refactor git crate
Rust / build (macos-latest, stable) (push) Waiting to run
Rust / build (ubuntu-latest, stable) (push) Waiting to run
Rust / build (windows-latest, stable) (push) Waiting to run
Rust / build (macos-latest, stable) (pull_request) Waiting to run
Rust / build (ubuntu-latest, stable) (pull_request) Waiting to run
Rust / build (windows-latest, stable) (pull_request) Waiting to run

This commit is contained in:
2026-09-10 16:23:13 +07:00
parent 705d4ea046
commit 94ea95c58c
12 changed files with 4068 additions and 3985 deletions
Generated
+1
View File
@@ -7971,6 +7971,7 @@ dependencies = [
"gix",
"gix-worktree",
"gix-worktree-state",
"ignore",
"nostr",
"signed_core",
"tempfile",
+1
View File
@@ -13,6 +13,7 @@ gix-worktree = "0.56"
gix-worktree-state = "0.34"
anyhow.workspace = true
diffy = "0.5"
ignore = "0.4"
[dev-dependencies]
tempfile = "3"
+96
View File
@@ -0,0 +1,96 @@
use std::path::{Path, PathBuf};
use anyhow::{Context, Result};
use signed_core::{Announcement, RepoAddr};
use crate::remote::{clone_repo, fetch_all};
/// On-disk cache of cloned repositories, keyed by owner pubkey / repo id.
#[derive(Debug, Clone)]
pub struct GitCache {
root: PathBuf,
}
impl GitCache {
pub fn new(root: PathBuf) -> Self {
Self { root }
}
/// The root directory holding the mirror clones.
pub fn root(&self) -> &Path {
&self.root
}
/// Local path of the clone for a repository.
pub fn repo_path(&self, addr: &RepoAddr) -> PathBuf {
self.root
.join(addr.public_key.to_hex())
.join(sanitize_path_component(&addr.identifier))
}
/// Open an existing clone.
pub fn open(&self, addr: &RepoAddr) -> Result<Option<gix::Repository>> {
let path = self.repo_path(addr);
match gix::open(&path) {
Ok(repo) => Ok(Some(repo)),
Err(gix::open::Error::NotARepository { .. }) => Ok(None),
Err(gix::open::Error::Io(e)) if e.kind() == std::io::ErrorKind::NotFound => Ok(None),
Err(e) => Err(e.into()),
}
}
/// Open the existing clone, fetching it first.
pub fn ensure_clone<U: AsRef<str>>(
&self,
addr: &RepoAddr,
clone_urls: &[U],
) -> Result<gix::Repository> {
let path = self.repo_path(addr);
if let Some(repo) = self.open(addr)? {
fetch_all(&repo).ok();
return Ok(repo);
}
if let Some(parent) = path.parent() {
std::fs::create_dir_all(parent)
.with_context(|| format!("failed to create {}", parent.display()))?;
}
clone_repo(clone_urls, &path)?;
self.open(addr)?
.ok_or_else(|| anyhow::anyhow!("clone finished but the repository cannot be opened"))
}
}
/// Map an untrusted repository id or display name to a safe single path component.
///
/// Everything outside `[A-Za-z0-9._-]` becomes `_`.
/// An id that maps to exactly `.` or `..` becomes `_`.
pub fn sanitize_path_component(id: &str) -> String {
let sanitized: String = id
.chars()
.map(|c| {
if c.is_ascii_alphanumeric() || matches!(c, '-' | '_' | '.') {
c
} else {
'_'
}
})
.collect();
if sanitized == "." || sanitized == ".." {
return "_".to_owned();
}
sanitized
}
/// The refs namespace of a fork's import in the target mirror.
pub fn fork_namespace(announcement: &Announcement) -> String {
format!(
"{}/{}",
announcement.owner.to_hex(),
sanitize_path_component(&announcement.id)
)
}
+304
View File
@@ -0,0 +1,304 @@
use std::path::Path;
use anyhow::Result;
use gix::diff::blob::unified_diff::{ConsumeHunk, DiffLineKind as GixLineKind, HunkHeader};
/// The kind of a [`DiffLine`].
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum DiffLineKind {
/// An unchanged context line, present on both sides.
Context,
/// A line added by the commit.
Addition,
/// A line removed by the commit.
Deletion,
}
/// One line of a file diff.
#[derive(Debug, Clone)]
pub struct DiffLine {
pub kind: DiffLineKind,
/// 1-based line number in the old version, if the line exists there.
pub old: Option<u32>,
/// 1-based line number in the new version, if the line exists there.
pub new: Option<u32>,
/// Line content without the trailing newline.
pub text: String,
}
/// A hunk of a file diff, like `@@ -a,b +c,d @@`.
#[derive(Debug, Clone)]
pub struct DiffHunk {
/// 1-based start line in the old version.
pub old_start: u32,
/// Number of old lines covered by the hunk.
pub old_lines: u32,
/// 1-based start line in the new version.
pub new_start: u32,
/// Number of new lines covered by the hunk.
pub new_lines: u32,
pub lines: Vec<DiffLine>,
}
/// How a file changed in a commit.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum DiffStatus {
Added,
Modified,
Deleted,
Renamed,
Copied,
}
/// The diff of one file in a commit.
#[derive(Debug, Clone)]
pub struct FileDiff {
/// Path of the file relative to the repo root.
///
/// For renames and copies, this is the destination path.
pub path: String,
/// Previous path, for renames and copies.
pub old_path: Option<String>,
pub status: DiffStatus,
/// Number of added lines, 0 for binary files.
pub insertions: usize,
/// Number of removed lines, 0 for binary files.
pub deletions: usize,
/// True if either version is binary, then `hunks` is empty.
pub binary: bool,
pub hunks: Vec<DiffHunk>,
}
/// The changes of one commit.
#[derive(Debug, Clone)]
pub struct CommitDiff {
pub files: Vec<FileDiff>,
}
/// The changes of the commit `id`, short or full, in the repository at `workdir`.
///
/// Compared against its first parent, the empty tree for the root commit.
pub fn worktree_commit_diff(workdir: &Path, id: &str) -> Result<CommitDiff> {
commit_diff(&gix::open(workdir)?, id)
}
fn commit_diff(repo: &gix::Repository, id: &str) -> Result<CommitDiff> {
let commit_id = repo.rev_parse_single(id.as_bytes())?;
let commit = commit_id.object()?.into_commit();
let new_tree = commit.tree()?;
let old_tree = match commit.parent_ids().next() {
Some(parent) => Some(parent.object()?.into_commit().tree()?),
None => None,
};
tree_diff(repo, old_tree.as_ref(), &new_tree)
}
/// The changes between two commits, `base`..`tip`, like `git diff base tip`.
///
/// Directories and submodules are skipped, files are sorted by path.
pub fn worktree_commit_range_diff(workdir: &Path, base: &str, tip: &str) -> Result<CommitDiff> {
let repo = gix::open(workdir)?;
let base_tree = repo
.rev_parse_single(base.as_bytes())?
.object()?
.into_commit()
.tree()?;
let tip_tree = repo
.rev_parse_single(tip.as_bytes())?
.object()?
.into_commit()
.tree()?;
tree_diff(&repo, Some(&base_tree), &tip_tree)
}
/// The changes between two trees. Used by both [`commit_diff`] and [`worktree_commit_range_diff`].
fn tree_diff(
repo: &gix::Repository,
old_tree: Option<&gix::Tree<'_>>,
new_tree: &gix::Tree<'_>,
) -> Result<CommitDiff> {
use gix::diff::blob::platform::prepare_diff::Operation;
use gix::object::tree::diff::Change;
use gix::objs::tree::EntryKind;
let changes = repo.diff_tree_to_tree(old_tree, Some(new_tree), None)?;
let mut cache = repo.diff_resource_cache_for_tree_diff()?;
let mut files = Vec::new();
for change in changes {
let attached = Change::from_change_ref(change.to_ref(), repo, repo);
// Skip directory trees and submodule gitlinks, only files are listed.
let (path, old_path, status) = match attached {
Change::Addition {
location,
entry_mode,
..
} if !matches!(entry_mode.kind(), EntryKind::Tree | EntryKind::Commit) => {
(location.to_owned(), None, DiffStatus::Added)
}
Change::Deletion {
location,
entry_mode,
..
} if !matches!(entry_mode.kind(), EntryKind::Tree | EntryKind::Commit) => {
(location.to_owned(), None, DiffStatus::Deleted)
}
Change::Modification {
location,
previous_entry_mode,
entry_mode,
..
} if !matches!(entry_mode.kind(), EntryKind::Tree | EntryKind::Commit)
&& !matches!(
previous_entry_mode.kind(),
EntryKind::Tree | EntryKind::Commit
) =>
{
(location.to_owned(), None, DiffStatus::Modified)
}
Change::Rewrite {
location,
source_location,
source_entry_mode,
entry_mode,
copy,
..
} if !matches!(entry_mode.kind(), EntryKind::Tree | EntryKind::Commit)
&& !matches!(
source_entry_mode.kind(),
EntryKind::Tree | EntryKind::Commit
) =>
{
let status = if copy {
DiffStatus::Copied
} else {
DiffStatus::Renamed
};
(
location.to_owned(),
Some(source_location.to_owned()),
status,
)
}
_ => continue,
};
// Always diff with the built-in algorithm.
// External diff drivers would shell out, out of scope for a read-only viewer.
let platform = attached.diff(&mut cache)?;
platform
.resource_cache
.options
.skip_internal_diff_if_external_is_configured = true;
let outcome = platform.resource_cache.prepare_diff()?;
let (binary, hunks, insertions, deletions) = match outcome.operation {
Operation::InternalDiff { algorithm } => {
let input = outcome.interned_input();
let diff = gix::diff::blob::diff_with_slider_heuristics(algorithm, &input);
let mut hunks = Vec::new();
let mut insertions = 0usize;
let mut deletions = 0usize;
let collector = HunkCollector {
hunks: &mut hunks,
insertions: &mut insertions,
deletions: &mut deletions,
};
gix::diff::blob::UnifiedDiff::new(&diff, &input, collector, Default::default())
.consume()?;
(false, hunks, insertions, deletions)
}
Operation::SourceOrDestinationIsBinary => (true, Vec::new(), 0, 0),
Operation::ExternalCommand { .. } => unreachable!("external diff drivers are disabled"),
};
files.push(FileDiff {
path: String::from_utf8_lossy(&path).into_owned(),
old_path: old_path.map(|p| String::from_utf8_lossy(&p).into_owned()),
status,
insertions,
deletions,
binary,
hunks,
});
}
files.sort_by(|a, b| a.path.cmp(&b.path));
Ok(CommitDiff { files })
}
/// Collects the hunks of one blob diff while tracking per-line numbers.
struct HunkCollector<'a> {
hunks: &'a mut Vec<DiffHunk>,
insertions: &'a mut usize,
deletions: &'a mut usize,
}
impl ConsumeHunk for HunkCollector<'_> {
type Out = ();
fn consume_hunk(
&mut self,
header: HunkHeader,
lines: &[(GixLineKind, &[u8])],
) -> std::io::Result<()> {
let mut old_ln = header.before_hunk_start;
let mut new_ln = header.after_hunk_start;
let mut out = Vec::with_capacity(lines.len());
for (kind, content) in lines {
let text = String::from_utf8_lossy(content).into_owned();
let line = match kind {
GixLineKind::Context => {
let line = DiffLine {
kind: DiffLineKind::Context,
old: Some(old_ln),
new: Some(new_ln),
text,
};
old_ln += 1;
new_ln += 1;
line
}
GixLineKind::Remove => {
*self.deletions += 1;
let line = DiffLine {
kind: DiffLineKind::Deletion,
old: Some(old_ln),
new: None,
text,
};
old_ln += 1;
line
}
GixLineKind::Add => {
*self.insertions += 1;
let line = DiffLine {
kind: DiffLineKind::Addition,
old: None,
new: Some(new_ln),
text,
};
new_ln += 1;
line
}
};
out.push(line);
}
self.hunks.push(DiffHunk {
old_start: header.before_hunk_start,
old_lines: header.before_hunk_len,
new_start: header.after_hunk_start,
new_lines: header.after_hunk_len,
lines: out,
});
Ok(())
}
fn finish(self) {}
}
+267
View File
@@ -0,0 +1,267 @@
use std::collections::HashSet;
use std::path::{Path, PathBuf};
use anyhow::Result;
/// In-memory object cache for history walks, see [`open_with_cache`].
///
/// Without one, a walk re-decodes the same commit objects from the object database.
/// Sized generously: a walk can cover a large portion of the repository's history.
const OBJECT_CACHE_BYTES: usize = 64 * 1024 * 1024;
/// Metadata of a commit, as shown in the repository browser's file header.
#[derive(Debug, Clone)]
pub struct FileCommit {
/// Shortened commit id, 7+ hex chars, disambiguated if needed.
pub id: String,
/// First line of the commit message.
pub summary: String,
/// Rest of the commit message after the title.
///
/// `None` for single-line commit messages.
pub description: Option<String>,
/// Author name.
pub author: String,
/// Author time, seconds since the Unix epoch.
pub time: i64,
}
/// Open the repository at `workdir` with an in-memory object cache.
///
/// Only history walks use it, they re-decode the same commit objects repeatedly.
/// Single-object reads open the repository plain.
pub(crate) fn open_with_cache(workdir: &Path) -> Result<gix::Repository> {
let mut repo = gix::open(workdir)?;
repo.object_cache_size_if_unset(OBJECT_CACHE_BYTES);
Ok(repo)
}
/// A [`FileCommit`] from a commit, with author, message title, body and shortened id.
///
/// The diff panel fetches the full commit on demand.
fn file_commit(commit: &gix::Commit<'_>) -> Result<FileCommit> {
file_commit_with_description(commit, true)
}
/// A [`FileCommit`] without the message body, for history lists that never display it.
///
/// Skipping the body saves an allocation per listed commit.
fn file_commit_summary(commit: &gix::Commit<'_>) -> Result<FileCommit> {
file_commit_with_description(commit, false)
}
/// [`file_commit`] and [`file_commit_summary`], `include_description` picks the body.
fn file_commit_with_description(
commit: &gix::Commit<'_>,
include_description: bool,
) -> Result<FileCommit> {
let author = commit.author()?;
let message = commit.message()?;
Ok(FileCommit {
id: commit.id().shorten_or_id().to_string(),
summary: String::from_utf8_lossy(message.title).trim().to_string(),
description: if include_description {
message
.body
.map(|body| String::from_utf8_lossy(body).trim().to_string())
.filter(|body| !body.is_empty())
} else {
None
},
author: String::from_utf8_lossy(author.name).trim().to_string(),
time: author.time()?.seconds,
})
}
/// Find the most recent commit that changed `rel`, a path relative to the worktree.
///
/// `Ok(None)` when no commit touched the file, e.g. an untracked file.
pub fn last_commit(repo: &gix::Repository, rel: &Path) -> Result<Option<FileCommit>> {
let rel = rel.to_path_buf();
Ok(last_commits(repo, std::slice::from_ref(&rel))?
.into_iter()
.next()
.map(|(_, commit)| commit))
}
/// Newest commit touching each of `rels`, like `git log -1 -- <rel>` per path.
/// `rels` are paths relative to the worktree.
///
/// Paths without any commit, like untracked files, are absent from the result.
pub fn worktree_last_commits(
workdir: &Path,
rels: &[PathBuf],
) -> Result<Vec<(PathBuf, FileCommit)>> {
last_commits(&open_with_cache(workdir)?, rels)
}
/// The walk behind [`last_commit`] and [`worktree_last_commits`].
///
/// Stops as soon as every pending path has its commit.
fn last_commits(repo: &gix::Repository, rels: &[PathBuf]) -> Result<Vec<(PathBuf, FileCommit)>> {
use gix::traverse::commit::simple::CommitTimeOrder;
let Some(head) = repo.head_id().ok() else {
return Ok(Vec::new());
};
// De-duplicate while preserving order.
let mut pending: Vec<PathBuf> = Vec::with_capacity(rels.len());
let mut seen: HashSet<&Path> = HashSet::with_capacity(rels.len());
for rel in rels {
if seen.insert(rel.as_path()) {
pending.push(rel.clone());
}
}
let walk = repo
.rev_walk([head])
.sorting(gix::revision::walk::Sorting::ByCommitTime(
CommitTimeOrder::NewestFirst,
));
let mut found = Vec::new();
for info in walk.all()? {
if pending.is_empty() {
break;
}
let info = info?;
let commit = info.object()?;
let tree = commit.tree()?;
let parent_tree = match info.parent_ids().next() {
Some(parent) => Some(parent.object()?.into_commit().tree()?),
None => None,
};
// Compare each unresolved path against this commit and its first parent.
// Resolved paths leave the pending set.
let mut ix = 0;
while ix < pending.len() {
let rel = &pending[ix];
let blob = tree.lookup_entry_by_path(rel)?;
let parent_blob = match &parent_tree {
Some(tree) => tree.lookup_entry_by_path(rel)?,
None => None,
};
if blob.map(|entry| entry.id().detach()) != parent_blob.map(|entry| entry.id().detach())
{
found.push((rel.clone(), file_commit(&commit)?));
pending.swap_remove(ix);
} else {
ix += 1;
}
}
}
Ok(found)
}
/// Cap on [`CommitList::commits`]. The virtual list renders a window at a time,
/// the tab badge shows the real count.
///
/// A huge history is never fully materialized in memory.
pub const MAX_LISTED_COMMITS: usize = 20_000;
/// Commits reachable from `HEAD`, newest first, possibly capped.
pub struct CommitList {
/// Number of commits reachable from HEAD.
pub total: usize,
/// Newest commits, capped at [`MAX_LISTED_COMMITS`].
pub commits: Vec<FileCommit>,
}
/// All commits reachable from `HEAD`, newest first, with author and summary.
///
/// Returns an empty list for a repository without any commits yet.
pub fn all_commits(repo: &gix::Repository) -> Result<CommitList> {
use gix::traverse::commit::simple::CommitTimeOrder;
let Some(head) = repo.head_id().ok() else {
return Ok(CommitList {
total: 0,
commits: Vec::new(),
});
};
let walk = repo
.rev_walk([head])
.sorting(gix::revision::walk::Sorting::ByCommitTime(
CommitTimeOrder::NewestFirst,
));
let mut commits = Vec::new();
let mut total = 0;
for info in walk.all()? {
let info = info?;
total += 1;
if commits.len() < MAX_LISTED_COMMITS {
commits.push(file_commit_summary(&info.object()?)?);
}
}
Ok(CommitList { total, commits })
}
/// Like [`all_commits`], but opens the repository at `workdir` first.
///
/// For non-bare clones the clone root is the worktree.
pub fn worktree_all_commits(workdir: &Path) -> Result<CommitList> {
all_commits(&open_with_cache(workdir)?)
}
/// Commits in the range `base`..`tip`, newest first, like `git log base..tip`.
pub fn worktree_commit_range_commits(
workdir: &Path,
base: &str,
tip: &str,
) -> Result<Vec<FileCommit>> {
use gix::traverse::commit::simple::CommitTimeOrder;
let repo = open_with_cache(workdir)?;
let base_id = repo.rev_parse_single(base.as_bytes())?;
let tip_id = repo.rev_parse_single(tip.as_bytes())?;
let walk = repo
.rev_walk([tip_id])
.sorting(gix::revision::walk::Sorting::ByCommitTime(
CommitTimeOrder::NewestFirst,
))
.with_hidden([base_id]);
let mut commits = Vec::new();
for info in walk.all()? {
let info = info?;
commits.push(file_commit_summary(&info.object()?)?);
}
Ok(commits)
}
/// The commit HEAD points to, like `git log -1`.
///
/// `Ok(None)` for a repository without commits yet, an unborn HEAD.
pub fn head_commit(repo: &gix::Repository) -> Result<Option<FileCommit>> {
let Some(head) = repo.head_id().ok() else {
return Ok(None);
};
let commit = head.object()?.into_commit();
Ok(Some(file_commit(&commit)?))
}
/// Full metadata of the commit `id`, short or full, in the repository at `workdir`.
/// Like [`head_commit`] for an arbitrary commit.
///
/// `Ok(None)` when the id cannot be resolved.
pub fn worktree_commit(workdir: &Path, id: &str) -> Result<Option<FileCommit>> {
let repo = gix::open(workdir)?;
match repo.rev_parse_single(id.as_bytes()) {
Ok(commit_id) => {
let commit = commit_id.object()?.into_commit();
Ok(Some(file_commit(&commit)?))
}
Err(_) => Ok(None),
}
}
File diff suppressed because it is too large Load Diff
+323
View File
@@ -0,0 +1,323 @@
use std::io::Write;
use std::path::Path;
use std::process::{Command, Stdio};
use anyhow::{Context, Result, bail};
use diffy::patch_set::{FileOperation, FilePatch, ParseOptions, PatchSet};
use diffy::{Hunk, Line};
use crate::diff::{CommitDiff, DiffHunk, DiffLine, DiffLineKind, DiffStatus, FileDiff};
use crate::history::FileCommit;
/// Apply a `git format-patch` patch or series with `git am`,
/// uses the git CLI because it handles the mbox format natively.
///
/// TODO: Replaced with a pure-Rust implementation later without changing callers.
pub fn apply_patch(repo_path: &Path, patch: &str) -> Result<()> {
let mut child = Command::new("git")
.arg("am")
.current_dir(repo_path)
.stdin(Stdio::piped())
.stderr(Stdio::piped())
.spawn()
.context("failed to spawn `git am`")?;
child
.stdin
.as_mut()
.expect("stdin piped")
.write_all(patch.as_bytes())?;
let output = child.wait_with_output()?;
if !output.status.success() {
bail!("git am failed: {}", String::from_utf8_lossy(&output.stderr));
}
Ok(())
}
/// The `git format-patch` mbox series of `base..tip`, like `git format-patch --stdout`.
/// Fails when the range has no commits.
///
/// The mbox is returned untrimmed. Trailing newlines are part of the format.
pub fn format_patch_between(repo_path: &Path, base: &str, tip: &str) -> Result<String> {
let output = Command::new("git")
.arg("-C")
.arg(repo_path)
.args(["format-patch", "--stdout", &format!("{base}..{tip}")])
.env("GIT_TERMINAL_PROMPT", "0")
.stderr(Stdio::piped())
.output()
.context("failed to spawn `git format-patch`")?;
if !output.status.success() {
bail!(
"git format-patch failed: {}",
String::from_utf8_lossy(&output.stderr).trim()
);
}
let patch = String::from_utf8_lossy(&output.stdout).into_owned();
if patch.trim().is_empty() {
bail!("no commits between {base} and {tip}");
}
Ok(patch)
}
/// Split a `git format-patch` series into its individual patches, mbox messages.
///
/// A single patch yields one element.
/// A malformed input yields one element covering it.
pub fn split_patch_series(patch: &str) -> Vec<&str> {
let mut starts = vec![0usize];
let mut search_from = 1;
while let Some(rel) = patch[search_from..].find("\nFrom ") {
let ix = search_from + rel + 1;
let hex = patch[ix + 5..]
.split(|c: char| !c.is_ascii_hexdigit())
.next()
.unwrap_or("");
if hex.len() == 40 {
starts.push(ix);
}
search_from = ix + 1;
}
starts
.iter()
.enumerate()
.map(|(i, &start)| {
let end = starts.get(i + 1).copied().unwrap_or(patch.len());
&patch[start..end]
})
.collect()
}
/// Parse `git format-patch` output, a single patch or a series.
///
/// Backed by [`diffy::patch_set`], which implements git's extended diff format:
/// `diff --git` headers, rename and copy detection, binary detection, and
/// C-style quoted or octal-escaped paths.
pub fn patch_diffs(patch: &str) -> Result<CommitDiff> {
if !patch.lines().any(|line| line.starts_with("diff --git ")) {
return Ok(CommitDiff { files: Vec::new() });
}
let mut files = Vec::new();
for file in PatchSet::parse(patch, ParseOptions::gitdiff()) {
files.push(file_diff(file?)?);
}
Ok(CommitDiff { files })
}
/// The [`FileDiff`] of one parsed file patch.
fn file_diff(file: FilePatch<'_, str>) -> Result<FileDiff> {
// The `---`/`+++` paths carry the `a/`/`b/` prefix, so the first path
// component is dropped, the same way `git apply -p1` does.
// Rename and copy paths come from their own headers, unprefixed.
let stripped;
let operation = match file.operation() {
operation @ (FileOperation::Rename { .. } | FileOperation::Copy { .. }) => operation,
operation => {
stripped = operation.strip_prefix(1);
&stripped
}
};
let (path, old_path, status) = match operation {
FileOperation::Create(path) => (path.as_ref(), None, DiffStatus::Added),
FileOperation::Delete(path) => (path.as_ref(), None, DiffStatus::Deleted),
FileOperation::Modify { modified, .. } => (modified.as_ref(), None, DiffStatus::Modified),
FileOperation::Rename { from, to } => {
(to.as_ref(), Some(from.as_ref()), DiffStatus::Renamed)
}
FileOperation::Copy { from, to } => (to.as_ref(), Some(from.as_ref()), DiffStatus::Copied),
};
let mut insertions = 0usize;
let mut deletions = 0usize;
let mut hunks = Vec::new();
let patch = file.patch();
if let Some(text) = patch.as_text() {
for hunk in text.hunks() {
let hunk = hunk_diff(hunk);
insertions += hunk
.lines
.iter()
.filter(|line| line.kind == DiffLineKind::Addition)
.count();
deletions += hunk
.lines
.iter()
.filter(|line| line.kind == DiffLineKind::Deletion)
.count();
hunks.push(hunk);
}
}
Ok(FileDiff {
path: path.to_owned(),
old_path: old_path.map(str::to_owned),
status,
insertions,
deletions,
binary: patch.is_binary(),
hunks,
})
}
/// The [`DiffHunk`] of one parsed hunk, including the line number of every line.
///
/// `diffy` reports only the hunk header ranges. The per-line numbers are
/// counted from them the way the header encodes them: context lines advance
/// both sides, deletions only the old, insertions only the new.
fn hunk_diff(hunk: &Hunk<'_, str>) -> DiffHunk {
let old_range = hunk.old_range();
let new_range = hunk.new_range();
let mut old = old_range.start() as u32;
let mut new = new_range.start() as u32;
let mut lines = Vec::with_capacity(hunk.lines().len());
for line in hunk.lines() {
let (kind, text) = match line {
Line::Context(text) => (DiffLineKind::Context, *text),
Line::Delete(text) => (DiffLineKind::Deletion, *text),
Line::Insert(text) => (DiffLineKind::Addition, *text),
};
let (old_no, new_no) = match kind {
DiffLineKind::Context => {
let numbers = (Some(old), Some(new));
old += 1;
new += 1;
numbers
}
DiffLineKind::Addition => {
let number = Some(new);
new += 1;
(None, number)
}
DiffLineKind::Deletion => {
let number = Some(old);
old += 1;
(number, None)
}
};
lines.push(DiffLine {
kind,
old: old_no,
new: new_no,
text: line_text(text),
});
}
DiffHunk {
old_start: old_range.start() as u32,
old_lines: old_range.len() as u32,
new_start: new_range.start() as u32,
new_lines: new_range.len() as u32,
lines,
}
}
/// The content of a parsed line without its line ending.
///
/// `diffy` keeps the trailing `\n`, the way `str::lines` splits it off.
fn line_text(text: &str) -> String {
let text = text.strip_suffix('\n').unwrap_or(text);
text.strip_suffix('\r').unwrap_or(text).to_owned()
}
/// Commits of a `git format-patch` output, a single patch or a series.
///
/// Entries appear in patch order, oldest first as `git format-patch` produces them.
pub fn patch_commits(patch: &str) -> Vec<FileCommit> {
let lines: Vec<&str> = patch.lines().collect();
let mut commits = Vec::new();
let mut i = 0;
while i < lines.len() {
// A patch starts with its `From <id> <date>` envelope line.
let Some(rest) = lines[i].strip_prefix("From ") else {
i += 1;
continue;
};
let Some(id) = rest.split_whitespace().next() else {
i += 1;
continue;
};
if id.len() != 40 {
i += 1;
continue;
}
let mut author = String::new();
let mut summary = String::new();
let mut time = 0i64;
// Envelope headers run up to the blank line before the commit message.
i += 1;
while i < lines.len() && !lines[i].is_empty() {
let header = lines[i];
if let Some(value) = header.strip_prefix("From: ") {
author = name_from_address(value);
} else if let Some(value) = header.strip_prefix("Subject: ") {
summary = strip_patch_prefix(value);
} else if let Some(value) = header.strip_prefix("Date: ") {
time = gix::date::parse(value.trim(), None)
.map(|t| t.seconds)
.unwrap_or(0);
}
i += 1;
}
commits.push(FileCommit {
id: id.to_string(),
summary,
description: None,
author,
time,
});
}
commits
}
/// The name part of a `From: Name <email>` header value.
fn name_from_address(from: &str) -> String {
match from.trim().find('<') {
Some(ix) => from[..ix].trim().to_string(),
None => from.trim().to_string(),
}
}
/// Strip the patch prefix from a `Subject:` header.
///
/// Examples are `[PATCH]`, `[PATCH 1/2]` and `[RFC PATCH]`.
fn strip_patch_prefix(subject: &str) -> String {
let trimmed = subject.trim();
let Some(rest) = trimmed.strip_prefix('[') else {
return trimmed.to_string();
};
let Some(end) = rest.find(']') else {
return trimmed.to_string();
};
if rest[..end].to_ascii_lowercase().contains("patch") {
rest[end + 1..].trim().to_string()
} else {
trimmed.to_string()
}
}
+357
View File
@@ -0,0 +1,357 @@
use std::collections::HashMap;
use std::path::Path;
use std::process::{Command, Stdio};
use anyhow::{Context, Result, bail};
use gix::interrupt::IS_INTERRUPTED;
use gix::progress::Discard;
/// Clone into `path` from the first working URL in `clone_urls`.
///
/// Unlike [`GitCache::ensure_clone`], the clone is not kept in any cache.
pub fn clone_repo<U: AsRef<str>>(clone_urls: &[U], path: &Path) -> Result<()> {
if path.exists() {
bail!("destination {} already exists", path.display());
}
try_each_url(clone_urls, "clone", |url| {
let repo = clone(url, path)?;
// The initial clone uses the default refspecs. Also fetch the `refs/nostr/*` PR refs.
fetch_all(&repo).ok();
Ok(())
})
}
/// Fetch all configured refspecs from `origin`, plus the `refs/nostr/*` namespace.
pub fn fetch_all(repo: &gix::Repository) -> Result<()> {
let options = gix::remote::ref_map::Options {
extra_refspecs: vec![
gix::refspec::parse(
gix::bstr::BStr::new("+refs/nostr/*:refs/nostr/*"),
gix::refspec::parse::Operation::Fetch,
)?
.to_owned(),
],
..Default::default()
};
repo.find_remote("origin")?
.connect(gix::remote::Direction::Fetch)?
.prepare_fetch(Discard, options)?
.receive(Discard, &IS_INTERRUPTED)?;
Ok(())
}
/// Push `commit` to `reference` on the server at `url`, from `repo_path`.
pub fn push_commit_ref(repo_path: &Path, url: &str, commit: &str, reference: &str) -> Result<()> {
let output = Command::new("git")
.arg("-C")
.arg(repo_path)
.args(["push"])
.arg(url)
.arg(format!("{commit}:{reference}"))
.env("GIT_TERMINAL_PROMPT", "0")
.stderr(Stdio::piped())
.output()
.context("failed to spawn `git push`")?;
if !output.status.success() {
bail!(
"git push failed: {}",
String::from_utf8_lossy(&output.stderr).trim()
);
}
Ok(())
}
/// Rewrite a grasp server URL to the https URL the git transport actually uses.
///
/// GRASP servers announce `grasp://<host>/<owner>/<repo>` clone URLs.
/// The transport is git smart HTTP, so the scheme is rewritten for gix.
fn transport_url(url: &str) -> String {
url.strip_prefix("grasp://")
.map(|rest| format!("https://{rest}"))
.unwrap_or_else(|| url.to_owned())
}
/// Run `attempt` against each URL in `urls` until one succeeds.
///
/// Returns the last error wrapped in `failed to {verb} from any mirror`,
/// or `no clone URLs provided` when the list is empty.
fn try_each_url<U: AsRef<str>, F>(urls: &[U], verb: &str, mut attempt: F) -> Result<()>
where
F: FnMut(&str) -> Result<()>,
{
let mut last_err: Option<anyhow::Error> = None;
for url in urls {
match attempt(url.as_ref()) {
Ok(()) => return Ok(()),
Err(e) => last_err = Some(e),
}
}
match last_err {
Some(e) => Err(e).context(format!("failed to {verb} from any mirror")),
None => bail!("no clone URLs provided"),
}
}
fn clone(url: &str, path: &Path) -> Result<gix::Repository> {
let url = transport_url(url);
let url = gix::url::parse(url).context("invalid clone URL")?;
let mut prepare = gix::prepare_clone(url, path)?;
let (mut checkout, _fetch) = prepare.fetch_then_checkout(Discard, &IS_INTERRUPTED)?;
let (repo, _checkout) = checkout.main_worktree(Discard, &IS_INTERRUPTED)?;
Ok(repo)
}
/// Push the `main` branch of the repository at `repo_path` to a grasp server.
pub fn push_main(repo_path: &Path, base_url: &str, owner: &str, repo_id: &str) -> Result<()> {
push_refspecs(
repo_path,
base_url,
owner,
repo_id,
&["refs/heads/main:refs/heads/main"],
)
}
/// Push every local branch and tag of the repository at `repo_path` to a grasp server.
///
/// This mirrors an initialized repository's whole history.
pub fn push_all(repo_path: &Path, base_url: &str, owner: &str, repo_id: &str) -> Result<()> {
push_refspecs(
repo_path,
base_url,
owner,
repo_id,
&["refs/heads/*:refs/heads/*", "refs/tags/*:refs/tags/*"],
)
}
/// Push `refspecs` to the grasp server URL derived from `base_url`, `owner` and `repo_id`.
fn push_refspecs(
repo_path: &Path,
base_url: &str,
owner: &str,
repo_id: &str,
refspecs: &[&str],
) -> Result<()> {
let url = format!("{base_url}/{owner}/{repo_id}.git");
let mut args: Vec<&str> = Vec::with_capacity(refspecs.len() + 2);
args.push("push");
args.push(&url);
args.extend_from_slice(refspecs);
let output = git_output(repo_path, &args, "git push")?;
if !output.status.success() {
bail!(
"git push to {base_url} failed: {}",
String::from_utf8_lossy(&output.stderr).trim()
);
}
Ok(())
}
/// Whether `url` advertises every ref in `expected` at the given commit.
///
/// Extra advertised refs are ignored: the question is whether the data this
/// push wanted to land is already there, not whether the remote is an exact mirror.
/// This is the convergence probe for a push that lost the compare-and-swap race
/// to the grasp server's own background ref alignment.
pub fn remote_has_refs(repo_path: &Path, url: &str, expected: &[(String, String)]) -> Result<bool> {
if expected.is_empty() {
return Ok(true);
}
let repo = gix::open(repo_path)?;
let url = transport_url(url);
// A URL-created remote has no configured fetch refspecs, and `ref_map` only
// keeps refs that match one. Match each expected ref by its exact name,
// like `git ls-remote <url> <name>` would; ref maps never write to the repository.
let refspecs = expected
.iter()
.map(|(name, _)| {
gix::refspec::parse(
gix::bstr::BStr::new(format!("+{name}:{name}").as_bytes()),
gix::refspec::parse::Operation::Fetch,
)
.map(|spec| spec.to_owned())
})
.collect::<Result<Vec<_>, _>>()
.context("invalid refspec")?;
let options = gix::remote::ref_map::Options {
extra_refspecs: refspecs,
..Default::default()
};
let (refs, _) = repo
.remote_at(url.as_str())
.with_context(|| format!("cannot use remote {url}"))?
.connect(gix::remote::Direction::Fetch)
.with_context(|| format!("cannot connect to {url}"))?
.ref_map(Discard, options)
.with_context(|| format!("listing refs of {url} failed"))?;
// Peeled tag entries carry the tag object in their direct oid, so mapping
// each advertised ref to its direct oid matches `git ls-remote` while
// skipping the duplicated `^{}` lines.
let advertised: HashMap<String, String> = refs
.remote_refs
.iter()
.filter_map(|reference| {
let (name, object, _peeled) = reference.unpack();
object.map(|oid| (String::from_utf8_lossy(name).into_owned(), oid.to_string()))
})
.collect();
Ok(expected
.iter()
.all(|(name, oid)| advertised.get(name.as_str()) == Some(oid)))
}
/// Add `origin` pointing at `url` when the repository has no remote yet.
///
/// No-op if `origin` already exists.
pub fn ensure_origin(repo_path: &Path, url: &str) -> Result<()> {
let repo = gix::open(repo_path)?;
if repo.find_remote("origin").is_ok() {
return Ok(());
}
// `git remote add` also configures the default fetch refspec.
edit_local_config(&repo, |config| {
config.set_raw_value("remote.origin.url", url)?;
config.set_raw_value("remote.origin.fetch", "+refs/heads/*:refs/remotes/origin/*")?;
Ok(())
})
}
/// Point `origin` at `url`, replacing an existing remote,
/// used after a clone whose `origin` points at the cloned-from path.
///
/// A working copy cloned from a local mirror is re-targeted at the grasp server.
pub fn set_origin(repo_path: &Path, url: &str) -> Result<()> {
let repo = gix::open(repo_path)?;
let had_origin = repo.find_remote("origin").is_ok();
edit_local_config(&repo, |config| {
// Replaces the existing url, like `git remote set-url origin <url>`.
// A pre-existing fetch refspec is left untouched.
config.set_raw_value("remote.origin.url", url)?;
if !had_origin {
config.set_raw_value("remote.origin.fetch", "+refs/heads/*:refs/remotes/origin/*")?;
}
Ok(())
})
}
/// Apply `edit` to the repository-local configuration and persist it.
///
/// The config file is locked while it is read, edited and written back,
/// like git would when running `git config` or `git remote`.
fn edit_local_config(
repo: &gix::Repository,
edit: impl FnOnce(&mut gix::config::File) -> Result<()>,
) -> Result<()> {
let config_path = repo.common_dir().join("config");
let mut lock = gix::lock::File::acquire_to_update_resource(
&config_path,
gix::lock::acquire::Fail::Immediately,
None,
)
.context("failed to lock repository config")?;
let mut config =
match gix::config::File::from_path_no_includes(config_path, gix::config::Source::Local) {
Ok(config) => config,
// A repository without a config file yet starts from scratch.
Err(gix::config::file::init::from_paths::Error::Io { source, .. })
if source.kind() == std::io::ErrorKind::NotFound =>
{
gix::config::File::default()
}
Err(error) => return Err(error).context("failed to read repository config"),
};
edit(&mut config)?;
config
.write_to(&mut lock)
.context("failed to write repository config")?;
lock.commit().context("failed to save repository config")?;
Ok(())
}
/// Fetch `refspec` into `repo_path` from the first working URL in `urls`.
/// When no URL works, the last error is returned.
///
/// Never touches the checked-out refs or the worktree.
pub fn fetch_repo_refs<U: AsRef<str>>(repo_path: &Path, urls: &[U], refspec: &str) -> Result<()> {
let repo = gix::open(repo_path)?;
let refspec = gix::refspec::parse(
gix::bstr::BStr::new(refspec),
gix::refspec::parse::Operation::Fetch,
)
.context("invalid fetch refspec")?
.to_owned();
try_each_url(urls, "fetch", |url| {
let url = transport_url(url);
let options = gix::remote::ref_map::Options {
extra_refspecs: vec![refspec.clone()],
..Default::default()
};
repo.remote_at(url.as_str())
.with_context(|| format!("fetch from {url} failed"))?
.connect(gix::remote::Direction::Fetch)
.with_context(|| format!("fetch from {url} failed"))?
.prepare_fetch(Discard, options)
.with_context(|| format!("fetch from {url} failed"))?
.receive(Discard, &IS_INTERRUPTED)
.with_context(|| format!("fetch from {url} failed"))?;
Ok(())
})
}
/// The URL of the `origin` remote of the repository at `workdir`.
///
/// `None` when it has no `origin` yet.
pub fn origin_url(workdir: &Path) -> Result<Option<String>> {
let Ok(repo) = gix::open(workdir) else {
return Ok(None);
};
let Ok(remote) = repo.find_remote("origin") else {
return Ok(None);
};
Ok(remote
.url(gix::remote::Direction::Fetch)
.map(|url| url.to_string()))
}
/// Run `git -C dir args`, disabling the terminal prompt and capturing stderr.
///
/// `what` names the command in the spawn error.
pub(crate) fn git_output(dir: &Path, args: &[&str], what: &str) -> Result<std::process::Output> {
Command::new("git")
.arg("-C")
.arg(dir)
.args(args)
.env("GIT_TERMINAL_PROMPT", "0")
.stderr(Stdio::piped())
.output()
.with_context(|| format!("failed to spawn `{what}`"))
}
+488
View File
@@ -0,0 +1,488 @@
use std::path::Path;
use anyhow::{Context, Result, bail};
use crate::history::open_with_cache;
use crate::worktree::{force_checkout, worktree_dirty};
/// The merge base of two revisions in the repository at `repo_path`,
/// revisions may be branch names, remote-tracking refs or commit ids.
///
/// `Ok(None)` when the revisions share no common ancestor.
///
/// Unresolvable revisions are errors.
pub fn merge_base(repo_path: &Path, a: &str, b: &str) -> Result<Option<String>> {
let repo = open_with_cache(repo_path)?;
let a = repo.rev_parse_single(a.as_bytes())?;
let b = repo.rev_parse_single(b.as_bytes())?;
match repo.merge_base(a, b) {
Ok(id) => Ok(Some(id.to_string())),
// No common ancestor, a valid outcome for a proposal.
Err(gix::repository::merge_base::Error::NotFound { .. }) => Ok(None),
Err(e) => Err(e.into()),
}
}
/// The commit HEAD points to in the repository at `repo_path`.
///
/// `None` when the repository has no commits yet, an unborn HEAD.
pub fn head_commit_id(repo_path: &Path) -> Result<Option<String>> {
let Ok(repo) = gix::open(repo_path) else {
return Ok(None);
};
match repo.head_id() {
Ok(id) => Ok(Some(id.to_string())),
Err(_) => Ok(None),
}
}
/// The commits in `base..HEAD` of the repository at `repo_path`, oldest first.
/// This is the order `git am` creates them.
///
/// `HEAD` alone when `base` is `None`.
pub fn commits_since(repo_path: &Path, base: Option<&str>) -> Result<Vec<String>> {
let repo = match gix::open(repo_path) {
Ok(repo) => repo,
Err(_) if base.is_none() => return Ok(Vec::new()),
Err(e) => return Err(e.into()),
};
let head = match repo.head_id() {
Ok(head) => head,
Err(_) if base.is_none() => return Ok(Vec::new()),
Err(e) => return Err(e).context("repository has no commits"),
};
let Some(base) = base else {
// `HEAD` alone when no base is given.
return Ok(vec![head.to_string()]);
};
let base = repo.rev_parse_single(base.as_bytes())?;
let mut commits = Vec::new();
for info in repo
.rev_walk([head])
.sorting(gix::revision::walk::Sorting::ByCommitTime(
gix::traverse::commit::simple::CommitTimeOrder::NewestFirst,
))
.with_hidden([base])
.all()?
{
commits.push(info?.id().to_string());
}
// Oldest first, like `git rev-list --reverse`, the order `git am` creates them.
commits.reverse();
Ok(commits)
}
/// The identity written to reflogs and commits created by this crate itself.
///
/// Like `git -c user.name=… -c user.email=…` per invocation: the repository works
/// without a global git identity, and `gix` runs no hooks and never signs.
pub(crate) fn repository_signature() -> (gix::actor::Signature, gix::date::parse::TimeBuf) {
let seconds = std::time::SystemTime::now()
.duration_since(std::time::UNIX_EPOCH)
.map(|duration| duration.as_secs() as i64)
.unwrap_or_default();
let signature = gix::actor::Signature {
name: gix::bstr::BString::from("Signed"),
email: gix::bstr::BString::from("signed@localhost"),
time: gix::date::Time { seconds, offset: 0 },
};
(signature, gix::date::parse::TimeBuf::default())
}
/// Create a repository at `path` with an initial `main` branch.
/// Write a `README.md` from `name` and `description`, then create the initial commit.
///
/// Returns the initial commit id.
pub fn init_repository(path: &Path, name: &str, description: &str) -> Result<String> {
use gix::refs::transaction::{Change, LogChange, PreviousValue, RefEdit, RefLog};
std::fs::create_dir_all(path)
.with_context(|| format!("failed to create {}", path.display()))?;
let repo = gix::init(path)?;
let (signature, mut time_buf) = repository_signature();
let signature = signature.to_ref(&mut time_buf);
// The initial branch is `main`, regardless of `init.defaultBranch` in
// the user's git configuration: point the unborn HEAD there.
let head = gix::refs::FullName::try_from("HEAD")
.map_err(|e| anyhow::anyhow!("invalid ref name: {e}"))?;
repo.edit_references_as(
[RefEdit {
change: Change::Update {
log: LogChange {
mode: RefLog::AndReference,
force_create_reflog: false,
message: "checkout: moving to main".into(),
},
expected: PreviousValue::Any,
new: gix::refs::Target::Symbolic(
gix::refs::FullName::try_from("refs/heads/main")
.map_err(|e| anyhow::anyhow!("invalid ref name: {e}"))?,
),
},
name: head,
deref: false,
}],
Some(signature),
)?;
let readme = if description.trim().is_empty() {
format!("# {name}\n")
} else {
format!("# {name}\n\n{description}\n")
};
std::fs::write(path.join("README.md"), &readme).context("failed to write README.md")?;
let blob = repo.write_object(gix::objs::Blob {
data: readme.into_bytes(),
})?;
let tree = repo.write_object(gix::objs::Tree {
entries: vec![gix::objs::tree::Entry {
mode: gix::objs::tree::EntryKind::Blob.into(),
filename: gix::bstr::BString::from("README.md"),
oid: blob.into(),
}],
})?;
let commit = repo.commit_as(
signature,
signature,
"HEAD",
"Initial commit",
tree,
Vec::<gix::ObjectId>::new(),
)?;
// Populate the index so the fresh repository is clean,
// as `git add` and`git commit` would leave it.
let mut index = repo.index_from_tree(&tree)?;
index.write(gix::index::write::Options::default())?;
let commit = commit.to_string();
if commit.len() != 40 {
bail!("unexpected initial commit id: {commit}");
}
Ok(commit)
}
/// The earliest unique commit of the repository at `repo_path`.
/// Used as the NIP-34 announcement's `euc` marker.
///
/// `None` for a repository without commits.
pub fn root_commit(repo_path: &Path) -> Result<Option<String>> {
let Ok(repo) = gix::open(repo_path) else {
return Ok(None);
};
let Ok(head) = repo.head_id() else {
// An unborn HEAD with no commits yet has no root commit.
return Ok(None);
};
for info in repo
.rev_walk([head])
.sorting(gix::revision::walk::Sorting::ByCommitTime(
gix::traverse::commit::simple::CommitTimeOrder::NewestFirst,
))
.all()?
{
let info = info?;
if info.parent_ids().next().is_none() {
let id = info.id().to_string();
return Ok((id.len() == 40).then_some(id));
}
}
Ok(None)
}
/// Full ref names under `prefix`, sorted lexicographically, like `git for-each-ref`.
/// `prefix` is a ref namespace like `refs/fork/<owner>/<id>`.
///
/// Returns an empty list when nothing matches.
pub fn refs_with_prefix(repo_path: &Path, prefix: &str) -> Result<Vec<String>> {
let pattern = prefix.trim_end_matches('/');
let repo = gix::open(repo_path)?;
let mut names = Vec::new();
for reference in repo.references()?.all()? {
let reference = reference.map_err(|error| anyhow::anyhow!("{error}"))?;
let name = String::from_utf8_lossy(reference.name().as_bstr()).into_owned();
// Match the pattern itself and everything beneath it, like `git for-each-ref`.
let under_pattern = name
.strip_prefix(pattern)
.is_some_and(|rest| rest.is_empty() || rest.starts_with('/'));
if under_pattern {
names.push(name);
}
}
// Sort lexicographically, like `git for-each-ref`.
names.sort();
Ok(names)
}
/// Delete every ref under `prefix` of the repository at `repo_path`.
/// `prefix` is a ref namespace like `refs/fork/<owner>/<id>`.
pub fn delete_refs_with_prefix(repo_path: &Path, prefix: &str) -> Result<()> {
use gix::refs::transaction::{Change, PreviousValue, RefEdit, RefLog};
let refs = refs_with_prefix(repo_path, prefix)?;
if refs.is_empty() {
return Ok(());
}
let repo = gix::open(repo_path)?;
let edits: Vec<RefEdit> = refs
.iter()
.map(|name| {
let full = gix::refs::FullName::try_from(name.as_str())
.map_err(|e| anyhow::anyhow!("invalid ref name {name}: {e}"))?;
Ok(RefEdit {
change: Change::Delete {
expected: PreviousValue::Any,
log: RefLog::AndReference,
},
name: full,
deref: false,
})
})
.collect::<Result<Vec<_>>>()?;
// Delete all refs with the given prefix.
repo.edit_references(edits)?;
Ok(())
}
/// Short name of the branch HEAD points to at `workdir`,
/// `None` when detached or unreadable, like `git branch --show-current`.
pub fn worktree_current_branch(workdir: &Path) -> Option<String> {
let repo = gix::open(workdir).ok()?;
let head = repo.head().ok()?;
let name = head.referent_name()?;
Some(String::from_utf8_lossy(name.shorten()).into_owned())
}
/// Whether the reference `name` exists in the repository at `workdir`.
pub fn worktree_ref_exists(workdir: &Path, name: &str) -> bool {
let Ok(repo) = gix::open(workdir) else {
return false;
};
repo.find_reference(name).is_ok()
}
/// Fast-forward local branches that trail their remote-tracking counterpart.
///
/// Returns whether any branch moved.
pub fn fast_forward_branches(workdir: &Path) -> Result<bool> {
let repo = gix::open(workdir)?;
let current = worktree_current_branch(workdir);
let heads = refs_with_prefix(workdir, "refs/heads")?;
let (signature, mut time_buf) = repository_signature();
let signature = signature.to_ref(&mut time_buf);
let mut moved = false;
for head in heads {
let Some(branch) = head.strip_prefix("refs/heads/") else {
continue;
};
let remote = format!("refs/remotes/origin/{branch}");
// No remote-tracking counterpart means the remote lacks this branch.
let Ok(mut remote_reference) = repo.find_reference(&remote) else {
continue;
};
let Ok(mut local_reference) = repo.find_reference(&head) else {
continue;
};
let Ok(remote_oid) = remote_reference.peel_to_id() else {
continue;
};
let Ok(local_oid) = local_reference.peel_to_id() else {
continue;
};
let remote_oid = remote_oid.detach();
let local_oid = local_oid.detach();
if local_oid == remote_oid {
continue;
}
// Only fast-forward.
// Local-only commits or diverged history must never be rewritten by a refresh.
let Ok(base) = repo.merge_base(local_oid, remote_oid) else {
continue;
};
if base != local_oid {
continue;
}
let full = gix::refs::FullName::try_from(head.as_str())
.map_err(|e| anyhow::anyhow!("invalid ref name: {e}"))?;
let edit = |new: gix::refs::Target| {
use gix::refs::transaction::{Change, LogChange, PreviousValue, RefEdit, RefLog};
RefEdit {
change: Change::Update {
log: LogChange {
mode: RefLog::AndReference,
force_create_reflog: false,
message: format!("merge {remote}: Fast-forward").into(),
},
expected: PreviousValue::ExistingMustMatch(gix::refs::Target::Object(
local_oid,
)),
new,
},
name: full.clone(),
deref: false,
}
};
if current.as_deref() == Some(branch) {
// Merge so the checked-out worktree follows the branch.
// Only proceed on a clean worktree, like `git merge --ff-only`.
if worktree_dirty(workdir) {
continue;
}
let tree = repo.find_object(remote_oid)?.peel_to_tree()?.id;
// Check out the remote tree, discarding local changes.
force_checkout(&repo, &tree)?;
// Update the branch reference to point to the remote tree.
repo.edit_references_as(
[edit(gix::refs::Target::Object(remote_oid))],
Some(signature),
)?;
moved = true;
} else {
// Update the branch reference to point to the remote tree.
repo.edit_references_as(
[edit(gix::refs::Target::Object(remote_oid))],
Some(signature),
)?;
moved = true;
}
}
Ok(moved)
}
/// Short names of local branches, `refs/heads/*`, of `repo`, sorted alphabetically.
pub fn repo_branches(repo: &gix::Repository) -> Result<Vec<String>> {
let mut names = Vec::new();
for reference in repo.references()?.local_branches()? {
let reference = reference.map_err(|error| anyhow::anyhow!("{error}"))?;
names.push(String::from_utf8_lossy(reference.name().shorten()).into_owned());
}
names.sort();
Ok(names)
}
/// Short names of tags, `refs/tags/*`, of `repo`, sorted alphabetically.
pub fn repo_tags(repo: &gix::Repository) -> Result<Vec<String>> {
let mut names = Vec::new();
for reference in repo.references()?.tags()? {
let reference = reference.map_err(|error| anyhow::anyhow!("{error}"))?;
names.push(String::from_utf8_lossy(reference.name().shorten()).into_owned());
}
names.sort();
Ok(names)
}
/// Short names of local branches, `refs/heads/*`, sorted alphabetically.
pub fn worktree_branches(workdir: &Path) -> Result<Vec<String>> {
repo_branches(&gix::open(workdir)?)
}
/// Short name of the branch HEAD points to, or `None` when detached.
///
/// Detached after checking out a tag or a commit directly.
pub fn current_branch(repo: &gix::Repository) -> Result<Option<String>> {
let head = repo.head()?;
let Some(name) = head.referent_name() else {
return Ok(None);
};
Ok(Some(String::from_utf8_lossy(name.shorten()).into_owned()))
}
/// Branch, tag and HEAD refs of a repository.
///
/// Ready for a NIP-34 kind-30618 repository state announcement.
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct RepoRefState {
/// `(full refname, commit id)` pairs for heads and tags, sorted.
pub refs: Vec<(String, String)>,
/// Short branch name HEAD points to, or `None` when detached.
pub head: Option<String>,
}
/// Collect the refs of `repo`.
///
/// Local branches and tags become `(refname, commit-id)` pairs.
/// Also reports the branch HEAD points to.
pub fn repo_ref_state(repo: &gix::Repository) -> Result<RepoRefState> {
let mut refs = Vec::new();
for reference in repo.references()?.local_branches()? {
let reference = reference.map_err(|error| anyhow::anyhow!("{error}"))?;
refs.push((
String::from_utf8_lossy(reference.name().as_bstr()).into_owned(),
reference.id().to_string(),
));
}
for reference in repo.references()?.tags()? {
let reference = reference.map_err(|error| anyhow::anyhow!("{error}"))?;
refs.push((
String::from_utf8_lossy(reference.name().as_bstr()).into_owned(),
reference.id().to_string(),
));
}
refs.sort();
let head = match repo.head() {
Ok(head) => head
.referent_name()
.filter(|name| name.as_bstr().starts_with(b"refs/heads/"))
.map(|name| String::from_utf8_lossy(name.shorten()).into_owned()),
Err(_) => None,
};
Ok(RepoRefState { refs, head })
}
/// [`repo_ref_state`] for the repository at `workdir`.
pub fn worktree_ref_state(workdir: &Path) -> Result<RepoRefState> {
repo_ref_state(&gix::open(workdir)?)
}
+43
View File
@@ -0,0 +1,43 @@
use std::path::{Path, PathBuf};
use ignore::WalkBuilder;
/// Maximum directory nesting depth when scanning for local repositories.
///
/// Pathological trees can't stall the scan.
const SCAN_MAX_DEPTH: usize = 12;
/// Walk `root` recursively and collect the paths of git repositories below it,
/// honouring `.gitignore` (and `.ignore`) files.
pub fn find_git_repos(root: &Path) -> Vec<PathBuf> {
if !root.is_dir() {
return Vec::new();
}
let walker = WalkBuilder::new(root)
.max_depth(Some(SCAN_MAX_DEPTH))
// Honour `.gitignore` even when the scan root is not itself a repository.
.require_git(false)
.build();
let mut repos: Vec<PathBuf> = walker
.flatten()
.filter(|entry| entry.file_type().is_some_and(|kind| kind.is_dir()))
.map(ignore::DirEntry::into_path)
.filter(|dir| dir.join(".git").exists())
.filter_map(|dir| dir.canonicalize().ok())
.collect();
repos.sort();
repos.dedup();
// A repository nested inside another, like a submodule worktree, is not reported.
let mut roots: Vec<PathBuf> = Vec::with_capacity(repos.len());
for repo in repos {
if !roots.iter().any(|kept| repo.starts_with(kept)) {
roots.push(repo);
}
}
roots
}
File diff suppressed because it is too large Load Diff
+354
View File
@@ -0,0 +1,354 @@
use std::collections::HashSet;
use std::path::{Path, PathBuf};
use anyhow::{Context, Result};
use gix::progress::Discard;
use crate::history::{FileCommit, head_commit};
use crate::repo::{current_branch, repository_signature};
/// Whether the worktree of `workdir` has uncommitted changes.
///
/// Best-effort: any read failure is reported as clean.
pub fn worktree_dirty(workdir: &Path) -> bool {
let Ok(repo) = gix::open(workdir) else {
return false;
};
// Changes to tracked files, staged or not; untracked files are excluded.
match repo.is_dirty() {
Ok(true) => return true,
Ok(false) => {}
Err(_) => return false,
}
// Untracked files surface as `DirectoryContents` items of the index-vs-worktree walk,
// tracked files only appear there when modified.
let Ok(platform) = repo.status(Discard) else {
return false;
};
let Ok(mut changes) = platform.into_index_worktree_iter(Vec::<gix::bstr::BString>::new())
else {
return false;
};
for change in changes.by_ref() {
match change {
Ok(gix::status::index_worktree::Item::DirectoryContents { .. }) => return true,
Ok(_) => {}
Err(_) => return false,
}
}
false
}
/// Commits in `base..branch` of the checkout at `workdir`.
///
/// Best-effort: 0 when the range cannot be computed.
pub fn worktree_commits_ahead(workdir: &Path, base: &str, branch: &str) -> u32 {
let Ok(repo) = gix::open(workdir) else {
return 0;
};
let (Some(base), Some(branch)) = (resolve_commit(&repo, base), resolve_commit(&repo, branch))
else {
return 0;
};
let Ok(walk) = repo.rev_walk([branch]).with_hidden([base]).all() else {
return 0;
};
walk.filter_map(Result::ok).count().min(u32::MAX as usize) as u32
}
/// Resolve `rev` to a commit id, accepting full refs,
/// symbolic refs and the bare branch names callers pass, like git's DWIM.
fn resolve_commit<'a>(repo: &'a gix::Repository, rev: &str) -> Option<gix::Id<'a>> {
if let Ok(id) = repo.rev_parse_single(rev.as_bytes()) {
return Some(id);
}
// Branch names arrive bare, like git resolving `main`.
if rev.contains('/') {
return None;
}
repo.rev_parse_single(format!("refs/heads/{rev}").as_bytes())
.ok()
}
/// Relative paths of all entries in the worktree, files and directories.
///
/// The `.git` directory is skipped.
pub fn worktree_entries(repo: &gix::Repository) -> Result<Vec<PathBuf>> {
let workdir = repo.workdir().context("repository has no worktree")?;
let mut entries: Vec<(PathBuf, bool)> = Vec::new();
collect_entries(workdir, workdir, &mut entries)?;
entries.sort_by(|(a, a_is_dir), (b, b_is_dir)| {
b_is_dir
.cmp(a_is_dir)
.then_with(|| a.as_os_str().cmp(b.as_os_str()))
});
Ok(entries.into_iter().map(|(path, _)| path).collect())
}
/// Read a file from the worktree.
///
/// Returns `Ok(None)` if the path is missing or not a regular file.
pub fn worktree_read(repo: &gix::Repository, rel: &Path) -> Result<Option<Vec<u8>>> {
let workdir = repo.workdir().context("repository has no worktree")?;
let path = workdir.join(rel);
match std::fs::read(&path) {
Ok(bytes) => Ok(Some(bytes)),
Err(e) if e.kind() == std::io::ErrorKind::NotFound => Ok(None),
Err(e) if e.kind() == std::io::ErrorKind::IsADirectory => Ok(None),
Err(e) => Err(e).with_context(|| format!("failed to read {}", path.display())),
}
}
/// Find the README file in the repository root.
///
/// Falls back to any other file whose name starts with `readme`.
pub fn find_readme(repo: &gix::Repository) -> Result<Option<PathBuf>> {
let Some(workdir) = repo.workdir() else {
return Ok(None);
};
let mut candidates: Vec<PathBuf> = Vec::new();
for entry in std::fs::read_dir(workdir)? {
let entry = entry?;
let name = entry.file_name();
let Some(name) = name.to_str() else { continue };
if name.to_ascii_lowercase().starts_with("readme") {
candidates.push(entry.path());
}
}
candidates.sort_by_key(|path| {
let ext = path
.extension()
.map(|e| e.to_string_lossy().to_ascii_lowercase());
match ext.as_deref() {
Some("md") => 0,
Some("markdown") => 1,
Some("mdown") => 2,
Some("mkdn") => 3,
Some(_) => 5,
None => 4,
}
});
Ok(candidates
.into_iter()
.next()
.and_then(|path| path.strip_prefix(workdir).ok().map(Path::to_path_buf)))
}
/// Everything the browser needs to refresh after a branch or tag switch.
pub struct WorktreeSnapshot {
/// Relative paths of all worktree entries, directories first.
pub entries: Vec<PathBuf>,
/// README path relative to the worktree, if any.
pub readme_path: Option<PathBuf>,
/// Contents of the README, if any.
pub readme: Option<Vec<u8>>,
/// Branch HEAD points to, `None` when detached, for example on a tag.
pub current_branch: Option<String>,
/// Commit HEAD points to, if any, see [`head_commit`].
pub head_commit: Option<FileCommit>,
}
/// Snapshot the worktree after a branch or tag switch.
///
/// Collects entries, the README, the branch HEAD points to and its commit.
pub fn worktree_snapshot(workdir: &Path) -> Result<WorktreeSnapshot> {
let repo = gix::open(workdir)?;
let readme_path = find_readme(&repo)?;
let readme = match &readme_path {
Some(path) => worktree_read(&repo, path)?,
None => None,
};
Ok(WorktreeSnapshot {
entries: worktree_entries(&repo)?,
readme_path,
readme,
current_branch: current_branch(&repo)?,
head_commit: head_commit(&repo)?,
})
}
/// Check out `tree` into the worktree of `repo`
pub(crate) fn force_checkout(repo: &gix::Repository, tree: &gix::hash::oid) -> Result<()> {
let workdir = repo
.workdir()
.context("repository has no worktree")?
.to_path_buf();
let mut index = repo.index_from_tree(tree)?;
// Files the previous index tracked but `tree` no longer contains are removed,
// like git deleting files that vanish between branches.
if let Ok(previous) = repo.index_or_empty() {
let keep: HashSet<PathBuf> = index
.entries()
.iter()
.map(|entry| PathBuf::from(String::from_utf8_lossy(entry.path(&index)).into_owned()))
.collect();
for entry in previous.entries() {
let rel = entry.path(&previous);
let rel = PathBuf::from(String::from_utf8_lossy(rel).into_owned());
if keep.contains(&rel) {
continue;
}
let path = workdir.join(&rel);
match std::fs::remove_file(&path) {
Ok(()) => {}
Err(error) if error.kind() == std::io::ErrorKind::NotFound => {}
Err(error) => {
return Err(error)
.with_context(|| format!("failed to remove {}", path.display()));
}
}
}
}
let mut options =
repo.checkout_options(gix_worktree::stack::state::attributes::Source::IdMapping)?;
options.overwrite_existing = true;
let objects = repo.objects.clone().into_arc()?;
let files = gix::progress::Discard;
let bytes = gix::progress::Discard;
// Check out the index into the worktree.
gix_worktree_state::checkout(
&mut index,
workdir,
objects,
&files,
&bytes,
&gix::interrupt::IS_INTERRUPTED,
options,
)?;
// Write the index to disk.
index.write(gix::index::write::Options::default())?;
Ok(())
}
/// Point `HEAD` at `target` and record the switch in the reflog.
fn move_head(
repo: &gix::Repository,
signature: gix::actor::SignatureRef<'_>,
target: gix::refs::Target,
message: &str,
) -> Result<()> {
use gix::refs::transaction::{Change, LogChange, PreviousValue, RefEdit, RefLog};
let head = gix::refs::FullName::try_from("HEAD")
.map_err(|e| anyhow::anyhow!("invalid ref name: {e}"))?;
// Update the reference, creating a reflog entry.
repo.edit_references_as(
[RefEdit {
change: Change::Update {
log: LogChange {
mode: RefLog::AndReference,
force_create_reflog: false,
message: message.into(),
},
expected: PreviousValue::Any,
new: target,
},
name: head,
deref: false,
}],
Some(signature),
)?;
Ok(())
}
/// Check out the local branch `name`, HEAD stays attached to it.
pub fn worktree_checkout_branch(workdir: &Path, name: &str) -> Result<()> {
let repo = gix::open(workdir)?;
let full = format!("refs/heads/{name}");
let branch = gix::refs::FullName::try_from(full.as_str())
.map_err(|e| anyhow::anyhow!("invalid ref name: {e}"))?;
let mut reference = repo.find_reference(&full)?;
let tree = reference.peel_to_tree()?.id;
let (signature, mut time_buf) = repository_signature();
let signature = signature.to_ref(&mut time_buf);
// Move HEAD to the branch, creating a reflog entry.
move_head(
&repo,
signature,
gix::refs::Target::Symbolic(branch),
&format!("checkout: moving to {name}"),
)?;
// Check out the branch's tree, replacing index + worktree.
force_checkout(&repo, &tree)?;
Ok(())
}
/// Check out the tag `name`, HEAD becomes detached at the tagged commit.
pub fn worktree_checkout_tag(workdir: &Path, name: &str) -> Result<()> {
let repo = gix::open(workdir)?;
let full = format!("refs/tags/{name}");
let mut reference = repo.find_reference(&full)?;
let commit = reference.peel_to_id()?;
let tree = reference.peel_to_tree()?.id;
let (signature, mut time_buf) = repository_signature();
let signature = signature.to_ref(&mut time_buf);
// Move HEAD to the tag, creating a reflog entry.
move_head(
&repo,
signature,
gix::refs::Target::Object(commit.detach()),
&format!("checkout: moving to {name}"),
)?;
// Check out the tag's tree, replacing index + worktree.
force_checkout(&repo, &tree)?;
Ok(())
}
fn collect_entries(root: &Path, dir: &Path, out: &mut Vec<(PathBuf, bool)>) -> Result<()> {
for entry in std::fs::read_dir(dir)? {
let entry = entry?;
if entry.file_name() == ".git" {
continue;
}
let is_dir = entry.file_type()?.is_dir();
let path = entry.path();
let rel = path.strip_prefix(root)?.to_path_buf();
out.push((rel, is_dir));
if is_dir {
collect_entries(root, &path, out)?;
}
}
Ok(())
}