rhg: Add lazy/cached dirstate data file ID parsing on Repo

The `dirstate_parents`, `dirstate_data_file_uuid`, and `dirstate_map` members
of `Repo` can be access in any order and the `.hg/dirstate` file should only
be opened once.

Differential Revision: https://phab.mercurial-scm.org/D11838

rust/hg-core/src/dirstate_tree/on_disk.rs

@@ -1,791 +1,791 b''
 //! The "version 2" disk representation of the dirstate
 //!
 //! See `mercurial/helptext/internals/dirstate-v2.txt`
 
 use crate::dirstate::TruncatedTimestamp;
 use crate::dirstate_tree::dirstate_map::{self, DirstateMap, NodeRef};
 use crate::dirstate_tree::path_with_basename::WithBasename;
 use crate::errors::HgError;
 use crate::utils::hg_path::HgPath;
 use crate::DirstateEntry;
 use crate::DirstateError;
 use crate::DirstateParents;
 use bitflags::bitflags;
 use bytes_cast::unaligned::{U16Be, U32Be};
 use bytes_cast::BytesCast;
 use format_bytes::format_bytes;
 use std::borrow::Cow;
 use std::convert::{TryFrom, TryInto};
 
 /// Added at the start of `.hg/dirstate` when the "v2" format is used.
 /// This a redundant sanity check more than an actual "magic number" since
 /// `.hg/requires` already governs which format should be used.
 pub const V2_FORMAT_MARKER: &[u8; 12] = b"dirstate-v2\n";
 
 /// Keep space for 256-bit hashes
 const STORED_NODE_ID_BYTES: usize = 32;
 
 /// … even though only 160 bits are used for now, with SHA-1
 const USED_NODE_ID_BYTES: usize = 20;
 
 pub(super) const IGNORE_PATTERNS_HASH_LEN: usize = 20;
 pub(super) type IgnorePatternsHash = [u8; IGNORE_PATTERNS_HASH_LEN];
 
 /// Must match constants of the same names in `mercurial/dirstateutils/v2.py`
 const TREE_METADATA_SIZE: usize = 44;
 const NODE_SIZE: usize = 44;
 
 /// Make sure that size-affecting changes are made knowingly
 #[allow(unused)]
 fn static_assert_size_of() {
     let _ = std::mem::transmute::<TreeMetadata, [u8; TREE_METADATA_SIZE]>;
     let _ = std::mem::transmute::<DocketHeader, [u8; TREE_METADATA_SIZE + 81]>;
     let _ = std::mem::transmute::<Node, [u8; NODE_SIZE]>;
 }
 
 // Must match `HEADER` in `mercurial/dirstateutils/docket.py`
 #[derive(BytesCast)]
 #[repr(C)]
 struct DocketHeader {
     marker: [u8; V2_FORMAT_MARKER.len()],
     parent_1: [u8; STORED_NODE_ID_BYTES],
     parent_2: [u8; STORED_NODE_ID_BYTES],
 
     metadata: TreeMetadata,
 
     /// Counted in bytes
     data_size: Size,
 
     uuid_size: u8,
 }
 
 pub struct Docket<'on_disk> {
     header: &'on_disk DocketHeader,
-    uuid: &'on_disk [u8],
+    pub uuid: &'on_disk [u8],
 }
 
 /// Fields are documented in the *Tree metadata in the docket file*
 /// section of `mercurial/helptext/internals/dirstate-v2.txt`
 #[derive(BytesCast)]
 #[repr(C)]
 struct TreeMetadata {
     root_nodes: ChildNodes,
     nodes_with_entry_count: Size,
     nodes_with_copy_source_count: Size,
     unreachable_bytes: Size,
     unused: [u8; 4],
 
     /// See *Optional hash of ignore patterns* section of
     /// `mercurial/helptext/internals/dirstate-v2.txt`
     ignore_patterns_hash: IgnorePatternsHash,
 }
 
 /// Fields are documented in the *The data file format*
 /// section of `mercurial/helptext/internals/dirstate-v2.txt`
 #[derive(BytesCast)]
 #[repr(C)]
 pub(super) struct Node {
     full_path: PathSlice,
 
     /// In bytes from `self.full_path.start`
     base_name_start: PathSize,
 
     copy_source: OptPathSlice,
     children: ChildNodes,
     pub(super) descendants_with_entry_count: Size,
     pub(super) tracked_descendants_count: Size,
     flags: U16Be,
     size: U32Be,
     mtime: PackedTruncatedTimestamp,
 }
 
 bitflags! {
     #[repr(C)]
     struct Flags: u16 {
         const WDIR_TRACKED = 1 << 0;
         const P1_TRACKED = 1 << 1;
         const P2_INFO = 1 << 2;
         const MODE_EXEC_PERM = 1 << 3;
         const MODE_IS_SYMLINK = 1 << 4;
         const HAS_FALLBACK_EXEC = 1 <<  5;
         const FALLBACK_EXEC = 1 <<  6;
         const HAS_FALLBACK_SYMLINK = 1 <<  7;
         const FALLBACK_SYMLINK = 1 <<  8;
         const EXPECTED_STATE_IS_MODIFIED = 1 << 9;
         const HAS_MODE_AND_SIZE = 1 <<10;
         const HAS_MTIME = 1 <<11;
         const MTIME_SECOND_AMBIGUOUS = 1 << 12;
         const DIRECTORY = 1 <<13;
         const ALL_UNKNOWN_RECORDED = 1 <<14;
         const ALL_IGNORED_RECORDED = 1 <<15;
     }
 }
 
 /// Duration since the Unix epoch
 #[derive(BytesCast, Copy, Clone)]
 #[repr(C)]
 struct PackedTruncatedTimestamp {
     truncated_seconds: U32Be,
     nanoseconds: U32Be,
 }
 
 /// Counted in bytes from the start of the file
 ///
 /// NOTE: not supporting `.hg/dirstate` files larger than 4 GiB.
 type Offset = U32Be;
 
 /// Counted in number of items
 ///
 /// NOTE: we choose not to support counting more than 4 billion nodes anywhere.
 type Size = U32Be;
 
 /// Counted in bytes
 ///
 /// NOTE: we choose not to support file names/paths longer than 64 KiB.
 type PathSize = U16Be;
 
 /// A contiguous sequence of `len` times `Node`, representing the child nodes
 /// of either some other node or of the repository root.
 ///
 /// Always sorted by ascending `full_path`, to allow binary search.
 /// Since nodes with the same parent nodes also have the same parent path,
 /// only the `base_name`s need to be compared during binary search.
 #[derive(BytesCast, Copy, Clone)]
 #[repr(C)]
 struct ChildNodes {
     start: Offset,
     len: Size,
 }
 
 /// A `HgPath` of `len` bytes
 #[derive(BytesCast, Copy, Clone)]
 #[repr(C)]
 struct PathSlice {
     start: Offset,
     len: PathSize,
 }
 
 /// Either nothing if `start == 0`, or a `HgPath` of `len` bytes
 type OptPathSlice = PathSlice;
 
 /// Unexpected file format found in `.hg/dirstate` with the "v2" format.
 ///
 /// This should only happen if Mercurial is buggy or a repository is corrupted.
 #[derive(Debug)]
 pub struct DirstateV2ParseError;
 
 impl From<DirstateV2ParseError> for HgError {
     fn from(_: DirstateV2ParseError) -> Self {
         HgError::corrupted("dirstate-v2 parse error")
     }
 }
 
 impl From<DirstateV2ParseError> for crate::DirstateError {
     fn from(error: DirstateV2ParseError) -> Self {
         HgError::from(error).into()
     }
 }
 
 impl<'on_disk> Docket<'on_disk> {
     pub fn parents(&self) -> DirstateParents {
         use crate::Node;
         let p1 = Node::try_from(&self.header.parent_1[..USED_NODE_ID_BYTES])
             .unwrap()
             .clone();
         let p2 = Node::try_from(&self.header.parent_2[..USED_NODE_ID_BYTES])
             .unwrap()
             .clone();
         DirstateParents { p1, p2 }
     }
 
     pub fn tree_metadata(&self) -> &[u8] {
         self.header.metadata.as_bytes()
     }
 
     pub fn data_size(&self) -> usize {
         // This `unwrap` could only panic on a 16-bit CPU
         self.header.data_size.get().try_into().unwrap()
     }
 
     pub fn data_filename(&self) -> String {
         String::from_utf8(format_bytes!(b"dirstate.{}", self.uuid)).unwrap()
     }
 }
 
 pub fn read_docket(
     on_disk: &[u8],
 ) -> Result<Docket<'_>, DirstateV2ParseError> {
     let (header, uuid) =
         DocketHeader::from_bytes(on_disk).map_err(|_| DirstateV2ParseError)?;
     let uuid_size = header.uuid_size as usize;
     if header.marker == *V2_FORMAT_MARKER && uuid.len() == uuid_size {
         Ok(Docket { header, uuid })
     } else {
         Err(DirstateV2ParseError)
     }
 }
 
 pub(super) fn read<'on_disk>(
     on_disk: &'on_disk [u8],
     metadata: &[u8],
 ) -> Result<DirstateMap<'on_disk>, DirstateV2ParseError> {
     if on_disk.is_empty() {
         return Ok(DirstateMap::empty(on_disk));
     }
     let (meta, _) = TreeMetadata::from_bytes(metadata)
         .map_err(|_| DirstateV2ParseError)?;
     let dirstate_map = DirstateMap {
         on_disk,
         root: dirstate_map::ChildNodes::OnDisk(read_nodes(
             on_disk,
             meta.root_nodes,
         )?),
         nodes_with_entry_count: meta.nodes_with_entry_count.get(),
         nodes_with_copy_source_count: meta.nodes_with_copy_source_count.get(),
         ignore_patterns_hash: meta.ignore_patterns_hash,
         unreachable_bytes: meta.unreachable_bytes.get(),
     };
     Ok(dirstate_map)
 }
 
 impl Node {
     pub(super) fn full_path<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'on_disk HgPath, DirstateV2ParseError> {
         read_hg_path(on_disk, self.full_path)
     }
 
     pub(super) fn base_name_start<'on_disk>(
         &self,
     ) -> Result<usize, DirstateV2ParseError> {
         let start = self.base_name_start.get();
         if start < self.full_path.len.get() {
             let start = usize::try_from(start)
                 // u32 -> usize, could only panic on a 16-bit CPU
                 .expect("dirstate-v2 base_name_start out of bounds");
             Ok(start)
         } else {
             Err(DirstateV2ParseError)
         }
     }
 
     pub(super) fn base_name<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'on_disk HgPath, DirstateV2ParseError> {
         let full_path = self.full_path(on_disk)?;
         let base_name_start = self.base_name_start()?;
         Ok(HgPath::new(&full_path.as_bytes()[base_name_start..]))
     }
 
     pub(super) fn path<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<dirstate_map::NodeKey<'on_disk>, DirstateV2ParseError> {
         Ok(WithBasename::from_raw_parts(
             Cow::Borrowed(self.full_path(on_disk)?),
             self.base_name_start()?,
         ))
     }
 
     pub(super) fn has_copy_source<'on_disk>(&self) -> bool {
         self.copy_source.start.get() != 0
     }
 
     pub(super) fn copy_source<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<Option<&'on_disk HgPath>, DirstateV2ParseError> {
         Ok(if self.has_copy_source() {
             Some(read_hg_path(on_disk, self.copy_source)?)
         } else {
             None
         })
     }
 
     fn flags(&self) -> Flags {
         Flags::from_bits_truncate(self.flags.get())
     }
 
     fn has_entry(&self) -> bool {
         self.flags().intersects(
             Flags::WDIR_TRACKED | Flags::P1_TRACKED | Flags::P2_INFO,
         )
     }
 
     pub(super) fn node_data(
         &self,
     ) -> Result<dirstate_map::NodeData, DirstateV2ParseError> {
         if self.has_entry() {
             Ok(dirstate_map::NodeData::Entry(self.assume_entry()?))
         } else if let Some(mtime) = self.cached_directory_mtime()? {
             Ok(dirstate_map::NodeData::CachedDirectory { mtime })
         } else {
             Ok(dirstate_map::NodeData::None)
         }
     }
 
     pub(super) fn cached_directory_mtime(
         &self,
     ) -> Result<Option<TruncatedTimestamp>, DirstateV2ParseError> {
         // For now we do not have code to handle the absence of
         // ALL_UNKNOWN_RECORDED, so we ignore the mtime if the flag is
         // unset.
         if self.flags().contains(Flags::DIRECTORY)
             && self.flags().contains(Flags::HAS_MTIME)
             && self.flags().contains(Flags::ALL_UNKNOWN_RECORDED)
         {
             Ok(Some(self.mtime.try_into()?))
         } else {
             Ok(None)
         }
     }
 
     fn synthesize_unix_mode(&self) -> u32 {
         let file_type = if self.flags().contains(Flags::MODE_IS_SYMLINK) {
             libc::S_IFLNK
         } else {
             libc::S_IFREG
         };
         let permisions = if self.flags().contains(Flags::MODE_EXEC_PERM) {
             0o755
         } else {
             0o644
         };
         file_type | permisions
     }
 
     fn assume_entry(&self) -> Result<DirstateEntry, DirstateV2ParseError> {
         // TODO: convert through raw bits instead?
         let wdir_tracked = self.flags().contains(Flags::WDIR_TRACKED);
         let p1_tracked = self.flags().contains(Flags::P1_TRACKED);
         let p2_info = self.flags().contains(Flags::P2_INFO);
         let mode_size = if self.flags().contains(Flags::HAS_MODE_AND_SIZE)
             && !self.flags().contains(Flags::EXPECTED_STATE_IS_MODIFIED)
         {
             Some((self.synthesize_unix_mode(), self.size.into()))
         } else {
             None
         };
         let mtime = if self.flags().contains(Flags::HAS_MTIME)
             && !self.flags().contains(Flags::DIRECTORY)
             && !self.flags().contains(Flags::EXPECTED_STATE_IS_MODIFIED)
         {
             let mut m: TruncatedTimestamp = self.mtime.try_into()?;
             if self.flags().contains(Flags::MTIME_SECOND_AMBIGUOUS) {
                 m.second_ambiguous = true;
             }
             Some(m)
         } else {
             None
         };
         let fallback_exec = if self.flags().contains(Flags::HAS_FALLBACK_EXEC)
         {
             Some(self.flags().contains(Flags::FALLBACK_EXEC))
         } else {
             None
         };
         let fallback_symlink =
             if self.flags().contains(Flags::HAS_FALLBACK_SYMLINK) {
                 Some(self.flags().contains(Flags::FALLBACK_SYMLINK))
             } else {
                 None
             };
         Ok(DirstateEntry::from_v2_data(
             wdir_tracked,
             p1_tracked,
             p2_info,
             mode_size,
             mtime,
             fallback_exec,
             fallback_symlink,
         ))
     }
 
     pub(super) fn entry(
         &self,
     ) -> Result<Option<DirstateEntry>, DirstateV2ParseError> {
         if self.has_entry() {
             Ok(Some(self.assume_entry()?))
         } else {
             Ok(None)
         }
     }
 
     pub(super) fn children<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'on_disk [Node], DirstateV2ParseError> {
         read_nodes(on_disk, self.children)
     }
 
     pub(super) fn to_in_memory_node<'on_disk>(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<dirstate_map::Node<'on_disk>, DirstateV2ParseError> {
         Ok(dirstate_map::Node {
             children: dirstate_map::ChildNodes::OnDisk(
                 self.children(on_disk)?,
             ),
             copy_source: self.copy_source(on_disk)?.map(Cow::Borrowed),
             data: self.node_data()?,
             descendants_with_entry_count: self
                 .descendants_with_entry_count
                 .get(),
             tracked_descendants_count: self.tracked_descendants_count.get(),
         })
     }
 
     fn from_dirstate_entry(
         entry: &DirstateEntry,
     ) -> (Flags, U32Be, PackedTruncatedTimestamp) {
         let (
             wdir_tracked,
             p1_tracked,
             p2_info,
             mode_size_opt,
             mtime_opt,
             fallback_exec,
             fallback_symlink,
         ) = entry.v2_data();
         // TODO: convert throug raw flag bits instead?
         let mut flags = Flags::empty();
         flags.set(Flags::WDIR_TRACKED, wdir_tracked);
         flags.set(Flags::P1_TRACKED, p1_tracked);
         flags.set(Flags::P2_INFO, p2_info);
         let size = if let Some((m, s)) = mode_size_opt {
             let exec_perm = m & libc::S_IXUSR != 0;
             let is_symlink = m & libc::S_IFMT == libc::S_IFLNK;
             flags.set(Flags::MODE_EXEC_PERM, exec_perm);
             flags.set(Flags::MODE_IS_SYMLINK, is_symlink);
             flags.insert(Flags::HAS_MODE_AND_SIZE);
             s.into()
         } else {
             0.into()
         };
         let mtime = if let Some(m) = mtime_opt {
             flags.insert(Flags::HAS_MTIME);
             if m.second_ambiguous {
                 flags.insert(Flags::MTIME_SECOND_AMBIGUOUS);
             };
             m.into()
         } else {
             PackedTruncatedTimestamp::null()
         };
         if let Some(f_exec) = fallback_exec {
             flags.insert(Flags::HAS_FALLBACK_EXEC);
             if f_exec {
                 flags.insert(Flags::FALLBACK_EXEC);
             }
         }
         if let Some(f_symlink) = fallback_symlink {
             flags.insert(Flags::HAS_FALLBACK_SYMLINK);
             if f_symlink {
                 flags.insert(Flags::FALLBACK_SYMLINK);
             }
         }
         (flags, size, mtime)
     }
 }
 
 fn read_hg_path(
     on_disk: &[u8],
     slice: PathSlice,
 ) -> Result<&HgPath, DirstateV2ParseError> {
     read_slice(on_disk, slice.start, slice.len.get()).map(HgPath::new)
 }
 
 fn read_nodes(
     on_disk: &[u8],
     slice: ChildNodes,
 ) -> Result<&[Node], DirstateV2ParseError> {
     read_slice(on_disk, slice.start, slice.len.get())
 }
 
 fn read_slice<T, Len>(
     on_disk: &[u8],
     start: Offset,
     len: Len,
 ) -> Result<&[T], DirstateV2ParseError>
 where
     T: BytesCast,
     Len: TryInto<usize>,
 {
     // Either `usize::MAX` would result in "out of bounds" error since a single
     // `&[u8]` cannot occupy the entire addess space.
     let start = start.get().try_into().unwrap_or(std::usize::MAX);
     let len = len.try_into().unwrap_or(std::usize::MAX);
     on_disk
         .get(start..)
         .and_then(|bytes| T::slice_from_bytes(bytes, len).ok())
         .map(|(slice, _rest)| slice)
         .ok_or_else(|| DirstateV2ParseError)
 }
 
 pub(crate) fn for_each_tracked_path<'on_disk>(
     on_disk: &'on_disk [u8],
     metadata: &[u8],
     mut f: impl FnMut(&'on_disk HgPath),
 ) -> Result<(), DirstateV2ParseError> {
     let (meta, _) = TreeMetadata::from_bytes(metadata)
         .map_err(|_| DirstateV2ParseError)?;
     fn recur<'on_disk>(
         on_disk: &'on_disk [u8],
         nodes: ChildNodes,
         f: &mut impl FnMut(&'on_disk HgPath),
     ) -> Result<(), DirstateV2ParseError> {
         for node in read_nodes(on_disk, nodes)? {
             if let Some(entry) = node.entry()? {
                 if entry.state().is_tracked() {
                     f(node.full_path(on_disk)?)
                 }
             }
             recur(on_disk, node.children, f)?
         }
         Ok(())
     }
     recur(on_disk, meta.root_nodes, &mut f)
 }
 
 /// Returns new data and metadata, together with whether that data should be
 /// appended to the existing data file whose content is at
 /// `dirstate_map.on_disk` (true), instead of written to a new data file
 /// (false).
 pub(super) fn write(
     dirstate_map: &DirstateMap,
     can_append: bool,
 ) -> Result<(Vec<u8>, Vec<u8>, bool), DirstateError> {
     let append = can_append && dirstate_map.write_should_append();
 
     // This ignores the space for paths, and for nodes without an entry.
     // TODO: better estimate? Skip the `Vec` and write to a file directly?
     let size_guess = std::mem::size_of::<Node>()
         * dirstate_map.nodes_with_entry_count as usize;
 
     let mut writer = Writer {
         dirstate_map,
         append,
         out: Vec::with_capacity(size_guess),
     };
 
     let root_nodes = writer.write_nodes(dirstate_map.root.as_ref())?;
 
     let meta = TreeMetadata {
         root_nodes,
         nodes_with_entry_count: dirstate_map.nodes_with_entry_count.into(),
         nodes_with_copy_source_count: dirstate_map
             .nodes_with_copy_source_count
             .into(),
         unreachable_bytes: dirstate_map.unreachable_bytes.into(),
         unused: [0; 4],
         ignore_patterns_hash: dirstate_map.ignore_patterns_hash,
     };
     Ok((writer.out, meta.as_bytes().to_vec(), append))
 }
 
 struct Writer<'dmap, 'on_disk> {
     dirstate_map: &'dmap DirstateMap<'on_disk>,
     append: bool,
     out: Vec<u8>,
 }
 
 impl Writer<'_, '_> {
     fn write_nodes(
         &mut self,
         nodes: dirstate_map::ChildNodesRef,
     ) -> Result<ChildNodes, DirstateError> {
         // Reuse already-written nodes if possible
         if self.append {
             if let dirstate_map::ChildNodesRef::OnDisk(nodes_slice) = nodes {
                 let start = self.on_disk_offset_of(nodes_slice).expect(
                     "dirstate-v2 OnDisk nodes not found within on_disk",
                 );
                 let len = child_nodes_len_from_usize(nodes_slice.len());
                 return Ok(ChildNodes { start, len });
             }
         }
 
         // `dirstate_map::ChildNodes::InMemory` contains a `HashMap` which has
         // undefined iteration order. Sort to enable binary search in the
         // written file.
         let nodes = nodes.sorted();
         let nodes_len = nodes.len();
 
         // First accumulate serialized nodes in a `Vec`
         let mut on_disk_nodes = Vec::with_capacity(nodes_len);
         for node in nodes {
             let children =
                 self.write_nodes(node.children(self.dirstate_map.on_disk)?)?;
             let full_path = node.full_path(self.dirstate_map.on_disk)?;
             let full_path = self.write_path(full_path.as_bytes());
             let copy_source = if let Some(source) =
                 node.copy_source(self.dirstate_map.on_disk)?
             {
                 self.write_path(source.as_bytes())
             } else {
                 PathSlice {
                     start: 0.into(),
                     len: 0.into(),
                 }
             };
             on_disk_nodes.push(match node {
                 NodeRef::InMemory(path, node) => {
                     let (flags, size, mtime) = match &node.data {
                         dirstate_map::NodeData::Entry(entry) => {
                             Node::from_dirstate_entry(entry)
                         }
                         dirstate_map::NodeData::CachedDirectory { mtime } => (
                             // we currently never set a mtime if unknown file
                             // are present.
                             // So if we have a mtime for a directory, we know
                             // they are no unknown
                             // files and we
                             // blindly set ALL_UNKNOWN_RECORDED.
                             //
                             // We never set ALL_IGNORED_RECORDED since we
                             // don't track that case
                             // currently.
                             Flags::DIRECTORY
                                 | Flags::HAS_MTIME
                                 | Flags::ALL_UNKNOWN_RECORDED,
                             0.into(),
                             (*mtime).into(),
                         ),
                         dirstate_map::NodeData::None => (
                             Flags::DIRECTORY,
                             0.into(),
                             PackedTruncatedTimestamp::null(),
                         ),
                     };
                     Node {
                         children,
                         copy_source,
                         full_path,
                         base_name_start: u16::try_from(path.base_name_start())
                             // Could only panic for paths over 64 KiB
                             .expect("dirstate-v2 path length overflow")
                             .into(),
                         descendants_with_entry_count: node
                             .descendants_with_entry_count
                             .into(),
                         tracked_descendants_count: node
                             .tracked_descendants_count
                             .into(),
                         flags: flags.bits().into(),
                         size,
                         mtime,
                     }
                 }
                 NodeRef::OnDisk(node) => Node {
                     children,
                     copy_source,
                     full_path,
                     ..*node
                 },
             })
         }
         // … so we can write them contiguously, after writing everything else
         // they refer to.
         let start = self.current_offset();
         let len = child_nodes_len_from_usize(nodes_len);
         self.out.extend(on_disk_nodes.as_bytes());
         Ok(ChildNodes { start, len })
     }
 
     /// If the given slice of items is within `on_disk`, returns its offset
     /// from the start of `on_disk`.
     fn on_disk_offset_of<T>(&self, slice: &[T]) -> Option<Offset>
     where
         T: BytesCast,
     {
         fn address_range(slice: &[u8]) -> std::ops::RangeInclusive<usize> {
             let start = slice.as_ptr() as usize;
             let end = start + slice.len();
             start..=end
         }
         let slice_addresses = address_range(slice.as_bytes());
         let on_disk_addresses = address_range(self.dirstate_map.on_disk);
         if on_disk_addresses.contains(slice_addresses.start())
             && on_disk_addresses.contains(slice_addresses.end())
         {
             let offset = slice_addresses.start() - on_disk_addresses.start();
             Some(offset_from_usize(offset))
         } else {
             None
         }
     }
 
     fn current_offset(&mut self) -> Offset {
         let mut offset = self.out.len();
         if self.append {
             offset += self.dirstate_map.on_disk.len()
         }
         offset_from_usize(offset)
     }
 
     fn write_path(&mut self, slice: &[u8]) -> PathSlice {
         let len = path_len_from_usize(slice.len());
         // Reuse an already-written path if possible
         if self.append {
             if let Some(start) = self.on_disk_offset_of(slice) {
                 return PathSlice { start, len };
             }
         }
         let start = self.current_offset();
         self.out.extend(slice.as_bytes());
         PathSlice { start, len }
     }
 }
 
 fn offset_from_usize(x: usize) -> Offset {
     u32::try_from(x)
         // Could only panic for a dirstate file larger than 4 GiB
         .expect("dirstate-v2 offset overflow")
         .into()
 }
 
 fn child_nodes_len_from_usize(x: usize) -> Size {
     u32::try_from(x)
         // Could only panic with over 4 billion nodes
         .expect("dirstate-v2 slice length overflow")
         .into()
 }
 
 fn path_len_from_usize(x: usize) -> PathSize {
     u16::try_from(x)
         // Could only panic for paths over 64 KiB
         .expect("dirstate-v2 path length overflow")
         .into()
 }
 
 impl From<TruncatedTimestamp> for PackedTruncatedTimestamp {
     fn from(timestamp: TruncatedTimestamp) -> Self {
         Self {
             truncated_seconds: timestamp.truncated_seconds().into(),
             nanoseconds: timestamp.nanoseconds().into(),
         }
     }
 }
 
 impl TryFrom<PackedTruncatedTimestamp> for TruncatedTimestamp {
     type Error = DirstateV2ParseError;
 
     fn try_from(
         timestamp: PackedTruncatedTimestamp,
     ) -> Result<Self, Self::Error> {
         Self::from_already_truncated(
             timestamp.truncated_seconds.get(),
             timestamp.nanoseconds.get(),
             false,
         )
     }
 }
 impl PackedTruncatedTimestamp {
     fn null() -> Self {
         Self {
             truncated_seconds: 0.into(),
             nanoseconds: 0.into(),
         }
     }
 }

rust/hg-core/src/repo.rs

@@ -1,429 +1,464 b''
 use crate::changelog::Changelog;
 use crate::config::{Config, ConfigError, ConfigParseError};
 use crate::dirstate::DirstateParents;
 use crate::dirstate_tree::dirstate_map::DirstateMap;
 use crate::dirstate_tree::owning::OwningDirstateMap;
 use crate::errors::HgError;
 use crate::errors::HgResultExt;
 use crate::exit_codes;
 use crate::lock::{try_with_lock_no_wait, LockError};
 use crate::manifest::{Manifest, Manifestlog};
 use crate::revlog::filelog::Filelog;
 use crate::revlog::revlog::RevlogError;
 use crate::utils::files::get_path_from_bytes;
 use crate::utils::hg_path::HgPath;
 use crate::utils::SliceExt;
 use crate::vfs::{is_dir, is_file, Vfs};
 use crate::{requirements, NodePrefix};
 use crate::{DirstateError, Revision};
 use std::cell::{Ref, RefCell, RefMut};
 use std::collections::HashSet;
 use std::path::{Path, PathBuf};
 
 /// A repository on disk
 pub struct Repo {
     working_directory: PathBuf,
     dot_hg: PathBuf,
     store: PathBuf,
     requirements: HashSet<String>,
     config: Config,
     dirstate_parents: LazyCell<DirstateParents, HgError>,
+    dirstate_data_file_uuid: LazyCell<Option<Vec<u8>>, HgError>,
     dirstate_map: LazyCell<OwningDirstateMap, DirstateError>,
     changelog: LazyCell<Changelog, HgError>,
     manifestlog: LazyCell<Manifestlog, HgError>,
 }
 
 #[derive(Debug, derive_more::From)]
 pub enum RepoError {
     NotFound {
         at: PathBuf,
     },
     #[from]
     ConfigParseError(ConfigParseError),
     #[from]
     Other(HgError),
 }
 
 impl From<ConfigError> for RepoError {
     fn from(error: ConfigError) -> Self {
         match error {
             ConfigError::Parse(error) => error.into(),
             ConfigError::Other(error) => error.into(),
         }
     }
 }
 
 impl Repo {
     /// tries to find nearest repository root in current working directory or
     /// its ancestors
     pub fn find_repo_root() -> Result<PathBuf, RepoError> {
         let current_directory = crate::utils::current_dir()?;
         // ancestors() is inclusive: it first yields `current_directory`
         // as-is.
         for ancestor in current_directory.ancestors() {
             if is_dir(ancestor.join(".hg"))? {
                 return Ok(ancestor.to_path_buf());
             }
         }
         return Err(RepoError::NotFound {
             at: current_directory,
         });
     }
 
     /// Find a repository, either at the given path (which must contain a `.hg`
     /// sub-directory) or by searching the current directory and its
     /// ancestors.
     ///
     /// A method with two very different "modes" like this usually a code smell
     /// to make two methods instead, but in this case an `Option` is what rhg
     /// sub-commands get from Clap for the `-R` / `--repository` CLI argument.
     /// Having two methods would just move that `if` to almost all callers.
     pub fn find(
         config: &Config,
         explicit_path: Option<PathBuf>,
     ) -> Result<Self, RepoError> {
         if let Some(root) = explicit_path {
             if is_dir(root.join(".hg"))? {
                 Self::new_at_path(root.to_owned(), config)
             } else if is_file(&root)? {
                 Err(HgError::unsupported("bundle repository").into())
             } else {
                 Err(RepoError::NotFound {
                     at: root.to_owned(),
                 })
             }
         } else {
             let root = Self::find_repo_root()?;
             Self::new_at_path(root, config)
         }
     }
 
     /// To be called after checking that `.hg` is a sub-directory
     fn new_at_path(
         working_directory: PathBuf,
         config: &Config,
     ) -> Result<Self, RepoError> {
         let dot_hg = working_directory.join(".hg");
 
         let mut repo_config_files = Vec::new();
         repo_config_files.push(dot_hg.join("hgrc"));
         repo_config_files.push(dot_hg.join("hgrc-not-shared"));
 
         let hg_vfs = Vfs { base: &dot_hg };
         let mut reqs = requirements::load_if_exists(hg_vfs)?;
         let relative =
             reqs.contains(requirements::RELATIVE_SHARED_REQUIREMENT);
         let shared =
             reqs.contains(requirements::SHARED_REQUIREMENT) || relative;
 
         // From `mercurial/localrepo.py`:
         //
         // if .hg/requires contains the sharesafe requirement, it means
         // there exists a `.hg/store/requires` too and we should read it
         // NOTE: presence of SHARESAFE_REQUIREMENT imply that store requirement
         // is present. We never write SHARESAFE_REQUIREMENT for a repo if store
         // is not present, refer checkrequirementscompat() for that
         //
         // However, if SHARESAFE_REQUIREMENT is not present, it means that the
         // repository was shared the old way. We check the share source
         // .hg/requires for SHARESAFE_REQUIREMENT to detect whether the
         // current repository needs to be reshared
         let share_safe = reqs.contains(requirements::SHARESAFE_REQUIREMENT);
 
         let store_path;
         if !shared {
             store_path = dot_hg.join("store");
         } else {
             let bytes = hg_vfs.read("sharedpath")?;
             let mut shared_path =
                 get_path_from_bytes(bytes.trim_end_matches(|b| b == b'\n'))
                     .to_owned();
             if relative {
                 shared_path = dot_hg.join(shared_path)
             }
             if !is_dir(&shared_path)? {
                 return Err(HgError::corrupted(format!(
                     ".hg/sharedpath points to nonexistent directory {}",
                     shared_path.display()
                 ))
                 .into());
             }
 
             store_path = shared_path.join("store");
 
             let source_is_share_safe =
                 requirements::load(Vfs { base: &shared_path })?
                     .contains(requirements::SHARESAFE_REQUIREMENT);
 
             if share_safe && !source_is_share_safe {
                 return Err(match config
                     .get(b"share", b"safe-mismatch.source-not-safe")
                 {
                     Some(b"abort") | None => HgError::abort(
                         "abort: share source does not support share-safe requirement\n\
                         (see `hg help config.format.use-share-safe` for more information)",
                         exit_codes::ABORT,
                     ),
                     _ => HgError::unsupported("share-safe downgrade"),
                 }
                 .into());
             } else if source_is_share_safe && !share_safe {
                 return Err(
                     match config.get(b"share", b"safe-mismatch.source-safe") {
                         Some(b"abort") | None => HgError::abort(
                             "abort: version mismatch: source uses share-safe \
                             functionality while the current share does not\n\
                             (see `hg help config.format.use-share-safe` for more information)",
                         exit_codes::ABORT,
                         ),
                         _ => HgError::unsupported("share-safe upgrade"),
                     }
                     .into(),
                 );
             }
 
             if share_safe {
                 repo_config_files.insert(0, shared_path.join("hgrc"))
             }
         }
         if share_safe {
             reqs.extend(requirements::load(Vfs { base: &store_path })?);
         }
 
         let repo_config = if std::env::var_os("HGRCSKIPREPO").is_none() {
             config.combine_with_repo(&repo_config_files)?
         } else {
             config.clone()
         };
 
         let repo = Self {
             requirements: reqs,
             working_directory,
             store: store_path,
             dot_hg,
             config: repo_config,
             dirstate_parents: LazyCell::new(Self::read_dirstate_parents),
+            dirstate_data_file_uuid: LazyCell::new(
+                Self::read_dirstate_data_file_uuid,
+            ),
             dirstate_map: LazyCell::new(Self::new_dirstate_map),
             changelog: LazyCell::new(Changelog::open),
             manifestlog: LazyCell::new(Manifestlog::open),
         };
 
         requirements::check(&repo)?;
 
         Ok(repo)
     }
 
     pub fn working_directory_path(&self) -> &Path {
         &self.working_directory
     }
 
     pub fn requirements(&self) -> &HashSet<String> {
         &self.requirements
     }
 
     pub fn config(&self) -> &Config {
         &self.config
     }
 
     /// For accessing repository files (in `.hg`), except for the store
     /// (`.hg/store`).
     pub fn hg_vfs(&self) -> Vfs<'_> {
         Vfs { base: &self.dot_hg }
     }
 
     /// For accessing repository store files (in `.hg/store`)
     pub fn store_vfs(&self) -> Vfs<'_> {
         Vfs { base: &self.store }
     }
 
     /// For accessing the working copy
     pub fn working_directory_vfs(&self) -> Vfs<'_> {
         Vfs {
             base: &self.working_directory,
         }
     }
 
     pub fn try_with_wlock_no_wait<R>(
         &self,
         f: impl FnOnce() -> R,
     ) -> Result<R, LockError> {
         try_with_lock_no_wait(self.hg_vfs(), "wlock", f)
     }
 
     pub fn has_dirstate_v2(&self) -> bool {
         self.requirements
             .contains(requirements::DIRSTATE_V2_REQUIREMENT)
     }
 
     pub fn has_sparse(&self) -> bool {
         self.requirements.contains(requirements::SPARSE_REQUIREMENT)
     }
 
     pub fn has_narrow(&self) -> bool {
         self.requirements.contains(requirements::NARROW_REQUIREMENT)
     }
 
     fn dirstate_file_contents(&self) -> Result<Vec<u8>, HgError> {
         Ok(self
             .hg_vfs()
             .read("dirstate")
             .io_not_found_as_none()?
             .unwrap_or(Vec::new()))
     }
 
     pub fn dirstate_parents(&self) -> Result<DirstateParents, HgError> {
         Ok(*self.dirstate_parents.get_or_init(self)?)
     }
 
     fn read_dirstate_parents(&self) -> Result<DirstateParents, HgError> {
         let dirstate = self.dirstate_file_contents()?;
         let parents = if dirstate.is_empty() {
+            if self.has_dirstate_v2() {
+                self.dirstate_data_file_uuid.set(None);
+            }
             DirstateParents::NULL
         } else if self.has_dirstate_v2() {
-            crate::dirstate_tree::on_disk::read_docket(&dirstate)?.parents()
+            let docket =
+                crate::dirstate_tree::on_disk::read_docket(&dirstate)?;
+            self.dirstate_data_file_uuid
+                .set(Some(docket.uuid.to_owned()));
+            docket.parents()
         } else {
             crate::dirstate::parsers::parse_dirstate_parents(&dirstate)?
                 .clone()
         };
         self.dirstate_parents.set(parents);
         Ok(parents)
     }
 
+    fn read_dirstate_data_file_uuid(
+        &self,
+    ) -> Result<Option<Vec<u8>>, HgError> {
+        assert!(
+            self.has_dirstate_v2(),
+            "accessing dirstate data file ID without dirstate-v2"
+        );
+        let dirstate = self.dirstate_file_contents()?;
+        if dirstate.is_empty() {
+            self.dirstate_parents.set(DirstateParents::NULL);
+            Ok(None)
+        } else {
+            let docket =
+                crate::dirstate_tree::on_disk::read_docket(&dirstate)?;
+            self.dirstate_parents.set(docket.parents());
+            Ok(Some(docket.uuid.to_owned()))
+        }
+    }
+
     fn new_dirstate_map(&self) -> Result<OwningDirstateMap, DirstateError> {
         let dirstate_file_contents = self.dirstate_file_contents()?;
         if dirstate_file_contents.is_empty() {
             self.dirstate_parents.set(DirstateParents::NULL);
+            if self.has_dirstate_v2() {
+                self.dirstate_data_file_uuid.set(None);
+            }
             Ok(OwningDirstateMap::new_empty(Vec::new()))
         } else if self.has_dirstate_v2() {
             let docket = crate::dirstate_tree::on_disk::read_docket(
                 &dirstate_file_contents,
             )?;
             self.dirstate_parents.set(docket.parents());
+            self.dirstate_data_file_uuid
+                .set(Some(docket.uuid.to_owned()));
             let data_size = docket.data_size();
             let metadata = docket.tree_metadata();
             let mut map = if let Some(data_mmap) = self
                 .hg_vfs()
                 .mmap_open(docket.data_filename())
                 .io_not_found_as_none()?
             {
                 OwningDirstateMap::new_empty(data_mmap)
             } else {
                 OwningDirstateMap::new_empty(Vec::new())
             };
             let (on_disk, placeholder) = map.get_pair_mut();
             *placeholder = DirstateMap::new_v2(on_disk, data_size, metadata)?;
             Ok(map)
         } else {
             let mut map = OwningDirstateMap::new_empty(dirstate_file_contents);
             let (on_disk, placeholder) = map.get_pair_mut();
             let (inner, parents) = DirstateMap::new_v1(on_disk)?;
             self.dirstate_parents
                 .set(parents.unwrap_or(DirstateParents::NULL));
             *placeholder = inner;
             Ok(map)
         }
     }
 
     pub fn dirstate_map(
         &self,
     ) -> Result<Ref<OwningDirstateMap>, DirstateError> {
         self.dirstate_map.get_or_init(self)
     }
 
     pub fn dirstate_map_mut(
         &self,
     ) -> Result<RefMut<OwningDirstateMap>, DirstateError> {
         self.dirstate_map.get_mut_or_init(self)
     }
 
     pub fn changelog(&self) -> Result<Ref<Changelog>, HgError> {
         self.changelog.get_or_init(self)
     }
 
     pub fn changelog_mut(&self) -> Result<RefMut<Changelog>, HgError> {
         self.changelog.get_mut_or_init(self)
     }
 
     pub fn manifestlog(&self) -> Result<Ref<Manifestlog>, HgError> {
         self.manifestlog.get_or_init(self)
     }
 
     pub fn manifestlog_mut(&self) -> Result<RefMut<Manifestlog>, HgError> {
         self.manifestlog.get_mut_or_init(self)
     }
 
     /// Returns the manifest of the *changeset* with the given node ID
     pub fn manifest_for_node(
         &self,
         node: impl Into<NodePrefix>,
     ) -> Result<Manifest, RevlogError> {
         self.manifestlog()?.data_for_node(
             self.changelog()?
                 .data_for_node(node.into())?
                 .manifest_node()?
                 .into(),
         )
     }
 
     /// Returns the manifest of the *changeset* with the given revision number
     pub fn manifest_for_rev(
         &self,
         revision: Revision,
     ) -> Result<Manifest, RevlogError> {
         self.manifestlog()?.data_for_node(
             self.changelog()?
                 .data_for_rev(revision)?
                 .manifest_node()?
                 .into(),
         )
     }
 
     pub fn filelog(&self, path: &HgPath) -> Result<Filelog, HgError> {
         Filelog::open(self, path)
     }
 }
 
 /// Lazily-initialized component of `Repo` with interior mutability
 ///
 /// This differs from `OnceCell` in that the value can still be "deinitialized"
 /// later by setting its inner `Option` to `None`.
 struct LazyCell<T, E> {
     value: RefCell<Option<T>>,
     // `Fn`s that don’t capture environment are zero-size, so this box does
     // not allocate:
     init: Box<dyn Fn(&Repo) -> Result<T, E>>,
 }
 
 impl<T, E> LazyCell<T, E> {
     fn new(init: impl Fn(&Repo) -> Result<T, E> + 'static) -> Self {
         Self {
             value: RefCell::new(None),
             init: Box::new(init),
         }
     }
 
     fn set(&self, value: T) {
         *self.value.borrow_mut() = Some(value)
     }
 
     fn get_or_init(&self, repo: &Repo) -> Result<Ref<T>, E> {
         let mut borrowed = self.value.borrow();
         if borrowed.is_none() {
             drop(borrowed);
             // Only use `borrow_mut` if it is really needed to avoid panic in
             // case there is another outstanding borrow but mutation is not
             // needed.
             *self.value.borrow_mut() = Some((self.init)(repo)?);
             borrowed = self.value.borrow()
         }
         Ok(Ref::map(borrowed, |option| option.as_ref().unwrap()))
     }
 
     fn get_mut_or_init(&self, repo: &Repo) -> Result<RefMut<T>, E> {
         let mut borrowed = self.value.borrow_mut();
         if borrowed.is_none() {
             *borrowed = Some((self.init)(repo)?);
         }
         Ok(RefMut::map(borrowed, |option| option.as_mut().unwrap()))
     }
 }