windows support

Author: skshetry

.github/workflows/ci.yml

@@ -13,10 +13,16 @@ env:
 jobs:
   ci:
     name: Check
-    runs-on: ubuntu-latest
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
     steps:
       - uses: actions/checkout@v4
       - uses: actions-rust-lang/setup-rust-toolchain@v1
       - run: cargo fmt --all --check
       - run: cargo clippy --all-targets --all-features
       - run: cargo test
+      - run: cargo build --verbose
+      - run: cargo build --release --verbose
+      - run: cargo build --profile dist --verbose

Cargo.lock

@@ -33,11 +33,14 @@ ignore = "0.4.22"
 itertools = "0.12.1"
 env_logger = "0.11.2"
 log = "0.4.20"
-uzers = "0.11.3"
 config = "0.14.0"
 serde_derive = "1.0.196"
 directories = "5.0.1"
 git2 = { version = "0.18.2", default-features = false }
+whoami = "1.4.1"
+
+[target.'cfg(windows)'.dependencies]
+file-id = "0.2.1"
 
 [dev-dependencies]
 tempfile = "3.10.0"
@@ -67,7 +70,7 @@ cargo-dist-version = "0.10.0"
 # The installers to generate for each app
 installers = ["shell", "powershell", "msi"]
 # Target platforms to build apps for (Rust target-triple syntax)
-targets = ["aarch64-apple-darwin", "x86_64-apple-darwin", "x86_64-unknown-linux-gnu", "x86_64-unknown-linux-musl"]
+targets = ["aarch64-apple-darwin", "x86_64-apple-darwin", "x86_64-unknown-linux-gnu", "x86_64-unknown-linux-musl", "x86_64-pc-windows-msvc"]
 # CI backends to support
 ci = ["github"]
 # Publish jobs to run in CI

Cargo.toml

@@ -1,17 +1,17 @@
-use crate::fsutils;
+use crate::fsutils::{compute_checksum, size_from_meta};
 use crate::hash::file_md5;
 use crate::objects::{Object, Tree, TreeEntry};
 use crate::odb::Odb;
 use crate::state::{State, StateHash, StateValue};
+use crate::timeutils::unix_time;
 use ignore;
 use ignore::gitignore::Gitignore;
 use jwalk::{Parallelism, WalkDir};
 use log::debug;
 use rayon::prelude::*;
-use std::fs;
-use std::os::unix::fs::MetadataExt;
 use std::path::{Path, PathBuf};
 use std::time::Instant;
+use std::{fs, u128};
 struct FileInfo {
     checksum: String,
     path: PathBuf,
@@ -20,9 +20,38 @@ struct FileInfo {
 
 impl FileInfo {
     fn from_metadata(path: &Path, meta: &fs::Metadata) -> Self {
+        let ut = unix_time(meta.modified().unwrap()).unwrap();
+        let ino: u128 = {
+            #[cfg(unix)]
+            {
+                use std::os::unix::fs::MetadataExt;
+                u128::from(meta.ino())
+            }
+
+            #[cfg(windows)]
+            {
+                use file_id::{get_file_id, FileId};
+                match get_file_id(path).unwrap() {
+                    FileId::LowRes {
+                        volume_serial_number: _,
+                        file_index,
+                    } => u128::from(file_index),
+                    FileId::HighRes {
+                        volume_serial_number: _,
+                        file_id: id,
+                    } => id,
+                    FileId::Inode {
+                        device_id: _,
+                        inode_number,
+                    } => u128::from(inode_number),
+                }
+            }
+        };
+        let size = size_from_meta(meta);
+        let checksum = compute_checksum(ut, ino, size);
         Self {
-            checksum: fsutils::checksum_from_metadata(meta),
-            size: meta.size(),
+            checksum,
+            size,
             path: path.to_path_buf(),
         }
     }

src/build.rs

@@ -1,30 +1,30 @@
 use crate::hash::md5;
+use core::panic;
 use std::fs;
-use std::os::unix::fs::{MetadataExt, PermissionsExt};
-use std::path::PathBuf;
+use std::path::{Path, PathBuf};
 
-pub fn checksum_from_metadata(meta: &fs::Metadata) -> String {
-    #[allow(clippy::cast_precision_loss)]
-    let m = meta.mtime() as f64 + (meta.mtime_nsec() as f64 / 1_000_000_000f64);
+pub fn compute_checksum(ut: f64, ino: u128, size: u64) -> String {
     let st = "([".to_owned()
-        + &meta.ino().to_string()
+        + &ino.to_string()
         + ", "
-        + &m.to_string()
+        + &ut.to_string()
         + ", "
-        + &meta.size().to_string()
+        + &size.to_string()
         + "],)";
     let hash = md5(&mut st.as_bytes());
     u128::from_str_radix(&hash, 16).unwrap().to_string()
 }
 
-pub fn checksum(path: &PathBuf) -> String {
-    let meta = fs::metadata(path).unwrap();
-    checksum_from_metadata(&meta)
+#[cfg(unix)]
+pub fn size_from_meta(meta: &fs::Metadata) -> u64 {
+    use std::os::unix::fs::MetadataExt;
+    meta.size()
 }
 
-pub fn size(path: PathBuf) -> u64 {
-    let meta = fs::metadata(path).unwrap();
-    meta.size()
+#[cfg(windows)]
+pub fn size_from_meta(meta: &fs::Metadata) -> u64 {
+    use std::os::windows::fs::MetadataExt;
+    meta.file_size()
 }
 
 pub fn transfer_file(from: &PathBuf, to: &PathBuf) {
@@ -33,7 +33,8 @@ pub fn transfer_file(from: &PathBuf, to: &PathBuf) {
         .unwrap_or_else(|_| panic!("transfer failed: {from:?} {to:?}"));
 }
 
-pub fn protect_file(path: &PathBuf) {
-    let permission = fs::Permissions::from_mode(0o444);
-    fs::set_permissions(path, permission).unwrap_or_default();
+pub fn protect_file(path: &Path) {
+    if let Ok(m) = path.metadata() {
+        m.permissions().set_readonly(true);
+    }
 }

src/fsutils.rs

@@ -13,6 +13,7 @@ pub mod odb;
 pub mod repo;
 pub mod state;
 pub mod status;
+pub mod timeutils;
 pub mod transfer;
 
 pub use build::build;

src/lib.rs

@@ -1,11 +1,12 @@
 use crate::hash::md5;
 use crate::json_format;
-use serde::{Deserialize, Serialize};
+use itertools::Itertools;
+use serde::{Deserialize, Deserializer, Serialize, Serializer};
 use std::convert::From;
 use std::error::Error;
 use std::fs::File;
 use std::io::BufReader;
-use std::path::PathBuf;
+use std::path::{Path, PathBuf};
 
 #[derive(Clone)]
 pub enum Object {
@@ -17,6 +18,7 @@ pub type HashFile = String;
 
 #[derive(Deserialize, Clone, PartialEq, Debug, PartialOrd, Ord, Eq)]
 pub struct TreeEntry {
+    #[serde(deserialize_with = "posixstr_to_ospath")]
     pub relpath: PathBuf,
     #[serde(rename = "md5")]
     pub oid: String,
@@ -27,6 +29,7 @@ pub struct TreeEntry {
 struct TreeEntrySerializer {
     #[serde(rename = "md5")]
     pub oid: String,
+    #[serde(serialize_with = "posixify_path")]
     pub relpath: PathBuf,
 }
 
@@ -54,6 +57,21 @@ impl Serialize for TreeEntry {
     }
 }
 
+fn posixify_path<S>(x: &Path, s: S) -> Result<S::Ok, S::Error>
+where
+    S: Serializer,
+{
+    s.serialize_str(&x.iter().map(|p| p.to_str().unwrap()).join("/"))
+}
+
+fn posixstr_to_ospath<'de, D>(deserializer: D) -> Result<PathBuf, D::Error>
+where
+    D: Deserializer<'de>,
+{
+    let s: &str = Deserialize::deserialize(deserializer)?;
+    Ok(s.split('/').collect())
+}
+
 impl Tree {
     pub fn serialize(&self) -> serde_json::Result<String> {
         // make it compatible with `json.dumps()` separator

src/objects.rs

@@ -1,5 +1,4 @@
 use std::ffi::OsString;
-use std::os::unix::fs::MetadataExt;
 use std::path::{Path, PathBuf};
 
 use log::debug;
@@ -8,6 +7,7 @@ use crate::config::Config;
 use crate::hash::md5;
 use crate::odb::Odb;
 use crate::state::State;
+use crate::timeutils::unix_time;
 use std::error::Error;
 use std::fs;
 use std::{env, io};
@@ -32,25 +32,23 @@ fn btime(tmp_dir: &Path) -> Result<f64, io::Error> {
 
     match result {
         Ok(()) => match fs::metadata(btime) {
-            #[allow(clippy::cast_precision_loss)]
-            Ok(meta) => Ok(meta.mtime() as f64 + (meta.mtime_nsec() as f64 / 1_000_000_000f64)),
+            Ok(meta) => Ok(unix_time(meta.modified()?).unwrap()),
             Err(e) => Err(e),
         },
         Err(e) => Err(e),
     }
 }
 
-#[cfg(unix)]
 fn db_dirname(root: &Path, tmp_dir: &Path) -> String {
-    use std::os::unix::ffi::OsStrExt;
     let btime = btime(tmp_dir).unwrap();
-    let user = uzers::get_current_username().unwrap();
+    let user = whoami::username();
     let dvc_major = 3;
-    let salt = 2;
+    let salt = 0;
 
     let mut st: OsString = "('".into();
     st.push(root.as_os_str());
     st.push("', ");
+    st.push("None, "); // subdir
     st.push(btime.to_string());
     st.push(", '");
     st.push(user);
@@ -60,7 +58,7 @@ fn db_dirname(root: &Path, tmp_dir: &Path) -> String {
     st.push(salt.to_string());
     st.push(")");
 
-    md5(&mut st.as_bytes())
+    md5(&mut st.as_encoded_bytes())
 }
 
 #[cfg(not(any(
@@ -73,6 +71,12 @@ fn db_dirs() -> PathBuf {
     "/var/tmp/dvc/repo".into()
 }
 
+#[cfg(target_os = "windows")]
+fn db_dirs() -> PathBuf {
+    Path::new(&env::var("CSIDL_COMMON_APPDATA").unwrap_or("C:/ProgramData/iterative/dvc".into()))
+        .to_owned()
+}
+
 #[cfg(target_os = "macos")]
 fn db_dirs() -> PathBuf {
     "/Library/Caches/dvc/repo".into()