rust-nodemap: pure Rust example

To run, use `cargo run --release --example nodemap`

This demonstrates that simple scenarios entirely written
in Rust can content themselves with `NodeTree<T>`.

The example mmaps both the nodemap file and the changelog index.
We had of course to include an implementation of `RevlogIndex`
directly, which isn't much at this stage. It felt a bit
prematurate to include it in the lib.

Here are some first performance measurements, obtained with
this example, on a clone of mozilla-central with 440000
changesets:

  (create) Nodemap constructed in RAM in 153.638305ms
  (query CAE63161B68962) found in 22.362us: Ok(Some(269489))
  (bench) Did 3 queries in 36.418µs (mean 12.139µs)
  (bench) Did 50 queries in 184.318µs (mean 3.686µs)
  (bench) Did 100000 queries in 31.053461ms (mean 310ns)

To be fair, even between bench runs, results tend to depend whether
the file is still in kernel caches, and it's not so easy to
get back to a real cold start. The worst we've seen was in the
50us ballpark.

In any busy server setting, the pages would always be in RAM.

We hope it's good enough not to be significantly slower on any
concrete Mercurial operation than the C nodetree when fully in RAM,
and of course this implementation has the serious headstart advantage
of persistence.

Differential Revision: https://phab.mercurial-scm.org/D7797

rust/hg-core/examples/nodemap/index.rs

@@ -0,0 +1,95 b''
+// Copyright 2019-2020 Georges Racinet <[e-mail]>
+//
+// This software may be used and distributed according to the terms of the
+// GNU General Public License version 2 or any later version.
+
+//! Minimal `RevlogIndex`, readable from standard Mercurial file format
+use hg::*;
+use memmap::*;
+use std::fs::File;
+use std::ops::Deref;
+use std::path::Path;
+use std::slice;
+
+pub struct Index {
+    data: Box<dyn Deref<Target = [IndexEntry]> + Send>,
+}
+
+/// A fixed sized index entry. All numbers are big endian
+#[repr(C)]
+pub struct IndexEntry {
+    not_used_yet: [u8; 24],
+    p1: Revision,
+    p2: Revision,
+    node: Node,
+    unused_node: [u8; 12],
+}
+
+pub const INDEX_ENTRY_SIZE: usize = 64;
+
+impl IndexEntry {
+    fn parents(&self) -> [Revision; 2] {
+        [Revision::from_be(self.p1), Revision::from_be(self.p1)]
+    }
+}
+
+impl RevlogIndex for Index {
+    fn len(&self) -> usize {
+        self.data.len()
+    }
+
+    fn node(&self, rev: Revision) -> Option<&Node> {
+        if rev == NULL_REVISION {
+            return None;
+        }
+        let i = rev as usize;
+        if i >= self.len() {
+            None
+        } else {
+            Some(&self.data[i].node)
+        }
+    }
+}
+
+impl Graph for &Index {
+    fn parents(&self, rev: Revision) -> Result<[Revision; 2], GraphError> {
+        let [p1, p2] = (*self).data[rev as usize].parents();
+        let len = (*self).len();
+        if p1 < NULL_REVISION
+            || p2 < NULL_REVISION
+            || p1 as usize >= len
+            || p2 as usize >= len
+        {
+            return Err(GraphError::ParentOutOfRange(rev));
+        }
+        Ok([p1, p2])
+    }
+}
+
+struct IndexMmap(Mmap);
+
+impl Deref for IndexMmap {
+    type Target = [IndexEntry];
+
+    fn deref(&self) -> &[IndexEntry] {
+        let ptr = self.0.as_ptr() as *const IndexEntry;
+        // Any misaligned data will be ignored.
+        debug_assert_eq!(
+            self.0.len() % std::mem::align_of::<IndexEntry>(),
+            0,
+            "Misaligned data in mmap"
+        );
+        unsafe { slice::from_raw_parts(ptr, self.0.len() / INDEX_ENTRY_SIZE) }
+    }
+}
+
+impl Index {
+    pub fn load_mmap(path: impl AsRef<Path>) -> Self {
+        let file = File::open(path).unwrap();
+        let msg = "Index file is missing, or missing permission";
+        let mmap = unsafe { MmapOptions::new().map(&file) }.expect(msg);
+        Self {
+            data: Box::new(IndexMmap(mmap)),
+        }
+    }
+}

rust/hg-core/examples/nodemap/main.rs

@@ -0,0 +1,146 b''
+// Copyright 2019-2020 Georges Racinet <[e-mail]>
+//
+// This software may be used and distributed according to the terms of the
+// GNU General Public License version 2 or any later version.
+
+use clap::*;
+use hg::revlog::node::*;
+use hg::revlog::nodemap::*;
+use hg::revlog::*;
+use memmap::MmapOptions;
+use rand::Rng;
+use std::fs::File;
+use std::io;
+use std::io::Write;
+use std::path::{Path, PathBuf};
+use std::str::FromStr;
+use std::time::Instant;
+
+mod index;
+use index::Index;
+
+fn mmap_index(repo_path: &Path) -> Index {
+    let mut path = PathBuf::from(repo_path);
+    path.extend([".hg", "store", "00changelog.i"].iter());
+    Index::load_mmap(path)
+}
+
+fn mmap_nodemap(path: &Path) -> NodeTree {
+    let file = File::open(path).unwrap();
+    let mmap = unsafe { MmapOptions::new().map(&file).unwrap() };
+    let len = mmap.len();
+    NodeTree::load_bytes(Box::new(mmap), len)
+}
+
+/// Scan the whole index and create the corresponding nodemap file at `path`
+fn create(index: &Index, path: &Path) -> io::Result<()> {
+    let mut file = File::create(path)?;
+    let start = Instant::now();
+    let mut nm = NodeTree::default();
+    for rev in 0..index.len() {
+        let rev = rev as Revision;
+        nm.insert(index, index.node(rev).unwrap(), rev).unwrap();
+    }
+    eprintln!("Nodemap constructed in RAM in {:?}", start.elapsed());
+    file.write(&nm.into_readonly_and_added_bytes().1)?;
+    eprintln!("Nodemap written to disk");
+    Ok(())
+}
+
+fn query(index: &Index, nm: &NodeTree, prefix: &str) {
+    let start = Instant::now();
+    let res = nm.find_hex(index, prefix);
+    println!("Result found in {:?}: {:?}", start.elapsed(), res);
+}
+
+fn bench(index: &Index, nm: &NodeTree, queries: usize) {
+    let len = index.len() as u32;
+    let mut rng = rand::thread_rng();
+    let nodes: Vec<Node> = (0..queries)
+        .map(|_| {
+            index
+                .node((rng.gen::<u32>() % len) as Revision)
+                .unwrap()
+                .clone()
+        })
+        .collect();
+    if queries < 10 {
+        let nodes_hex: Vec<String> =
+            nodes.iter().map(|n| n.encode_hex()).collect();
+        println!("Nodes: {:?}", nodes_hex);
+    }
+    let mut last: Option<Revision> = None;
+    let start = Instant::now();
+    for node in nodes.iter() {
+        last = nm.find_bin(index, node.into()).unwrap();
+    }
+    let elapsed = start.elapsed();
+    println!(
+        "Did {} queries in {:?} (mean {:?}), last was {:?} with result {:?}",
+        queries,
+        elapsed,
+        elapsed / (queries as u32),
+        nodes.last().unwrap().encode_hex(),
+        last
+    );
+}
+
+fn main() {
+    let matches = App::new("Nodemap pure Rust example")
+        .arg(
+            Arg::with_name("REPOSITORY")
+                .help("Path to the repository, always necessary for its index")
+                .required(true),
+        )
+        .arg(
+            Arg::with_name("NODEMAP_FILE")
+                .help("Path to the nodemap file, independent of REPOSITORY")
+                .required(true),
+        )
+        .subcommand(
+            SubCommand::with_name("create")
+                .about("Create NODEMAP_FILE by scanning repository index"),
+        )
+        .subcommand(
+            SubCommand::with_name("query")
+                .about("Query NODEMAP_FILE for PREFIX")
+                .arg(Arg::with_name("PREFIX").required(true)),
+        )
+        .subcommand(
+            SubCommand::with_name("bench")
+                .about(
+                    "Perform #QUERIES random successful queries on NODEMAP_FILE")
+                .arg(Arg::with_name("QUERIES").required(true)),
+        )
+        .get_matches();
+
+    let repo = matches.value_of("REPOSITORY").unwrap();
+    let nm_path = matches.value_of("NODEMAP_FILE").unwrap();
+
+    let index = mmap_index(&Path::new(repo));
+
+    if let Some(_) = matches.subcommand_matches("create") {
+        println!("Creating nodemap file {} for repository {}", nm_path, repo);
+        create(&index, &Path::new(nm_path)).unwrap();
+        return;
+    }
+
+    let nm = mmap_nodemap(&Path::new(nm_path));
+    if let Some(matches) = matches.subcommand_matches("query") {
+        let prefix = matches.value_of("PREFIX").unwrap();
+        println!(
+            "Querying {} in nodemap file {} of repository {}",
+            prefix, nm_path, repo
+        );
+        query(&index, &nm, prefix);
+    }
+    if let Some(matches) = matches.subcommand_matches("bench") {
+        let queries =
+            usize::from_str(matches.value_of("QUERIES").unwrap()).unwrap();
+        println!(
+            "Doing {} random queries in nodemap file {} of repository {}",
+            queries, nm_path, repo
+        );
+        bench(&index, &nm, queries);
+    }
+}

rust/Cargo.lock

@@ -17,6 +17,16 b' dependencies = ['
 ]
 
 [[package]]
+name = "atty"
+version = "0.2.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "hermit-abi 0.1.6 (registry+https://github.com/rust-lang/crates.io-index)",
+ "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
+ "winapi 0.3.8 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
+[[package]]
 name = "autocfg"
 version = "0.1.7"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -55,6 +65,20 b' version = "0.1.10"'
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
 [[package]]
+name = "clap"
+version = "2.33.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "ansi_term 0.11.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "atty 0.2.14 (registry+https://github.com/rust-lang/crates.io-index)",
+ "bitflags 1.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "strsim 0.8.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "textwrap 0.11.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-width 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
+ "vec_map 0.8.1 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
+[[package]]
 name = "cloudabi"
 version = "0.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -167,10 +191,12 b' version = "0.1.0"'
 dependencies = [
  "byteorder 1.3.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "cc 1.0.50 (registry+https://github.com/rust-lang/crates.io-index)",
+ "clap 2.33.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "hex 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "lazy_static 1.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
  "memchr 2.3.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "memmap 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "pretty_assertions 0.6.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "rand 0.6.5 (registry+https://github.com/rust-lang/crates.io-index)",
  "rand_pcg 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -214,6 +240,15 b' version = "2.3.0"'
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
 [[package]]
+name = "memmap"
+version = "0.7.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)",
+ "winapi 0.3.8 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
+[[package]]
 name = "memoffset"
 version = "0.5.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -523,6 +558,11 b' version = "0.7.0"'
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
 [[package]]
+name = "strsim"
+version = "0.8.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
+[[package]]
 name = "syn"
 version = "1.0.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -546,6 +586,14 b' dependencies = ['
 ]
 
 [[package]]
+name = "textwrap"
+version = "0.11.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "unicode-width 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
+[[package]]
 name = "thread_local"
 version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -562,11 +610,21 b' dependencies = ['
 ]
 
 [[package]]
+name = "unicode-width"
+version = "0.1.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
+[[package]]
 name = "unicode-xid"
 version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
 [[package]]
+name = "vec_map"
+version = "0.8.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+
+[[package]]
 name = "wasi"
 version = "0.9.0+wasi-snapshot-preview1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -601,6 +659,7 b' source = "registry+https://github.com/ru'
 [metadata]
 "checksum aho-corasick 0.7.8 (registry+https://github.com/rust-lang/crates.io-index)" = "743ad5a418686aad3b87fd14c43badd828cf26e214a00f92a384291cf22e1811"
 "checksum ansi_term 0.11.0 (registry+https://github.com/rust-lang/crates.io-index)" = "ee49baf6cb617b853aa8d93bf420db2383fab46d314482ca2803b40d5fde979b"
+"checksum atty 0.2.14 (registry+https://github.com/rust-lang/crates.io-index)" = "d9b39be18770d11421cdb1b9947a45dd3f37e93092cbf377614828a319d5fee8"
 "checksum autocfg 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)" = "1d49d90015b3c36167a20fe2810c5cd875ad504b39cff3d4eae7977e6b7c1cb2"
 "checksum autocfg 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "f8aac770f1885fd7e387acedd76065302551364496e46b3dd00860b2f8359b9d"
 "checksum bitflags 1.2.1 (registry+https://github.com/rust-lang/crates.io-index)" = "cf1de2fe8c75bc145a2f577add951f8134889b4795d47466a54a5c846d691693"
@@ -608,6 +667,7 b' source = "registry+https://github.com/ru'
 "checksum c2-chacha 0.2.3 (registry+https://github.com/rust-lang/crates.io-index)" = "214238caa1bf3a496ec3392968969cab8549f96ff30652c9e56885329315f6bb"
 "checksum cc 1.0.50 (registry+https://github.com/rust-lang/crates.io-index)" = "95e28fa049fda1c330bcf9d723be7663a899c4679724b34c81e9f5a326aab8cd"
 "checksum cfg-if 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)" = "4785bdd1c96b2a846b2bd7cc02e86b6b3dbf14e7e53446c4f54c92a361040822"
+"checksum clap 2.33.0 (registry+https://github.com/rust-lang/crates.io-index)" = "5067f5bb2d80ef5d68b4c87db81601f0b75bca627bc2ef76b141d7b846a3c6d9"
 "checksum cloudabi 0.0.3 (registry+https://github.com/rust-lang/crates.io-index)" = "ddfc5b9aa5d4507acaf872de71051dfd0e309860e88966e1051e462a077aac4f"
 "checksum cpython 0.4.1 (registry+https://github.com/rust-lang/crates.io-index)" = "bfaf3847ab963e40c4f6dd8d6be279bdf74007ae2413786a0dcbb28c52139a95"
 "checksum crossbeam-deque 0.7.2 (registry+https://github.com/rust-lang/crates.io-index)" = "c3aa945d63861bfe624b55d153a39684da1e8c0bc8fba932f7ee3a3c16cea3ca"
@@ -624,6 +684,7 b' source = "registry+https://github.com/ru'
 "checksum lazy_static 1.4.0 (registry+https://github.com/rust-lang/crates.io-index)" = "e2abad23fbc42b3700f2f279844dc832adb2b2eb069b2df918f455c4e18cc646"
 "checksum libc 0.2.66 (registry+https://github.com/rust-lang/crates.io-index)" = "d515b1f41455adea1313a4a2ac8a8a477634fbae63cc6100e3aebb207ce61558"
 "checksum memchr 2.3.0 (registry+https://github.com/rust-lang/crates.io-index)" = "3197e20c7edb283f87c071ddfc7a2cca8f8e0b888c242959846a6fce03c72223"
+"checksum memmap 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)" = "6585fd95e7bb50d6cc31e20d4cf9afb4e2ba16c5846fc76793f11218da9c475b"
 "checksum memoffset 0.5.3 (registry+https://github.com/rust-lang/crates.io-index)" = "75189eb85871ea5c2e2c15abbdd541185f63b408415e5051f5cac122d8c774b9"
 "checksum num-traits 0.2.11 (registry+https://github.com/rust-lang/crates.io-index)" = "c62be47e61d1842b9170f0fdeec8eba98e60e90e5446449a0545e5152acd7096"
 "checksum num_cpus 1.12.0 (registry+https://github.com/rust-lang/crates.io-index)" = "46203554f085ff89c235cd12f7075f3233af9b11ed7c9e16dfe2560d03313ce6"
@@ -660,11 +721,15 b' source = "registry+https://github.com/ru'
 "checksum scopeguard 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "b42e15e59b18a828bbf5c58ea01debb36b9b096346de35d941dcb89009f24a0d"
 "checksum semver 0.9.0 (registry+https://github.com/rust-lang/crates.io-index)" = "1d7eb9ef2c18661902cc47e535f9bc51b78acd254da71d375c2f6720d9a40403"
 "checksum semver-parser 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)" = "388a1df253eca08550bef6c72392cfe7c30914bf41df5269b68cbd6ff8f570a3"
+"checksum strsim 0.8.0 (registry+https://github.com/rust-lang/crates.io-index)" = "8ea5119cdb4c55b55d432abb513a0429384878c15dde60cc77b1c99de1a95a6a"
 "checksum syn 1.0.14 (registry+https://github.com/rust-lang/crates.io-index)" = "af6f3550d8dff9ef7dc34d384ac6f107e5d31c8f57d9f28e0081503f547ac8f5"
 "checksum tempfile 3.1.0 (registry+https://github.com/rust-lang/crates.io-index)" = "7a6e24d9338a0a5be79593e2fa15a648add6138caa803e2d5bc782c371732ca9"
+"checksum textwrap 0.11.0 (registry+https://github.com/rust-lang/crates.io-index)" = "d326610f408c7a4eb6f51c37c330e496b08506c9457c9d34287ecc38809fb060"
 "checksum thread_local 1.0.1 (registry+https://github.com/rust-lang/crates.io-index)" = "d40c6d1b69745a6ec6fb1ca717914848da4b44ae29d9b3080cbee91d72a69b14"
 "checksum twox-hash 1.5.0 (registry+https://github.com/rust-lang/crates.io-index)" = "3bfd5b7557925ce778ff9b9ef90e3ade34c524b5ff10e239c69a42d546d2af56"
+"checksum unicode-width 0.1.7 (registry+https://github.com/rust-lang/crates.io-index)" = "caaa9d531767d1ff2150b9332433f32a24622147e5ebb1f26409d5da67afd479"
 "checksum unicode-xid 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)" = "826e7639553986605ec5979c7dd957c7895e93eabed50ab2ffa7f6128a75097c"
+"checksum vec_map 0.8.1 (registry+https://github.com/rust-lang/crates.io-index)" = "05c78687fb1a80548ae3250346c3db86a80a7cdd77bda190189f2d0a0987c81a"
 "checksum wasi 0.9.0+wasi-snapshot-preview1 (registry+https://github.com/rust-lang/crates.io-index)" = "cccddf32554fecc6acb585f82a32a72e28b48f8c4c1883ddfeeeaa96f7d8e519"
 "checksum winapi 0.3.8 (registry+https://github.com/rust-lang/crates.io-index)" = "8093091eeb260906a183e6ae1abdba2ef5ef2257a21801128899c3fc699229c6"
 "checksum winapi-i686-pc-windows-gnu 0.4.0 (registry+https://github.com/rust-lang/crates.io-index)" = "ac3b87c63620426dd9b991e5ce0329eff545bccbbb34f3be09ff6fb6ab51b7b6"

rust/hg-core/Cargo.toml

@@ -23,8 +23,10 b' twox-hash = "1.5.0"'
 same-file = "1.0.6"
 
 [dev-dependencies]
+clap = "*"
+memmap = "0.7.0"
+pretty_assertions = "0.6.1"
 tempfile = "3.1.0"
-pretty_assertions = "0.6.1"
 
 [build-dependencies]
 cc = { version = "1.0.48", optional = true }