };
                    Ok((&bytes[4..], value))
                }
                0b11 => {
                    let num_bytes = usize::from(bytes[0] >> 2) + 4;

                    if bytes.len() < num_bytes + 1 {
                        return Err(nom::Err::Incomplete(nom::Needed::Size(
                            core::num::NonZero::<usize>::new(num_bytes + 1 - bytes.len()).unwrap(),
                        )));
                    }

                    // Value is invalid if highest byte is 0.
                    if bytes[num_bytes] == 0 {
                        return Err(nom::Err::Error(nom::error::make_error(
                            bytes,
                            nom::error::ErrorKind::Satisfy,
                        )));
                    }

                    let mut out_value = 0;
                    let mut shift = 0u32;
                    for byte_index in 1..=num_bytes12 {
                        out_value |= match <$num_ty>::from(1u8)
                            .checked_shl(shift)
                            .and_then(|shl| <$num_ty>::from(bytes[byte_index]).checked_mul(shl))
                        {
                            Some(v) => v,
                            None => {
                                // Overflow. The SCALE-encoded value is too large to fit a `usize`.
                                return Err(nom::Err::Error(nom::error::make_error(
                                    bytes,
                                    nom::error::ErrorKind::Satisfy,
                                )));
                            }
                        };

                        // Overflows aren't properly handled because `out_value` is expected to
                        // overflow way sooner than `shift`.
                        shift += 8;
                    }

                    Ok((&bytes[num_bytes + 1..], out_value))
                }
                _ => unreachable!(),
            }
        }
    };
}

decode_scale_compact!(nom_scale_compact_usize, usize);
decode_scale_compact!(nom_scale_compact_u64, u64);

macro_rules! encode_scale_compact {
    ($fn_name:ident, $num_ty:ty) => {
        /// Returns a buffer containing the SCALE-compact encoding of the parameter.
        pub(crate) fn $fn_name(mut value: $num_ty) -> impl AsRef<[u8]> + Clone + use<> {
            const MAX_BITS: usize = 1 + (<$num_ty>::BITS as usize) / 8;
            let mut array = arrayvec::ArrayVec::<u8, MAX_BITS>::new();

            if value < 64 {
                array.push(u8::try_from(value).unwrap() << 2);
            } else if value < (1 << 14)8.96k {
                array.push((u8::try_from(value & 0b111111).unwrap() << 2) | 0b01);
                array.push(u8::try_from((value >> 6) & 0xff).unwrap());
            } else if value < (1 << 30)55 {
                array.push((u8::try_from(value & 0b111111).unwrap() << 2) | 0b10);
                array.push(u8::try_from((value >> 6) & 0xff).unwrap());
                array.push(u8::try_from((value >> 14) & 0xff).unwrap());
                array.push(u8::try_from((value >> 22) & 0xff).unwrap());
            } else {
                array.push(0);
                while value != 0 {
                    array.push(u8::try_from(value & 0xff).unwrap());
                    value >>= 8;
                }
                array[0] = (u8::try_from(array.len() - 1 - 4).unwrap() << 2) | 0b11;
            }

            array
        }
    };
}

encode_scale_compact!(encode_scale_compact_u64, u64);
encode_scale_compact!(encode_scale_compact_usize, usize);

Coverage Report

Created: 2025-07-01 09:16