##// END OF EJS Templates
copies: no longer cache the ChangedFiles during copy tracing...
copies: no longer cache the ChangedFiles during copy tracing Now that the copies information for both parents are processed all at once, we no longer needs to cache this information, so we simplify the code. The simpler code is also a (tiny) bit faster overall. Repo Case Source-Rev Dest-Rev # of revisions old time new time Difference Factor time per rev --------------------------------------------------------------------------------------------------------------------------------------------------------------- mercurial x_revs_x_added_0_copies ad6b123de1c7 39cfcef4f463 : 1 revs, 0.000041 s, 0.000041 s, +0.000000 s, × 1.0000, 41 µs/rev mercurial x_revs_x_added_x_copies 2b1c78674230 0c1d10351869 : 6 revs, 0.000102 s, 0.000096 s, -0.000006 s, × 0.9412, 16 µs/rev mercurial x000_revs_x000_added_x_copies 81f8ff2a9bf2 dd3267698d84 : 1032 revs, 0.004254 s, 0.004039 s, -0.000215 s, × 0.9495, 3 µs/rev pypy x_revs_x_added_0_copies aed021ee8ae8 099ed31b181b : 9 revs, 0.000282 s, 0.000189 s, -0.000093 s, × 0.6702, 21 µs/rev pypy x_revs_x000_added_0_copies 4aa4e1f8e19a 359343b9ac0e : 1 revs, 0.000048 s, 0.000047 s, -0.000001 s, × 0.9792, 47 µs/rev pypy x_revs_x_added_x_copies ac52eb7bbbb0 72e022663155 : 7 revs, 0.000211 s, 0.000103 s, -0.000108 s, × 0.4882, 14 µs/rev pypy x_revs_x00_added_x_copies c3b14617fbd7 ace7255d9a26 : 1 revs, 0.000375 s, 0.000286 s, -0.000089 s, × 0.7627, 286 µs/rev pypy x_revs_x000_added_x000_copies df6f7a526b60 a83dc6a2d56f : 6 revs, 0.010574 s, 0.010436 s, -0.000138 s, × 0.9869, 1739 µs/rev pypy x000_revs_xx00_added_0_copies 89a76aede314 2f22446ff07e : 4785 revs, 0.049974 s, 0.047465 s, -0.002509 s, × 0.9498, 9 µs/rev pypy x000_revs_x000_added_x_copies 8a3b5bfd266e 2c68e87c3efe : 6780 revs, 0.084300 s, 0.082351 s, -0.001949 s, × 0.9769, 12 µs/rev pypy x000_revs_x000_added_x000_copies 89a76aede314 7b3dda341c84 : 5441 revs, 0.060128 s, 0.058757 s, -0.001371 s, × 0.9772, 10 µs/rev pypy x0000_revs_x_added_0_copies d1defd0dc478 c9cb1334cc78 : 43645 revs, 0.686542 s, 0.674129 s, -0.012413 s, × 0.9819, 15 µs/rev pypy x0000_revs_xx000_added_0_copies bf2c629d0071 4ffed77c095c : 2 revs, 0.009277 s, 0.009434 s, +0.000157 s, × 1.0169, 4717 µs/rev pypy x0000_revs_xx000_added_x000_copies 08ea3258278e d9fa043f30c0 : 11316 revs, 0.114733 s, 0.111935 s, -0.002798 s, × 0.9756, 9 µs/rev netbeans x_revs_x_added_0_copies fb0955ffcbcd a01e9239f9e7 : 2 revs, 0.000081 s, 0.000078 s, -0.000003 s, × 0.9630, 39 µs/rev netbeans x_revs_x000_added_0_copies 6f360122949f 20eb231cc7d0 : 2 revs, 0.000107 s, 0.000106 s, -0.000001 s, × 0.9907, 53 µs/rev netbeans x_revs_x_added_x_copies 1ada3faf6fb6 5a39d12eecf4 : 3 revs, 0.000173 s, 0.000162 s, -0.000011 s, × 0.9364, 54 µs/rev netbeans x_revs_x00_added_x_copies 35be93ba1e2c 9eec5e90c05f : 9 revs, 0.000698 s, 0.000695 s, -0.000003 s, × 0.9957, 77 µs/rev netbeans x000_revs_xx00_added_0_copies eac3045b4fdd 51d4ae7f1290 : 1421 revs, 0.009248 s, 0.008901 s, -0.000347 s, × 0.9625, 6 µs/rev netbeans x000_revs_x000_added_x_copies e2063d266acd 6081d72689dc : 1533 revs, 0.015446 s, 0.014333 s, -0.001113 s, × 0.9279, 9 µs/rev netbeans x000_revs_x000_added_x000_copies ff453e9fee32 411350406ec2 : 5750 revs, 0.074373 s, 0.071998 s, -0.002375 s, × 0.9681, 12 µs/rev netbeans x0000_revs_xx000_added_x000_copies 588c2d1ced70 1aad62e59ddd : 66949 revs, 0.639870 s, 0.615346 s, -0.024524 s, × 0.9617, 9 µs/rev mozilla-central x_revs_x_added_0_copies 3697f962bb7b 7015fcdd43a2 : 2 revs, 0.000088 s, 0.000085 s, -0.000003 s, × 0.9659, 42 µs/rev mozilla-central x_revs_x000_added_0_copies dd390860c6c9 40d0c5bed75d : 8 revs, 0.000199 s, 0.000199 s, +0.000000 s, × 1.0000, 24 µs/rev mozilla-central x_revs_x_added_x_copies 8d198483ae3b 14207ffc2b2f : 9 revs, 0.000171 s, 0.000169 s, -0.000002 s, × 0.9883, 18 µs/rev mozilla-central x_revs_x00_added_x_copies 98cbc58cc6bc 446a150332c3 : 7 revs, 0.000592 s, 0.000590 s, -0.000002 s, × 0.9966, 84 µs/rev mozilla-central x_revs_x000_added_x000_copies 3c684b4b8f68 0a5e72d1b479 : 3 revs, 0.003151 s, 0.003122 s, -0.000029 s, × 0.9908, 1040 µs/rev mozilla-central x_revs_x0000_added_x0000_copies effb563bb7e5 c07a39dc4e80 : 6 revs, 0.061612 s, 0.061192 s, -0.000420 s, × 0.9932, 10198 µs/rev mozilla-central x000_revs_xx00_added_0_copies 6100d773079a 04a55431795e : 1593 revs, 0.005381 s, 0.005137 s, -0.000244 s, × 0.9547, 3 µs/rev mozilla-central x000_revs_x000_added_x_copies 9f17a6fc04f9 2d37b966abed : 41 revs, 0.003742 s, 0.003585 s, -0.000157 s, × 0.9580, 87 µs/rev mozilla-central x000_revs_x000_added_x000_copies 7c97034feb78 4407bd0c6330 : 7839 revs, 0.061983 s, 0.060592 s, -0.001391 s, × 0.9776, 7 µs/rev mozilla-central x0000_revs_xx000_added_0_copies 9eec5917337d 67118cc6dcad : 615 revs, 0.019861 s, 0.019596 s, -0.000265 s, × 0.9867, 31 µs/rev mozilla-central x0000_revs_xx000_added_x000_copies f78c615a656c 96a38b690156 : 30263 revs, 0.188101 s, 0.183558 s, -0.004543 s, × 0.9758, 6 µs/rev mozilla-central x00000_revs_x0000_added_x0000_copies 6832ae71433c 4c222a1d9a00 : 153721 revs, 1.806696 s, 1.758083 s, -0.048613 s, × 0.9731, 11 µs/rev mozilla-central x00000_revs_x00000_added_x000_copies 76caed42cf7c 1daa622bbe42 : 204976 revs, 2.682987 s, 2.592955 s, -0.090032 s, × 0.9664, 12 µs/rev mozilla-try x_revs_x_added_0_copies aaf6dde0deb8 9790f499805a : 2 revs, 0.000852 s, 0.000844 s, -0.000008 s, × 0.9906, 422 µs/rev mozilla-try x_revs_x000_added_0_copies d8d0222927b4 5bb8ce8c7450 : 2 revs, 0.000859 s, 0.000861 s, +0.000002 s, × 1.0023, 430 µs/rev mozilla-try x_revs_x_added_x_copies 092fcca11bdb 936255a0384a : 4 revs, 0.000150 s, 0.000150 s, +0.000000 s, × 1.0000, 37 µs/rev mozilla-try x_revs_x00_added_x_copies b53d2fadbdb5 017afae788ec : 2 revs, 0.001158 s, 0.001166 s, +0.000008 s, × 1.0069, 583 µs/rev mozilla-try x_revs_x000_added_x000_copies 20408ad61ce5 6f0ee96e21ad : 1 revs, 0.027240 s, 0.027359 s, +0.000119 s, × 1.0044, 27359 µs/rev mozilla-try x_revs_x0000_added_x0000_copies effb563bb7e5 c07a39dc4e80 : 6 revs, 0.062824 s, 0.061848 s, -0.000976 s, × 0.9845, 10308 µs/rev mozilla-try x000_revs_xx00_added_0_copies 6100d773079a 04a55431795e : 1593 revs, 0.005463 s, 0.005110 s, -0.000353 s, × 0.9354, 3 µs/rev mozilla-try x000_revs_x000_added_x_copies 9f17a6fc04f9 2d37b966abed : 41 revs, 0.004238 s, 0.004168 s, -0.000070 s, × 0.9835, 101 µs/rev mozilla-try x000_revs_x000_added_x000_copies 1346fd0130e4 4c65cbdabc1f : 6657 revs, 0.064113 s, 0.063414 s, -0.000699 s, × 0.9891, 9 µs/rev mozilla-try x0000_revs_x_added_0_copies 63519bfd42ee a36a2a865d92 : 40314 revs, 0.294063 s, 0.288301 s, -0.005762 s, × 0.9804, 7 µs/rev mozilla-try x0000_revs_x_added_x_copies 9fe69ff0762d bcabf2a78927 : 38690 revs, 0.281493 s, 0.275798 s, -0.005695 s, × 0.9798, 7 µs/rev mozilla-try x0000_revs_xx000_added_x_copies 156f6e2674f2 4d0f2c178e66 : 8598 revs, 0.076323 s, 0.074640 s, -0.001683 s, × 0.9779, 8 µs/rev mozilla-try x0000_revs_xx000_added_0_copies 9eec5917337d 67118cc6dcad : 615 revs, 0.020390 s, 0.020327 s, -0.000063 s, × 0.9969, 33 µs/rev mozilla-try x0000_revs_xx000_added_x000_copies 89294cd501d9 7ccb2fc7ccb5 : 97052 revs, 3.023879 s, 2.970385 s, -0.053494 s, × 0.9823, 30 µs/rev mozilla-try x0000_revs_x0000_added_x0000_copies e928c65095ed e951f4ad123a : 52031 revs, 0.735549 s, 0.719432 s, -0.016117 s, × 0.9781, 13 µs/rev mozilla-try x00000_revs_x_added_0_copies 6a320851d377 1ebb79acd503 : 363753 revs, 18.568900 s, 18.165143 s, -0.403757 s, × 0.9783, 49 µs/rev mozilla-try x00000_revs_x00000_added_0_copies dc8a3ca7010e d16fde900c9c : 34414 revs, 0.502584 s, 0.486769 s, -0.015815 s, × 0.9685, 14 µs/rev mozilla-try x00000_revs_x_added_x_copies 5173c4b6f97c 95d83ee7242d : 362229 revs, 18.356645 s, 17.913924 s, -0.442721 s, × 0.9759, 49 µs/rev mozilla-try x00000_revs_x000_added_x_copies 9126823d0e9c ca82787bb23c : 359344 revs, 18.250393 s, 17.660113 s, -0.590280 s, × 0.9677, 49 µs/rev mozilla-try x00000_revs_x0000_added_x0000_copies 8d3fafa80d4b eb884023b810 : 192665 revs, 2.792459 s, 2.709446 s, -0.083013 s, × 0.9703, 14 µs/rev mozilla-try x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d : 228985 revs, 107.697264 s, 107.796891 s, +0.099627 s, × 1.0009, 470 µs/rev mozilla-try x00000_revs_x00000_added_x000_copies 9b2a99adc05e 8e29777b48e6 : 382065 revs, 63.961040 s, 63.575217 s, -0.385823 s, × 0.9940, 166 µs/rev Differential Revision: https://phab.mercurial-scm.org/D9423

File last commit:

r45500:26114bd6 default
r46765:a132aa59 default
Show More
nodemap.rs
1118 lines | 36.2 KiB | application/rls-services+xml | RustLexer
// Copyright 2018-2020 Georges Racinet <georges.racinet@octobus.net>
// and Mercurial contributors
//
// This software may be used and distributed according to the terms of the
// GNU General Public License version 2 or any later version.
//! Indexing facilities for fast retrieval of `Revision` from `Node`
//!
//! This provides a variation on the 16-ary radix tree that is
//! provided as "nodetree" in revlog.c, ready for append-only persistence
//! on disk.
//!
//! Following existing implicit conventions, the "nodemap" terminology
//! is used in a more abstract context.
use super::{
node::NULL_NODE, Node, NodeError, NodePrefix, NodePrefixRef, Revision,
RevlogIndex, NULL_REVISION,
};
use std::cmp::max;
use std::fmt;
use std::mem;
use std::ops::Deref;
use std::ops::Index;
use std::slice;
#[derive(Debug, PartialEq)]
pub enum NodeMapError {
MultipleResults,
InvalidNodePrefix(NodeError),
/// A `Revision` stored in the nodemap could not be found in the index
RevisionNotInIndex(Revision),
}
impl From<NodeError> for NodeMapError {
fn from(err: NodeError) -> Self {
NodeMapError::InvalidNodePrefix(err)
}
}
/// Mapping system from Mercurial nodes to revision numbers.
///
/// ## `RevlogIndex` and `NodeMap`
///
/// One way to think about their relationship is that
/// the `NodeMap` is a prefix-oriented reverse index of the `Node` information
/// carried by a [`RevlogIndex`].
///
/// Many of the methods in this trait take a `RevlogIndex` argument
/// which is used for validation of their results. This index must naturally
/// be the one the `NodeMap` is about, and it must be consistent.
///
/// Notably, the `NodeMap` must not store
/// information about more `Revision` values than there are in the index.
/// In these methods, an encountered `Revision` is not in the index, a
/// [`RevisionNotInIndex`] error is returned.
///
/// In insert operations, the rule is thus that the `NodeMap` must always
/// be updated after the `RevlogIndex`
/// be updated first, and the `NodeMap` second.
///
/// [`RevisionNotInIndex`]: enum.NodeMapError.html#variant.RevisionNotInIndex
/// [`RevlogIndex`]: ../trait.RevlogIndex.html
pub trait NodeMap {
/// Find the unique `Revision` having the given `Node`
///
/// If no Revision matches the given `Node`, `Ok(None)` is returned.
fn find_node(
&self,
index: &impl RevlogIndex,
node: &Node,
) -> Result<Option<Revision>, NodeMapError> {
self.find_bin(index, node.into())
}
/// Find the unique Revision whose `Node` starts with a given binary prefix
///
/// If no Revision matches the given prefix, `Ok(None)` is returned.
///
/// If several Revisions match the given prefix, a [`MultipleResults`]
/// error is returned.
fn find_bin<'a>(
&self,
idx: &impl RevlogIndex,
prefix: NodePrefixRef<'a>,
) -> Result<Option<Revision>, NodeMapError>;
/// Find the unique Revision whose `Node` hexadecimal string representation
/// starts with a given prefix
///
/// If no Revision matches the given prefix, `Ok(None)` is returned.
///
/// If several Revisions match the given prefix, a [`MultipleResults`]
/// error is returned.
fn find_hex(
&self,
idx: &impl RevlogIndex,
prefix: &str,
) -> Result<Option<Revision>, NodeMapError> {
self.find_bin(idx, NodePrefix::from_hex(prefix)?.borrow())
}
/// Give the size of the shortest node prefix that determines
/// the revision uniquely.
///
/// From a binary node prefix, if it is matched in the node map, this
/// returns the number of hexadecimal digits that would had sufficed
/// to find the revision uniquely.
///
/// Returns `None` if no `Revision` could be found for the prefix.
///
/// If several Revisions match the given prefix, a [`MultipleResults`]
/// error is returned.
fn unique_prefix_len_bin<'a>(
&self,
idx: &impl RevlogIndex,
node_prefix: NodePrefixRef<'a>,
) -> Result<Option<usize>, NodeMapError>;
/// Same as `unique_prefix_len_bin`, with the hexadecimal representation
/// of the prefix as input.
fn unique_prefix_len_hex(
&self,
idx: &impl RevlogIndex,
prefix: &str,
) -> Result<Option<usize>, NodeMapError> {
self.unique_prefix_len_bin(idx, NodePrefix::from_hex(prefix)?.borrow())
}
/// Same as `unique_prefix_len_bin`, with a full `Node` as input
fn unique_prefix_len_node(
&self,
idx: &impl RevlogIndex,
node: &Node,
) -> Result<Option<usize>, NodeMapError> {
self.unique_prefix_len_bin(idx, node.into())
}
}
pub trait MutableNodeMap: NodeMap {
fn insert<I: RevlogIndex>(
&mut self,
index: &I,
node: &Node,
rev: Revision,
) -> Result<(), NodeMapError>;
}
/// Low level NodeTree [`Blocks`] elements
///
/// These are exactly as for instance on persistent storage.
type RawElement = i32;
/// High level representation of values in NodeTree
/// [`Blocks`](struct.Block.html)
///
/// This is the high level representation that most algorithms should
/// use.
#[derive(Clone, Debug, Eq, PartialEq)]
enum Element {
Rev(Revision),
Block(usize),
None,
}
impl From<RawElement> for Element {
/// Conversion from low level representation, after endianness conversion.
///
/// See [`Block`](struct.Block.html) for explanation about the encoding.
fn from(raw: RawElement) -> Element {
if raw >= 0 {
Element::Block(raw as usize)
} else if raw == -1 {
Element::None
} else {
Element::Rev(-raw - 2)
}
}
}
impl From<Element> for RawElement {
fn from(element: Element) -> RawElement {
match element {
Element::None => 0,
Element::Block(i) => i as RawElement,
Element::Rev(rev) => -rev - 2,
}
}
}
/// A logical block of the `NodeTree`, packed with a fixed size.
///
/// These are always used in container types implementing `Index<Block>`,
/// such as `&Block`
///
/// As an array of integers, its ith element encodes that the
/// ith potential edge from the block, representing the ith hexadecimal digit
/// (nybble) `i` is either:
///
/// - absent (value -1)
/// - another `Block` in the same indexable container (value ≥ 0)
/// - a `Revision` leaf (value ≤ -2)
///
/// Endianness has to be fixed for consistency on shared storage across
/// different architectures.
///
/// A key difference with the C `nodetree` is that we need to be
/// able to represent the [`Block`] at index 0, hence -1 is the empty marker
/// rather than 0 and the `Revision` range upper limit of -2 instead of -1.
///
/// Another related difference is that `NULL_REVISION` (-1) is not
/// represented at all, because we want an immutable empty nodetree
/// to be valid.
#[derive(Copy, Clone)]
pub struct Block([u8; BLOCK_SIZE]);
/// Not derivable for arrays of length >32 until const generics are stable
impl PartialEq for Block {
fn eq(&self, other: &Self) -> bool {
self.0[..] == other.0[..]
}
}
pub const BLOCK_SIZE: usize = 64;
impl Block {
fn new() -> Self {
// -1 in 2's complement to create an absent node
let byte: u8 = 255;
Block([byte; BLOCK_SIZE])
}
fn get(&self, nybble: u8) -> Element {
let index = nybble as usize * mem::size_of::<RawElement>();
Element::from(RawElement::from_be_bytes([
self.0[index],
self.0[index + 1],
self.0[index + 2],
self.0[index + 3],
]))
}
fn set(&mut self, nybble: u8, element: Element) {
let values = RawElement::to_be_bytes(element.into());
let index = nybble as usize * mem::size_of::<RawElement>();
self.0[index] = values[0];
self.0[index + 1] = values[1];
self.0[index + 2] = values[2];
self.0[index + 3] = values[3];
}
}
impl fmt::Debug for Block {
/// sparse representation for testing and debugging purposes
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
f.debug_map()
.entries((0..16).filter_map(|i| match self.get(i) {
Element::None => None,
element => Some((i, element)),
}))
.finish()
}
}
/// A mutable 16-radix tree with the root block logically at the end
///
/// Because of the append only nature of our node trees, we need to
/// keep the original untouched and store new blocks separately.
///
/// The mutable root `Block` is kept apart so that we don't have to rebump
/// it on each insertion.
pub struct NodeTree {
readonly: Box<dyn Deref<Target = [Block]> + Send>,
growable: Vec<Block>,
root: Block,
masked_inner_blocks: usize,
}
impl Index<usize> for NodeTree {
type Output = Block;
fn index(&self, i: usize) -> &Block {
let ro_len = self.readonly.len();
if i < ro_len {
&self.readonly[i]
} else if i == ro_len + self.growable.len() {
&self.root
} else {
&self.growable[i - ro_len]
}
}
}
/// Return `None` unless the `Node` for `rev` has given prefix in `index`.
fn has_prefix_or_none(
idx: &impl RevlogIndex,
prefix: NodePrefixRef,
rev: Revision,
) -> Result<Option<Revision>, NodeMapError> {
idx.node(rev)
.ok_or_else(|| NodeMapError::RevisionNotInIndex(rev))
.map(|node| {
if prefix.is_prefix_of(node) {
Some(rev)
} else {
None
}
})
}
/// validate that the candidate's node starts indeed with given prefix,
/// and treat ambiguities related to `NULL_REVISION`.
///
/// From the data in the NodeTree, one can only conclude that some
/// revision is the only one for a *subprefix* of the one being looked up.
fn validate_candidate(
idx: &impl RevlogIndex,
prefix: NodePrefixRef,
candidate: (Option<Revision>, usize),
) -> Result<(Option<Revision>, usize), NodeMapError> {
let (rev, steps) = candidate;
if let Some(nz_nybble) = prefix.first_different_nybble(&NULL_NODE) {
rev.map_or(Ok((None, steps)), |r| {
has_prefix_or_none(idx, prefix, r)
.map(|opt| (opt, max(steps, nz_nybble + 1)))
})
} else {
// the prefix is only made of zeros; NULL_REVISION always matches it
// and any other *valid* result is an ambiguity
match rev {
None => Ok((Some(NULL_REVISION), steps + 1)),
Some(r) => match has_prefix_or_none(idx, prefix, r)? {
None => Ok((Some(NULL_REVISION), steps + 1)),
_ => Err(NodeMapError::MultipleResults),
},
}
}
}
impl NodeTree {
/// Initiate a NodeTree from an immutable slice-like of `Block`
///
/// We keep `readonly` and clone its root block if it isn't empty.
fn new(readonly: Box<dyn Deref<Target = [Block]> + Send>) -> Self {
let root = readonly.last().cloned().unwrap_or_else(Block::new);
NodeTree {
readonly,
growable: Vec::new(),
root,
masked_inner_blocks: 0,
}
}
/// Create from an opaque bunch of bytes
///
/// The created `NodeTreeBytes` from `buffer`,
/// of which exactly `amount` bytes are used.
///
/// - `buffer` could be derived from `PyBuffer` and `Mmap` objects.
/// - `offset` allows for the final file format to include fixed data
/// (generation number, behavioural flags)
/// - `amount` is expressed in bytes, and is not automatically derived from
/// `bytes`, so that a caller that manages them atomically can perform
/// temporary disk serializations and still rollback easily if needed.
/// First use-case for this would be to support Mercurial shell hooks.
///
/// panics if `buffer` is smaller than `amount`
pub fn load_bytes(
bytes: Box<dyn Deref<Target = [u8]> + Send>,
amount: usize,
) -> Self {
NodeTree::new(Box::new(NodeTreeBytes::new(bytes, amount)))
}
/// Retrieve added `Block` and the original immutable data
pub fn into_readonly_and_added(
self,
) -> (Box<dyn Deref<Target = [Block]> + Send>, Vec<Block>) {
let mut vec = self.growable;
let readonly = self.readonly;
if readonly.last() != Some(&self.root) {
vec.push(self.root);
}
(readonly, vec)
}
/// Retrieve added `Blocks` as bytes, ready to be written to persistent
/// storage
pub fn into_readonly_and_added_bytes(
self,
) -> (Box<dyn Deref<Target = [Block]> + Send>, Vec<u8>) {
let (readonly, vec) = self.into_readonly_and_added();
// Prevent running `v`'s destructor so we are in complete control
// of the allocation.
let vec = mem::ManuallyDrop::new(vec);
// Transmute the `Vec<Block>` to a `Vec<u8>`. Blocks are contiguous
// bytes, so this is perfectly safe.
let bytes = unsafe {
// Assert that `Block` hasn't been changed and has no padding
let _: [u8; 4 * BLOCK_SIZE] =
std::mem::transmute([Block::new(); 4]);
// /!\ Any use of `vec` after this is use-after-free.
// TODO: use `into_raw_parts` once stabilized
Vec::from_raw_parts(
vec.as_ptr() as *mut u8,
vec.len() * BLOCK_SIZE,
vec.capacity() * BLOCK_SIZE,
)
};
(readonly, bytes)
}
/// Total number of blocks
fn len(&self) -> usize {
self.readonly.len() + self.growable.len() + 1
}
/// Implemented for completeness
///
/// A `NodeTree` always has at least the mutable root block.
#[allow(dead_code)]
fn is_empty(&self) -> bool {
false
}
/// Main working method for `NodeTree` searches
///
/// The first returned value is the result of analysing `NodeTree` data
/// *alone*: whereas `None` guarantees that the given prefix is absent
/// from the `NodeTree` data (but still could match `NULL_NODE`), with
/// `Some(rev)`, it is to be understood that `rev` is the unique `Revision`
/// that could match the prefix. Actually, all that can be inferred from
/// the `NodeTree` data is that `rev` is the revision with the longest
/// common node prefix with the given prefix.
///
/// The second returned value is the size of the smallest subprefix
/// of `prefix` that would give the same result, i.e. not the
/// `MultipleResults` error variant (again, using only the data of the
/// `NodeTree`).
fn lookup(
&self,
prefix: NodePrefixRef,
) -> Result<(Option<Revision>, usize), NodeMapError> {
for (i, visit_item) in self.visit(prefix).enumerate() {
if let Some(opt) = visit_item.final_revision() {
return Ok((opt, i + 1));
}
}
Err(NodeMapError::MultipleResults)
}
fn visit<'n, 'p>(
&'n self,
prefix: NodePrefixRef<'p>,
) -> NodeTreeVisitor<'n, 'p> {
NodeTreeVisitor {
nt: self,
prefix,
visit: self.len() - 1,
nybble_idx: 0,
done: false,
}
}
/// Return a mutable reference for `Block` at index `idx`.
///
/// If `idx` lies in the immutable area, then the reference is to
/// a newly appended copy.
///
/// Returns (new_idx, glen, mut_ref) where
///
/// - `new_idx` is the index of the mutable `Block`
/// - `mut_ref` is a mutable reference to the mutable Block.
/// - `glen` is the new length of `self.growable`
///
/// Note: the caller wouldn't be allowed to query `self.growable.len()`
/// itself because of the mutable borrow taken with the returned `Block`
fn mutable_block(&mut self, idx: usize) -> (usize, &mut Block, usize) {
let ro_blocks = &self.readonly;
let ro_len = ro_blocks.len();
let glen = self.growable.len();
if idx < ro_len {
self.masked_inner_blocks += 1;
self.growable.push(ro_blocks[idx]);
(glen + ro_len, &mut self.growable[glen], glen + 1)
} else if glen + ro_len == idx {
(idx, &mut self.root, glen)
} else {
(idx, &mut self.growable[idx - ro_len], glen)
}
}
/// Main insertion method
///
/// This will dive in the node tree to find the deepest `Block` for
/// `node`, split it as much as needed and record `node` in there.
/// The method then backtracks, updating references in all the visited
/// blocks from the root.
///
/// All the mutated `Block` are copied first to the growable part if
/// needed. That happens for those in the immutable part except the root.
pub fn insert<I: RevlogIndex>(
&mut self,
index: &I,
node: &Node,
rev: Revision,
) -> Result<(), NodeMapError> {
let ro_len = &self.readonly.len();
let mut visit_steps: Vec<_> = self.visit(node.into()).collect();
let read_nybbles = visit_steps.len();
// visit_steps cannot be empty, since we always visit the root block
let deepest = visit_steps.pop().unwrap();
let (mut block_idx, mut block, mut glen) =
self.mutable_block(deepest.block_idx);
if let Element::Rev(old_rev) = deepest.element {
let old_node = index
.node(old_rev)
.ok_or_else(|| NodeMapError::RevisionNotInIndex(old_rev))?;
if old_node == node {
return Ok(()); // avoid creating lots of useless blocks
}
// Looping over the tail of nybbles in both nodes, creating
// new blocks until we find the difference
let mut new_block_idx = ro_len + glen;
let mut nybble = deepest.nybble;
for nybble_pos in read_nybbles..node.nybbles_len() {
block.set(nybble, Element::Block(new_block_idx));
let new_nybble = node.get_nybble(nybble_pos);
let old_nybble = old_node.get_nybble(nybble_pos);
if old_nybble == new_nybble {
self.growable.push(Block::new());
block = &mut self.growable[glen];
glen += 1;
new_block_idx += 1;
nybble = new_nybble;
} else {
let mut new_block = Block::new();
new_block.set(old_nybble, Element::Rev(old_rev));
new_block.set(new_nybble, Element::Rev(rev));
self.growable.push(new_block);
break;
}
}
} else {
// Free slot in the deepest block: no splitting has to be done
block.set(deepest.nybble, Element::Rev(rev));
}
// Backtrack over visit steps to update references
while let Some(visited) = visit_steps.pop() {
let to_write = Element::Block(block_idx);
if visit_steps.is_empty() {
self.root.set(visited.nybble, to_write);
break;
}
let (new_idx, block, _) = self.mutable_block(visited.block_idx);
if block.get(visited.nybble) == to_write {
break;
}
block.set(visited.nybble, to_write);
block_idx = new_idx;
}
Ok(())
}
/// Make the whole `NodeTree` logically empty, without touching the
/// immutable part.
pub fn invalidate_all(&mut self) {
self.root = Block::new();
self.growable = Vec::new();
self.masked_inner_blocks = self.readonly.len();
}
/// Return the number of blocks in the readonly part that are currently
/// masked in the mutable part.
///
/// The `NodeTree` structure has no efficient way to know how many blocks
/// are already unreachable in the readonly part.
///
/// After a call to `invalidate_all()`, the returned number can be actually
/// bigger than the whole readonly part, a conventional way to mean that
/// all the readonly blocks have been masked. This is what is really
/// useful to the caller and does not require to know how many were
/// actually unreachable to begin with.
pub fn masked_readonly_blocks(&self) -> usize {
if let Some(readonly_root) = self.readonly.last() {
if readonly_root == &self.root {
return 0;
}
} else {
return 0;
}
self.masked_inner_blocks + 1
}
}
pub struct NodeTreeBytes {
buffer: Box<dyn Deref<Target = [u8]> + Send>,
len_in_blocks: usize,
}
impl NodeTreeBytes {
fn new(
buffer: Box<dyn Deref<Target = [u8]> + Send>,
amount: usize,
) -> Self {
assert!(buffer.len() >= amount);
let len_in_blocks = amount / BLOCK_SIZE;
NodeTreeBytes {
buffer,
len_in_blocks,
}
}
}
impl Deref for NodeTreeBytes {
type Target = [Block];
fn deref(&self) -> &[Block] {
unsafe {
slice::from_raw_parts(
(&self.buffer).as_ptr() as *const Block,
self.len_in_blocks,
)
}
}
}
struct NodeTreeVisitor<'n, 'p> {
nt: &'n NodeTree,
prefix: NodePrefixRef<'p>,
visit: usize,
nybble_idx: usize,
done: bool,
}
#[derive(Debug, PartialEq, Clone)]
struct NodeTreeVisitItem {
block_idx: usize,
nybble: u8,
element: Element,
}
impl<'n, 'p> Iterator for NodeTreeVisitor<'n, 'p> {
type Item = NodeTreeVisitItem;
fn next(&mut self) -> Option<Self::Item> {
if self.done || self.nybble_idx >= self.prefix.len() {
return None;
}
let nybble = self.prefix.get_nybble(self.nybble_idx);
self.nybble_idx += 1;
let visit = self.visit;
let element = self.nt[visit].get(nybble);
if let Element::Block(idx) = element {
self.visit = idx;
} else {
self.done = true;
}
Some(NodeTreeVisitItem {
block_idx: visit,
nybble,
element,
})
}
}
impl NodeTreeVisitItem {
// Return `Some(opt)` if this item is final, with `opt` being the
// `Revision` that it may represent.
//
// If the item is not terminal, return `None`
fn final_revision(&self) -> Option<Option<Revision>> {
match self.element {
Element::Block(_) => None,
Element::Rev(r) => Some(Some(r)),
Element::None => Some(None),
}
}
}
impl From<Vec<Block>> for NodeTree {
fn from(vec: Vec<Block>) -> Self {
Self::new(Box::new(vec))
}
}
impl fmt::Debug for NodeTree {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
let readonly: &[Block] = &*self.readonly;
write!(
f,
"readonly: {:?}, growable: {:?}, root: {:?}",
readonly, self.growable, self.root
)
}
}
impl Default for NodeTree {
/// Create a fully mutable empty NodeTree
fn default() -> Self {
NodeTree::new(Box::new(Vec::new()))
}
}
impl NodeMap for NodeTree {
fn find_bin<'a>(
&self,
idx: &impl RevlogIndex,
prefix: NodePrefixRef<'a>,
) -> Result<Option<Revision>, NodeMapError> {
validate_candidate(idx, prefix.clone(), self.lookup(prefix)?)
.map(|(opt, _shortest)| opt)
}
fn unique_prefix_len_bin<'a>(
&self,
idx: &impl RevlogIndex,
prefix: NodePrefixRef<'a>,
) -> Result<Option<usize>, NodeMapError> {
validate_candidate(idx, prefix.clone(), self.lookup(prefix)?)
.map(|(opt, shortest)| opt.map(|_rev| shortest))
}
}
#[cfg(test)]
mod tests {
use super::NodeMapError::*;
use super::*;
use crate::revlog::node::{hex_pad_right, Node};
use std::collections::HashMap;
/// Creates a `Block` using a syntax close to the `Debug` output
macro_rules! block {
{$($nybble:tt : $variant:ident($val:tt)),*} => (
{
let mut block = Block::new();
$(block.set($nybble, Element::$variant($val)));*;
block
}
)
}
#[test]
fn test_block_debug() {
let mut block = Block::new();
block.set(1, Element::Rev(3));
block.set(10, Element::Block(0));
assert_eq!(format!("{:?}", block), "{1: Rev(3), 10: Block(0)}");
}
#[test]
fn test_block_macro() {
let block = block! {5: Block(2)};
assert_eq!(format!("{:?}", block), "{5: Block(2)}");
let block = block! {13: Rev(15), 5: Block(2)};
assert_eq!(format!("{:?}", block), "{5: Block(2), 13: Rev(15)}");
}
#[test]
fn test_raw_block() {
let mut raw = [255u8; 64];
let mut counter = 0;
for val in [0, 15, -2, -1, -3].iter() {
for byte in RawElement::to_be_bytes(*val).iter() {
raw[counter] = *byte;
counter += 1;
}
}
let block = Block(raw);
assert_eq!(block.get(0), Element::Block(0));
assert_eq!(block.get(1), Element::Block(15));
assert_eq!(block.get(3), Element::None);
assert_eq!(block.get(2), Element::Rev(0));
assert_eq!(block.get(4), Element::Rev(1));
}
type TestIndex = HashMap<Revision, Node>;
impl RevlogIndex for TestIndex {
fn node(&self, rev: Revision) -> Option<&Node> {
self.get(&rev)
}
fn len(&self) -> usize {
self.len()
}
}
/// Pad hexadecimal Node prefix with zeros on the right
///
/// This avoids having to repeatedly write very long hexadecimal
/// strings for test data, and brings actual hash size independency.
#[cfg(test)]
fn pad_node(hex: &str) -> Node {
Node::from_hex(&hex_pad_right(hex)).unwrap()
}
/// Pad hexadecimal Node prefix with zeros on the right, then insert
fn pad_insert(idx: &mut TestIndex, rev: Revision, hex: &str) {
idx.insert(rev, pad_node(hex));
}
fn sample_nodetree() -> NodeTree {
NodeTree::from(vec![
block![0: Rev(9)],
block![0: Rev(0), 1: Rev(9)],
block![0: Block(1), 1:Rev(1)],
])
}
#[test]
fn test_nt_debug() {
let nt = sample_nodetree();
assert_eq!(
format!("{:?}", nt),
"readonly: \
[{0: Rev(9)}, {0: Rev(0), 1: Rev(9)}, {0: Block(1), 1: Rev(1)}], \
growable: [], \
root: {0: Block(1), 1: Rev(1)}",
);
}
#[test]
fn test_immutable_find_simplest() -> Result<(), NodeMapError> {
let mut idx: TestIndex = HashMap::new();
pad_insert(&mut idx, 1, "1234deadcafe");
let nt = NodeTree::from(vec![block! {1: Rev(1)}]);
assert_eq!(nt.find_hex(&idx, "1")?, Some(1));
assert_eq!(nt.find_hex(&idx, "12")?, Some(1));
assert_eq!(nt.find_hex(&idx, "1234de")?, Some(1));
assert_eq!(nt.find_hex(&idx, "1a")?, None);
assert_eq!(nt.find_hex(&idx, "ab")?, None);
// and with full binary Nodes
assert_eq!(nt.find_node(&idx, idx.get(&1).unwrap())?, Some(1));
let unknown = Node::from_hex(&hex_pad_right("3d")).unwrap();
assert_eq!(nt.find_node(&idx, &unknown)?, None);
Ok(())
}
#[test]
fn test_immutable_find_one_jump() {
let mut idx = TestIndex::new();
pad_insert(&mut idx, 9, "012");
pad_insert(&mut idx, 0, "00a");
let nt = sample_nodetree();
assert_eq!(nt.find_hex(&idx, "0"), Err(MultipleResults));
assert_eq!(nt.find_hex(&idx, "01"), Ok(Some(9)));
assert_eq!(nt.find_hex(&idx, "00"), Err(MultipleResults));
assert_eq!(nt.find_hex(&idx, "00a"), Ok(Some(0)));
assert_eq!(nt.unique_prefix_len_hex(&idx, "00a"), Ok(Some(3)));
assert_eq!(nt.find_hex(&idx, "000"), Ok(Some(NULL_REVISION)));
}
#[test]
fn test_mutated_find() -> Result<(), NodeMapError> {
let mut idx = TestIndex::new();
pad_insert(&mut idx, 9, "012");
pad_insert(&mut idx, 0, "00a");
pad_insert(&mut idx, 2, "cafe");
pad_insert(&mut idx, 3, "15");
pad_insert(&mut idx, 1, "10");
let nt = NodeTree {
readonly: sample_nodetree().readonly,
growable: vec![block![0: Rev(1), 5: Rev(3)]],
root: block![0: Block(1), 1:Block(3), 12: Rev(2)],
masked_inner_blocks: 1,
};
assert_eq!(nt.find_hex(&idx, "10")?, Some(1));
assert_eq!(nt.find_hex(&idx, "c")?, Some(2));
assert_eq!(nt.unique_prefix_len_hex(&idx, "c")?, Some(1));
assert_eq!(nt.find_hex(&idx, "00"), Err(MultipleResults));
assert_eq!(nt.find_hex(&idx, "000")?, Some(NULL_REVISION));
assert_eq!(nt.unique_prefix_len_hex(&idx, "000")?, Some(3));
assert_eq!(nt.find_hex(&idx, "01")?, Some(9));
assert_eq!(nt.masked_readonly_blocks(), 2);
Ok(())
}
struct TestNtIndex {
index: TestIndex,
nt: NodeTree,
}
impl TestNtIndex {
fn new() -> Self {
TestNtIndex {
index: HashMap::new(),
nt: NodeTree::default(),
}
}
fn insert(
&mut self,
rev: Revision,
hex: &str,
) -> Result<(), NodeMapError> {
let node = pad_node(hex);
self.index.insert(rev, node.clone());
self.nt.insert(&self.index, &node, rev)?;
Ok(())
}
fn find_hex(
&self,
prefix: &str,
) -> Result<Option<Revision>, NodeMapError> {
self.nt.find_hex(&self.index, prefix)
}
fn unique_prefix_len_hex(
&self,
prefix: &str,
) -> Result<Option<usize>, NodeMapError> {
self.nt.unique_prefix_len_hex(&self.index, prefix)
}
/// Drain `added` and restart a new one
fn commit(self) -> Self {
let mut as_vec: Vec<Block> =
self.nt.readonly.iter().map(|block| block.clone()).collect();
as_vec.extend(self.nt.growable);
as_vec.push(self.nt.root);
Self {
index: self.index,
nt: NodeTree::from(as_vec).into(),
}
}
}
#[test]
fn test_insert_full_mutable() -> Result<(), NodeMapError> {
let mut idx = TestNtIndex::new();
idx.insert(0, "1234")?;
assert_eq!(idx.find_hex("1")?, Some(0));
assert_eq!(idx.find_hex("12")?, Some(0));
// let's trigger a simple split
idx.insert(1, "1a34")?;
assert_eq!(idx.nt.growable.len(), 1);
assert_eq!(idx.find_hex("12")?, Some(0));
assert_eq!(idx.find_hex("1a")?, Some(1));
// reinserting is a no_op
idx.insert(1, "1a34")?;
assert_eq!(idx.nt.growable.len(), 1);
assert_eq!(idx.find_hex("12")?, Some(0));
assert_eq!(idx.find_hex("1a")?, Some(1));
idx.insert(2, "1a01")?;
assert_eq!(idx.nt.growable.len(), 2);
assert_eq!(idx.find_hex("1a"), Err(NodeMapError::MultipleResults));
assert_eq!(idx.find_hex("12")?, Some(0));
assert_eq!(idx.find_hex("1a3")?, Some(1));
assert_eq!(idx.find_hex("1a0")?, Some(2));
assert_eq!(idx.find_hex("1a12")?, None);
// now let's make it split and create more than one additional block
idx.insert(3, "1a345")?;
assert_eq!(idx.nt.growable.len(), 4);
assert_eq!(idx.find_hex("1a340")?, Some(1));
assert_eq!(idx.find_hex("1a345")?, Some(3));
assert_eq!(idx.find_hex("1a341")?, None);
// there's no readonly block to mask
assert_eq!(idx.nt.masked_readonly_blocks(), 0);
Ok(())
}
#[test]
fn test_unique_prefix_len_zero_prefix() {
let mut idx = TestNtIndex::new();
idx.insert(0, "00000abcd").unwrap();
assert_eq!(idx.find_hex("000"), Err(NodeMapError::MultipleResults));
// in the nodetree proper, this will be found at the first nybble
// yet the correct answer for unique_prefix_len is not 1, nor 1+1,
// but the first difference with `NULL_NODE`
assert_eq!(idx.unique_prefix_len_hex("00000a"), Ok(Some(6)));
assert_eq!(idx.unique_prefix_len_hex("00000ab"), Ok(Some(6)));
// same with odd result
idx.insert(1, "00123").unwrap();
assert_eq!(idx.unique_prefix_len_hex("001"), Ok(Some(3)));
assert_eq!(idx.unique_prefix_len_hex("0012"), Ok(Some(3)));
// these are unchanged of course
assert_eq!(idx.unique_prefix_len_hex("00000a"), Ok(Some(6)));
assert_eq!(idx.unique_prefix_len_hex("00000ab"), Ok(Some(6)));
}
#[test]
fn test_insert_extreme_splitting() -> Result<(), NodeMapError> {
// check that the splitting loop is long enough
let mut nt_idx = TestNtIndex::new();
let nt = &mut nt_idx.nt;
let idx = &mut nt_idx.index;
let node0_hex = hex_pad_right("444444");
let mut node1_hex = hex_pad_right("444444").clone();
node1_hex.pop();
node1_hex.push('5');
let node0 = Node::from_hex(&node0_hex).unwrap();
let node1 = Node::from_hex(&node1_hex).unwrap();
idx.insert(0, node0.clone());
nt.insert(idx, &node0, 0)?;
idx.insert(1, node1.clone());
nt.insert(idx, &node1, 1)?;
assert_eq!(nt.find_bin(idx, (&node0).into())?, Some(0));
assert_eq!(nt.find_bin(idx, (&node1).into())?, Some(1));
Ok(())
}
#[test]
fn test_insert_partly_immutable() -> Result<(), NodeMapError> {
let mut idx = TestNtIndex::new();
idx.insert(0, "1234")?;
idx.insert(1, "1235")?;
idx.insert(2, "131")?;
idx.insert(3, "cafe")?;
let mut idx = idx.commit();
assert_eq!(idx.find_hex("1234")?, Some(0));
assert_eq!(idx.find_hex("1235")?, Some(1));
assert_eq!(idx.find_hex("131")?, Some(2));
assert_eq!(idx.find_hex("cafe")?, Some(3));
// we did not add anything since init from readonly
assert_eq!(idx.nt.masked_readonly_blocks(), 0);
idx.insert(4, "123A")?;
assert_eq!(idx.find_hex("1234")?, Some(0));
assert_eq!(idx.find_hex("1235")?, Some(1));
assert_eq!(idx.find_hex("131")?, Some(2));
assert_eq!(idx.find_hex("cafe")?, Some(3));
assert_eq!(idx.find_hex("123A")?, Some(4));
// we masked blocks for all prefixes of "123", including the root
assert_eq!(idx.nt.masked_readonly_blocks(), 4);
eprintln!("{:?}", idx.nt);
idx.insert(5, "c0")?;
assert_eq!(idx.find_hex("cafe")?, Some(3));
assert_eq!(idx.find_hex("c0")?, Some(5));
assert_eq!(idx.find_hex("c1")?, None);
assert_eq!(idx.find_hex("1234")?, Some(0));
// inserting "c0" is just splitting the 'c' slot of the mutable root,
// it doesn't mask anything
assert_eq!(idx.nt.masked_readonly_blocks(), 4);
Ok(())
}
#[test]
fn test_invalidate_all() -> Result<(), NodeMapError> {
let mut idx = TestNtIndex::new();
idx.insert(0, "1234")?;
idx.insert(1, "1235")?;
idx.insert(2, "131")?;
idx.insert(3, "cafe")?;
let mut idx = idx.commit();
idx.nt.invalidate_all();
assert_eq!(idx.find_hex("1234")?, None);
assert_eq!(idx.find_hex("1235")?, None);
assert_eq!(idx.find_hex("131")?, None);
assert_eq!(idx.find_hex("cafe")?, None);
// all the readonly blocks have been masked, this is the
// conventional expected response
assert_eq!(idx.nt.masked_readonly_blocks(), idx.nt.readonly.len() + 1);
Ok(())
}
#[test]
fn test_into_added_empty() {
assert!(sample_nodetree().into_readonly_and_added().1.is_empty());
assert!(sample_nodetree()
.into_readonly_and_added_bytes()
.1
.is_empty());
}
#[test]
fn test_into_added_bytes() -> Result<(), NodeMapError> {
let mut idx = TestNtIndex::new();
idx.insert(0, "1234")?;
let mut idx = idx.commit();
idx.insert(4, "cafe")?;
let (_, bytes) = idx.nt.into_readonly_and_added_bytes();
// only the root block has been changed
assert_eq!(bytes.len(), BLOCK_SIZE);
// big endian for -2
assert_eq!(&bytes[4..2 * 4], [255, 255, 255, 254]);
// big endian for -6
assert_eq!(&bytes[12 * 4..13 * 4], [255, 255, 255, 250]);
Ok(())
}
}