rhg: Add Repo::write_dirstate

This method is not used yet. It saves to disk any mutation that was done to
the `Repo`’s dirstate through `Repo::dirstate_map_mut`. It takes care of
dirstate-v1 v.s. dirstate-v2, dockets, data files, appending when possible,
etc.

Differential Revision: https://phab.mercurial-scm.org/D11839

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

@@ -1,1139 +1,1139 b''
 use bytes_cast::BytesCast;
 use micro_timer::timed;
 use std::borrow::Cow;
 use std::path::PathBuf;
 
 use super::on_disk;
 use super::on_disk::DirstateV2ParseError;
 use super::owning::OwningDirstateMap;
 use super::path_with_basename::WithBasename;
 use crate::dirstate::parsers::pack_entry;
 use crate::dirstate::parsers::packed_entry_size;
 use crate::dirstate::parsers::parse_dirstate_entries;
 use crate::dirstate::CopyMapIter;
 use crate::dirstate::StateMapIter;
 use crate::dirstate::TruncatedTimestamp;
 use crate::dirstate::SIZE_FROM_OTHER_PARENT;
 use crate::dirstate::SIZE_NON_NORMAL;
 use crate::matchers::Matcher;
 use crate::utils::hg_path::{HgPath, HgPathBuf};
 use crate::DirstateEntry;
 use crate::DirstateError;
 use crate::DirstateParents;
 use crate::DirstateStatus;
 use crate::EntryState;
 use crate::FastHashMap;
 use crate::PatternFileWarning;
 use crate::StatusError;
 use crate::StatusOptions;
 
 /// Append to an existing data file if the amount of unreachable data (not used
 /// anymore) is less than this fraction of the total amount of existing data.
 const ACCEPTABLE_UNREACHABLE_BYTES_RATIO: f32 = 0.5;
 
 pub struct DirstateMap<'on_disk> {
     /// Contents of the `.hg/dirstate` file
     pub(super) on_disk: &'on_disk [u8],
 
     pub(super) root: ChildNodes<'on_disk>,
 
     /// Number of nodes anywhere in the tree that have `.entry.is_some()`.
     pub(super) nodes_with_entry_count: u32,
 
     /// Number of nodes anywhere in the tree that have
     /// `.copy_source.is_some()`.
     pub(super) nodes_with_copy_source_count: u32,
 
     /// See on_disk::Header
     pub(super) ignore_patterns_hash: on_disk::IgnorePatternsHash,
 
     /// How many bytes of `on_disk` are not used anymore
     pub(super) unreachable_bytes: u32,
 }
 
 /// Using a plain `HgPathBuf` of the full path from the repository root as a
 /// map key would also work: all paths in a given map have the same parent
 /// path, so comparing full paths gives the same result as comparing base
 /// names. However `HashMap` would waste time always re-hashing the same
 /// string prefix.
 pub(super) type NodeKey<'on_disk> = WithBasename<Cow<'on_disk, HgPath>>;
 
 /// Similar to `&'tree Cow<'on_disk, HgPath>`, but can also be returned
 /// for on-disk nodes that don’t actually have a `Cow` to borrow.
 pub(super) enum BorrowedPath<'tree, 'on_disk> {
     InMemory(&'tree HgPathBuf),
     OnDisk(&'on_disk HgPath),
 }
 
 pub(super) enum ChildNodes<'on_disk> {
     InMemory(FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>),
     OnDisk(&'on_disk [on_disk::Node]),
 }
 
 pub(super) enum ChildNodesRef<'tree, 'on_disk> {
     InMemory(&'tree FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>),
     OnDisk(&'on_disk [on_disk::Node]),
 }
 
 pub(super) enum NodeRef<'tree, 'on_disk> {
     InMemory(&'tree NodeKey<'on_disk>, &'tree Node<'on_disk>),
     OnDisk(&'on_disk on_disk::Node),
 }
 
 impl<'tree, 'on_disk> BorrowedPath<'tree, 'on_disk> {
     pub fn detach_from_tree(&self) -> Cow<'on_disk, HgPath> {
         match *self {
             BorrowedPath::InMemory(in_memory) => Cow::Owned(in_memory.clone()),
             BorrowedPath::OnDisk(on_disk) => Cow::Borrowed(on_disk),
         }
     }
 }
 
 impl<'tree, 'on_disk> std::ops::Deref for BorrowedPath<'tree, 'on_disk> {
     type Target = HgPath;
 
     fn deref(&self) -> &HgPath {
         match *self {
             BorrowedPath::InMemory(in_memory) => in_memory,
             BorrowedPath::OnDisk(on_disk) => on_disk,
         }
     }
 }
 
 impl Default for ChildNodes<'_> {
     fn default() -> Self {
         ChildNodes::InMemory(Default::default())
     }
 }
 
 impl<'on_disk> ChildNodes<'on_disk> {
     pub(super) fn as_ref<'tree>(
         &'tree self,
     ) -> ChildNodesRef<'tree, 'on_disk> {
         match self {
             ChildNodes::InMemory(nodes) => ChildNodesRef::InMemory(nodes),
             ChildNodes::OnDisk(nodes) => ChildNodesRef::OnDisk(nodes),
         }
     }
 
     pub(super) fn is_empty(&self) -> bool {
         match self {
             ChildNodes::InMemory(nodes) => nodes.is_empty(),
             ChildNodes::OnDisk(nodes) => nodes.is_empty(),
         }
     }
 
     fn make_mut(
         &mut self,
         on_disk: &'on_disk [u8],
         unreachable_bytes: &mut u32,
     ) -> Result<
         &mut FastHashMap<NodeKey<'on_disk>, Node<'on_disk>>,
         DirstateV2ParseError,
     > {
         match self {
             ChildNodes::InMemory(nodes) => Ok(nodes),
             ChildNodes::OnDisk(nodes) => {
                 *unreachable_bytes +=
                     std::mem::size_of_val::<[on_disk::Node]>(nodes) as u32;
                 let nodes = nodes
                     .iter()
                     .map(|node| {
                         Ok((
                             node.path(on_disk)?,
                             node.to_in_memory_node(on_disk)?,
                         ))
                     })
                     .collect::<Result<_, _>>()?;
                 *self = ChildNodes::InMemory(nodes);
                 match self {
                     ChildNodes::InMemory(nodes) => Ok(nodes),
                     ChildNodes::OnDisk(_) => unreachable!(),
                 }
             }
         }
     }
 }
 
 impl<'tree, 'on_disk> ChildNodesRef<'tree, 'on_disk> {
     pub(super) fn get(
         &self,
         base_name: &HgPath,
         on_disk: &'on_disk [u8],
     ) -> Result<Option<NodeRef<'tree, 'on_disk>>, DirstateV2ParseError> {
         match self {
             ChildNodesRef::InMemory(nodes) => Ok(nodes
                 .get_key_value(base_name)
                 .map(|(k, v)| NodeRef::InMemory(k, v))),
             ChildNodesRef::OnDisk(nodes) => {
                 let mut parse_result = Ok(());
                 let search_result = nodes.binary_search_by(|node| {
                     match node.base_name(on_disk) {
                         Ok(node_base_name) => node_base_name.cmp(base_name),
                         Err(e) => {
                             parse_result = Err(e);
                             // Dummy comparison result, `search_result` won’t
                             // be used since `parse_result` is an error
                             std::cmp::Ordering::Equal
                         }
                     }
                 });
                 parse_result.map(|()| {
                     search_result.ok().map(|i| NodeRef::OnDisk(&nodes[i]))
                 })
             }
         }
     }
 
     /// Iterate in undefined order
     pub(super) fn iter(
         &self,
     ) -> impl Iterator<Item = NodeRef<'tree, 'on_disk>> {
         match self {
             ChildNodesRef::InMemory(nodes) => itertools::Either::Left(
                 nodes.iter().map(|(k, v)| NodeRef::InMemory(k, v)),
             ),
             ChildNodesRef::OnDisk(nodes) => {
                 itertools::Either::Right(nodes.iter().map(NodeRef::OnDisk))
             }
         }
     }
 
     /// Iterate in parallel in undefined order
     pub(super) fn par_iter(
         &self,
     ) -> impl rayon::iter::ParallelIterator<Item = NodeRef<'tree, 'on_disk>>
     {
         use rayon::prelude::*;
         match self {
             ChildNodesRef::InMemory(nodes) => rayon::iter::Either::Left(
                 nodes.par_iter().map(|(k, v)| NodeRef::InMemory(k, v)),
             ),
             ChildNodesRef::OnDisk(nodes) => rayon::iter::Either::Right(
                 nodes.par_iter().map(NodeRef::OnDisk),
             ),
         }
     }
 
     pub(super) fn sorted(&self) -> Vec<NodeRef<'tree, 'on_disk>> {
         match self {
             ChildNodesRef::InMemory(nodes) => {
                 let mut vec: Vec<_> = nodes
                     .iter()
                     .map(|(k, v)| NodeRef::InMemory(k, v))
                     .collect();
                 fn sort_key<'a>(node: &'a NodeRef) -> &'a HgPath {
                     match node {
                         NodeRef::InMemory(path, _node) => path.base_name(),
                         NodeRef::OnDisk(_) => unreachable!(),
                     }
                 }
                 // `sort_unstable_by_key` doesn’t allow keys borrowing from the
                 // value: https://github.com/rust-lang/rust/issues/34162
                 vec.sort_unstable_by(|a, b| sort_key(a).cmp(sort_key(b)));
                 vec
             }
             ChildNodesRef::OnDisk(nodes) => {
                 // Nodes on disk are already sorted
                 nodes.iter().map(NodeRef::OnDisk).collect()
             }
         }
     }
 }
 
 impl<'tree, 'on_disk> NodeRef<'tree, 'on_disk> {
     pub(super) fn full_path(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'tree HgPath, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => Ok(path.full_path()),
             NodeRef::OnDisk(node) => node.full_path(on_disk),
         }
     }
 
     /// Returns a `BorrowedPath`, which can be turned into a `Cow<'on_disk,
     /// HgPath>` detached from `'tree`
     pub(super) fn full_path_borrowed(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<BorrowedPath<'tree, 'on_disk>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => match path.full_path() {
                 Cow::Borrowed(on_disk) => Ok(BorrowedPath::OnDisk(on_disk)),
                 Cow::Owned(in_memory) => Ok(BorrowedPath::InMemory(in_memory)),
             },
             NodeRef::OnDisk(node) => {
                 Ok(BorrowedPath::OnDisk(node.full_path(on_disk)?))
             }
         }
     }
 
     pub(super) fn base_name(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<&'tree HgPath, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(path, _node) => Ok(path.base_name()),
             NodeRef::OnDisk(node) => node.base_name(on_disk),
         }
     }
 
     pub(super) fn children(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<ChildNodesRef<'tree, 'on_disk>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => Ok(node.children.as_ref()),
             NodeRef::OnDisk(node) => {
                 Ok(ChildNodesRef::OnDisk(node.children(on_disk)?))
             }
         }
     }
 
     pub(super) fn has_copy_source(&self) -> bool {
         match self {
             NodeRef::InMemory(_path, node) => node.copy_source.is_some(),
             NodeRef::OnDisk(node) => node.has_copy_source(),
         }
     }
 
     pub(super) fn copy_source(
         &self,
         on_disk: &'on_disk [u8],
     ) -> Result<Option<&'tree HgPath>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => {
                 Ok(node.copy_source.as_ref().map(|s| &**s))
             }
             NodeRef::OnDisk(node) => node.copy_source(on_disk),
         }
     }
 
     pub(super) fn entry(
         &self,
     ) -> Result<Option<DirstateEntry>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => {
                 Ok(node.data.as_entry().copied())
             }
             NodeRef::OnDisk(node) => node.entry(),
         }
     }
 
     pub(super) fn state(
         &self,
     ) -> Result<Option<EntryState>, DirstateV2ParseError> {
         Ok(self.entry()?.map(|e| e.state()))
     }
 
     pub(super) fn cached_directory_mtime(
         &self,
     ) -> Result<Option<TruncatedTimestamp>, DirstateV2ParseError> {
         match self {
             NodeRef::InMemory(_path, node) => Ok(match node.data {
                 NodeData::CachedDirectory { mtime } => Some(mtime),
                 _ => None,
             }),
             NodeRef::OnDisk(node) => node.cached_directory_mtime(),
         }
     }
 
     pub(super) fn descendants_with_entry_count(&self) -> u32 {
         match self {
             NodeRef::InMemory(_path, node) => {
                 node.descendants_with_entry_count
             }
             NodeRef::OnDisk(node) => node.descendants_with_entry_count.get(),
         }
     }
 
     pub(super) fn tracked_descendants_count(&self) -> u32 {
         match self {
             NodeRef::InMemory(_path, node) => node.tracked_descendants_count,
             NodeRef::OnDisk(node) => node.tracked_descendants_count.get(),
         }
     }
 }
 
 /// Represents a file or a directory
 #[derive(Default)]
 pub(super) struct Node<'on_disk> {
     pub(super) data: NodeData,
 
     pub(super) copy_source: Option<Cow<'on_disk, HgPath>>,
 
     pub(super) children: ChildNodes<'on_disk>,
 
     /// How many (non-inclusive) descendants of this node have an entry.
     pub(super) descendants_with_entry_count: u32,
 
     /// How many (non-inclusive) descendants of this node have an entry whose
     /// state is "tracked".
     pub(super) tracked_descendants_count: u32,
 }
 
 pub(super) enum NodeData {
     Entry(DirstateEntry),
     CachedDirectory { mtime: TruncatedTimestamp },
     None,
 }
 
 impl Default for NodeData {
     fn default() -> Self {
         NodeData::None
     }
 }
 
 impl NodeData {
     fn has_entry(&self) -> bool {
         match self {
             NodeData::Entry(_) => true,
             _ => false,
         }
     }
 
     fn as_entry(&self) -> Option<&DirstateEntry> {
         match self {
             NodeData::Entry(entry) => Some(entry),
             _ => None,
         }
     }
 }
 
 impl<'on_disk> DirstateMap<'on_disk> {
     pub(super) fn empty(on_disk: &'on_disk [u8]) -> Self {
         Self {
             on_disk,
             root: ChildNodes::default(),
             nodes_with_entry_count: 0,
             nodes_with_copy_source_count: 0,
             ignore_patterns_hash: [0; on_disk::IGNORE_PATTERNS_HASH_LEN],
             unreachable_bytes: 0,
         }
     }
 
     #[timed]
     pub fn new_v2(
         on_disk: &'on_disk [u8],
         data_size: usize,
         metadata: &[u8],
     ) -> Result<Self, DirstateError> {
         if let Some(data) = on_disk.get(..data_size) {
             Ok(on_disk::read(data, metadata)?)
         } else {
             Err(DirstateV2ParseError.into())
         }
     }
 
     #[timed]
     pub fn new_v1(
         on_disk: &'on_disk [u8],
     ) -> Result<(Self, Option<DirstateParents>), DirstateError> {
         let mut map = Self::empty(on_disk);
         if map.on_disk.is_empty() {
             return Ok((map, None));
         }
 
         let parents = parse_dirstate_entries(
             map.on_disk,
             |path, entry, copy_source| {
                 let tracked = entry.state().is_tracked();
                 let node = Self::get_or_insert_node(
                     map.on_disk,
                     &mut map.unreachable_bytes,
                     &mut map.root,
                     path,
                     WithBasename::to_cow_borrowed,
                     |ancestor| {
                         if tracked {
                             ancestor.tracked_descendants_count += 1
                         }
                         ancestor.descendants_with_entry_count += 1
                     },
                 )?;
                 assert!(
                     !node.data.has_entry(),
                     "duplicate dirstate entry in read"
                 );
                 assert!(
                     node.copy_source.is_none(),
                     "duplicate dirstate entry in read"
                 );
                 node.data = NodeData::Entry(*entry);
                 node.copy_source = copy_source.map(Cow::Borrowed);
                 map.nodes_with_entry_count += 1;
                 if copy_source.is_some() {
                     map.nodes_with_copy_source_count += 1
                 }
                 Ok(())
             },
         )?;
         let parents = Some(parents.clone());
 
         Ok((map, parents))
     }
 
     /// Assuming dirstate-v2 format, returns whether the next write should
     /// append to the existing data file that contains `self.on_disk` (true),
     /// or create a new data file from scratch (false).
     pub(super) fn write_should_append(&self) -> bool {
         let ratio = self.unreachable_bytes as f32 / self.on_disk.len() as f32;
         ratio < ACCEPTABLE_UNREACHABLE_BYTES_RATIO
     }
 
     fn get_node<'tree>(
         &'tree self,
         path: &HgPath,
     ) -> Result<Option<NodeRef<'tree, 'on_disk>>, DirstateV2ParseError> {
         let mut children = self.root.as_ref();
         let mut components = path.components();
         let mut component =
             components.next().expect("expected at least one components");
         loop {
             if let Some(child) = children.get(component, self.on_disk)? {
                 if let Some(next_component) = components.next() {
                     component = next_component;
                     children = child.children(self.on_disk)?;
                 } else {
                     return Ok(Some(child));
                 }
             } else {
                 return Ok(None);
             }
         }
     }
 
     /// Returns a mutable reference to the node at `path` if it exists
     ///
     /// This takes `root` instead of `&mut self` so that callers can mutate
     /// other fields while the returned borrow is still valid
     fn get_node_mut<'tree>(
         on_disk: &'on_disk [u8],
         unreachable_bytes: &mut u32,
         root: &'tree mut ChildNodes<'on_disk>,
         path: &HgPath,
     ) -> Result<Option<&'tree mut Node<'on_disk>>, DirstateV2ParseError> {
         let mut children = root;
         let mut components = path.components();
         let mut component =
             components.next().expect("expected at least one components");
         loop {
             if let Some(child) = children
                 .make_mut(on_disk, unreachable_bytes)?
                 .get_mut(component)
             {
                 if let Some(next_component) = components.next() {
                     component = next_component;
                     children = &mut child.children;
                 } else {
                     return Ok(Some(child));
                 }
             } else {
                 return Ok(None);
             }
         }
     }
 
     pub(super) fn get_or_insert<'tree, 'path>(
         &'tree mut self,
         path: &HgPath,
     ) -> Result<&'tree mut Node<'on_disk>, DirstateV2ParseError> {
         Self::get_or_insert_node(
             self.on_disk,
             &mut self.unreachable_bytes,
             &mut self.root,
             path,
             WithBasename::to_cow_owned,
             |_| {},
         )
     }
 
     fn get_or_insert_node<'tree, 'path>(
         on_disk: &'on_disk [u8],
         unreachable_bytes: &mut u32,
         root: &'tree mut ChildNodes<'on_disk>,
         path: &'path HgPath,
         to_cow: impl Fn(
             WithBasename<&'path HgPath>,
         ) -> WithBasename<Cow<'on_disk, HgPath>>,
         mut each_ancestor: impl FnMut(&mut Node),
     ) -> Result<&'tree mut Node<'on_disk>, DirstateV2ParseError> {
         let mut child_nodes = root;
         let mut inclusive_ancestor_paths =
             WithBasename::inclusive_ancestors_of(path);
         let mut ancestor_path = inclusive_ancestor_paths
             .next()
             .expect("expected at least one inclusive ancestor");
         loop {
             // TODO: can we avoid allocating an owned key in cases where the
             // map already contains that key, without introducing double
             // lookup?
             let child_node = child_nodes
                 .make_mut(on_disk, unreachable_bytes)?
                 .entry(to_cow(ancestor_path))
                 .or_default();
             if let Some(next) = inclusive_ancestor_paths.next() {
                 each_ancestor(child_node);
                 ancestor_path = next;
                 child_nodes = &mut child_node.children;
             } else {
                 return Ok(child_node);
             }
         }
     }
 
     fn add_or_remove_file(
         &mut self,
         path: &HgPath,
         old_state: Option<EntryState>,
         new_entry: DirstateEntry,
     ) -> Result<(), DirstateV2ParseError> {
         let had_entry = old_state.is_some();
         let was_tracked = old_state.map_or(false, |s| s.is_tracked());
         let tracked_count_increment =
             match (was_tracked, new_entry.state().is_tracked()) {
                 (false, true) => 1,
                 (true, false) => -1,
                 _ => 0,
             };
 
         let node = Self::get_or_insert_node(
             self.on_disk,
             &mut self.unreachable_bytes,
             &mut self.root,
             path,
             WithBasename::to_cow_owned,
             |ancestor| {
                 if !had_entry {
                     ancestor.descendants_with_entry_count += 1;
                 }
 
                 // We can’t use `+= increment` because the counter is unsigned,
                 // and we want debug builds to detect accidental underflow
                 // through zero
                 match tracked_count_increment {
                     1 => ancestor.tracked_descendants_count += 1,
                     -1 => ancestor.tracked_descendants_count -= 1,
                     _ => {}
                 }
             },
         )?;
         if !had_entry {
             self.nodes_with_entry_count += 1
         }
         node.data = NodeData::Entry(new_entry);
         Ok(())
     }
 
     fn iter_nodes<'tree>(
         &'tree self,
     ) -> impl Iterator<
         Item = Result<NodeRef<'tree, 'on_disk>, DirstateV2ParseError>,
     > + 'tree {
         // Depth first tree traversal.
         //
         // If we could afford internal iteration and recursion,
         // this would look like:
         //
         // ```
         // fn traverse_children(
         //     children: &ChildNodes,
         //     each: &mut impl FnMut(&Node),
         // ) {
         //     for child in children.values() {
         //         traverse_children(&child.children, each);
         //         each(child);
         //     }
         // }
         // ```
         //
         // However we want an external iterator and therefore can’t use the
         // call stack. Use an explicit stack instead:
         let mut stack = Vec::new();
         let mut iter = self.root.as_ref().iter();
         std::iter::from_fn(move || {
             while let Some(child_node) = iter.next() {
                 let children = match child_node.children(self.on_disk) {
                     Ok(children) => children,
                     Err(error) => return Some(Err(error)),
                 };
                 // Pseudo-recursion
                 let new_iter = children.iter();
                 let old_iter = std::mem::replace(&mut iter, new_iter);
                 stack.push((child_node, old_iter));
             }
             // Found the end of a `children.iter()` iterator.
             if let Some((child_node, next_iter)) = stack.pop() {
                 // "Return" from pseudo-recursion by restoring state from the
                 // explicit stack
                 iter = next_iter;
 
                 Some(Ok(child_node))
             } else {
                 // Reached the bottom of the stack, we’re done
                 None
             }
         })
     }
 
     fn count_dropped_path(unreachable_bytes: &mut u32, path: &Cow<HgPath>) {
         if let Cow::Borrowed(path) = path {
             *unreachable_bytes += path.len() as u32
         }
     }
 }
 
 /// Like `Iterator::filter_map`, but over a fallible iterator of `Result`s.
 ///
 /// The callback is only called for incoming `Ok` values. Errors are passed
 /// through as-is. In order to let it use the `?` operator the callback is
 /// expected to return a `Result` of `Option`, instead of an `Option` of
 /// `Result`.
 fn filter_map_results<'a, I, F, A, B, E>(
     iter: I,
     f: F,
 ) -> impl Iterator<Item = Result<B, E>> + 'a
 where
     I: Iterator<Item = Result<A, E>> + 'a,
     F: Fn(A) -> Result<Option<B>, E> + 'a,
 {
     iter.filter_map(move |result| match result {
         Ok(node) => f(node).transpose(),
         Err(e) => Some(Err(e)),
     })
 }
 
 impl OwningDirstateMap {
     pub fn clear(&mut self) {
         let map = self.get_map_mut();
         map.root = Default::default();
         map.nodes_with_entry_count = 0;
         map.nodes_with_copy_source_count = 0;
     }
 
     pub fn set_entry(
         &mut self,
         filename: &HgPath,
         entry: DirstateEntry,
     ) -> Result<(), DirstateV2ParseError> {
         let map = self.get_map_mut();
         map.get_or_insert(&filename)?.data = NodeData::Entry(entry);
         Ok(())
     }
 
     pub fn add_file(
         &mut self,
         filename: &HgPath,
         entry: DirstateEntry,
     ) -> Result<(), DirstateError> {
         let old_state = self.get(filename)?.map(|e| e.state());
         let map = self.get_map_mut();
         Ok(map.add_or_remove_file(filename, old_state, entry)?)
     }
 
     pub fn remove_file(
         &mut self,
         filename: &HgPath,
         in_merge: bool,
     ) -> Result<(), DirstateError> {
         let old_entry_opt = self.get(filename)?;
         let old_state = old_entry_opt.map(|e| e.state());
         let mut size = 0;
         if in_merge {
             // XXX we should not be able to have 'm' state and 'FROM_P2' if not
             // during a merge. So I (marmoute) am not sure we need the
             // conditionnal at all. Adding double checking this with assert
             // would be nice.
             if let Some(old_entry) = old_entry_opt {
                 // backup the previous state
                 if old_entry.state() == EntryState::Merged {
                     size = SIZE_NON_NORMAL;
                 } else if old_entry.state() == EntryState::Normal
                     && old_entry.size() == SIZE_FROM_OTHER_PARENT
                 {
                     // other parent
                     size = SIZE_FROM_OTHER_PARENT;
                 }
             }
         }
         if size == 0 {
             self.copy_map_remove(filename)?;
         }
         let map = self.get_map_mut();
         let entry = DirstateEntry::new_removed(size);
         Ok(map.add_or_remove_file(filename, old_state, entry)?)
     }
 
     pub fn drop_entry_and_copy_source(
         &mut self,
         filename: &HgPath,
     ) -> Result<(), DirstateError> {
         let was_tracked = self
             .get(filename)?
             .map_or(false, |e| e.state().is_tracked());
         let map = self.get_map_mut();
         struct Dropped {
             was_tracked: bool,
             had_entry: bool,
             had_copy_source: bool,
         }
 
         /// If this returns `Ok(Some((dropped, removed)))`, then
         ///
         /// * `dropped` is about the leaf node that was at `filename`
         /// * `removed` is whether this particular level of recursion just
         ///   removed a node in `nodes`.
         fn recur<'on_disk>(
             on_disk: &'on_disk [u8],
             unreachable_bytes: &mut u32,
             nodes: &mut ChildNodes<'on_disk>,
             path: &HgPath,
         ) -> Result<Option<(Dropped, bool)>, DirstateV2ParseError> {
             let (first_path_component, rest_of_path) =
                 path.split_first_component();
             let nodes = nodes.make_mut(on_disk, unreachable_bytes)?;
             let node = if let Some(node) = nodes.get_mut(first_path_component)
             {
                 node
             } else {
                 return Ok(None);
             };
             let dropped;
             if let Some(rest) = rest_of_path {
                 if let Some((d, removed)) = recur(
                     on_disk,
                     unreachable_bytes,
                     &mut node.children,
                     rest,
                 )? {
                     dropped = d;
                     if dropped.had_entry {
                         node.descendants_with_entry_count -= 1;
                     }
                     if dropped.was_tracked {
                         node.tracked_descendants_count -= 1;
                     }
 
                     // Directory caches must be invalidated when removing a
                     // child node
                     if removed {
                         if let NodeData::CachedDirectory { .. } = &node.data {
                             node.data = NodeData::None
                         }
                     }
                 } else {
                     return Ok(None);
                 }
             } else {
                 let had_entry = node.data.has_entry();
                 if had_entry {
                     node.data = NodeData::None
                 }
                 if let Some(source) = &node.copy_source {
                     DirstateMap::count_dropped_path(unreachable_bytes, source);
                     node.copy_source = None
                 }
                 dropped = Dropped {
                     was_tracked: node
                         .data
                         .as_entry()
                         .map_or(false, |entry| entry.state().is_tracked()),
                     had_entry,
                     had_copy_source: node.copy_source.take().is_some(),
                 };
             }
             // After recursion, for both leaf (rest_of_path is None) nodes and
             // parent nodes, remove a node if it just became empty.
             let remove = !node.data.has_entry()
                 && node.copy_source.is_none()
                 && node.children.is_empty();
             if remove {
                 let (key, _) =
                     nodes.remove_entry(first_path_component).unwrap();
                 DirstateMap::count_dropped_path(
                     unreachable_bytes,
                     key.full_path(),
                 )
             }
             Ok(Some((dropped, remove)))
         }
 
         if let Some((dropped, _removed)) = recur(
             map.on_disk,
             &mut map.unreachable_bytes,
             &mut map.root,
             filename,
         )? {
             if dropped.had_entry {
                 map.nodes_with_entry_count -= 1
             }
             if dropped.had_copy_source {
                 map.nodes_with_copy_source_count -= 1
             }
         } else {
             debug_assert!(!was_tracked);
         }
         Ok(())
     }
 
     pub fn has_tracked_dir(
         &mut self,
         directory: &HgPath,
     ) -> Result<bool, DirstateError> {
         let map = self.get_map_mut();
         if let Some(node) = map.get_node(directory)? {
             // A node without a `DirstateEntry` was created to hold child
             // nodes, and is therefore a directory.
             let state = node.state()?;
             Ok(state.is_none() && node.tracked_descendants_count() > 0)
         } else {
             Ok(false)
         }
     }
 
     pub fn has_dir(
         &mut self,
         directory: &HgPath,
     ) -> Result<bool, DirstateError> {
         let map = self.get_map_mut();
         if let Some(node) = map.get_node(directory)? {
             // A node without a `DirstateEntry` was created to hold child
             // nodes, and is therefore a directory.
             let state = node.state()?;
             Ok(state.is_none() && node.descendants_with_entry_count() > 0)
         } else {
             Ok(false)
         }
     }
 
     #[timed]
     pub fn pack_v1(
         &self,
         parents: DirstateParents,
     ) -> Result<Vec<u8>, DirstateError> {
         let map = self.get_map();
         // Optizimation (to be measured?): pre-compute size to avoid `Vec`
         // reallocations
         let mut size = parents.as_bytes().len();
         for node in map.iter_nodes() {
             let node = node?;
             if node.entry()?.is_some() {
                 size += packed_entry_size(
                     node.full_path(map.on_disk)?,
                     node.copy_source(map.on_disk)?,
                 );
             }
         }
 
         let mut packed = Vec::with_capacity(size);
         packed.extend(parents.as_bytes());
 
         for node in map.iter_nodes() {
             let node = node?;
             if let Some(entry) = node.entry()? {
                 pack_entry(
                     node.full_path(map.on_disk)?,
                     &entry,
                     node.copy_source(map.on_disk)?,
                     &mut packed,
                 );
             }
         }
         Ok(packed)
     }
 
     /// Returns new data and metadata together with whether that data should be
     /// appended to the existing data file whose content is at
     /// `map.on_disk` (true), instead of written to a new data file
     /// (false).
     #[timed]
     pub fn pack_v2(
         &self,
         can_append: bool,
-    ) -> Result<(Vec<u8>, Vec<u8>, bool), DirstateError> {
+    ) -> Result<(Vec<u8>, on_disk::TreeMetadata, bool), DirstateError> {
         let map = self.get_map();
         on_disk::write(map, can_append)
     }
 
     pub fn status<'a>(
         &'a mut self,
         matcher: &'a (dyn Matcher + Sync),
         root_dir: PathBuf,
         ignore_files: Vec<PathBuf>,
         options: StatusOptions,
     ) -> Result<(DirstateStatus<'a>, Vec<PatternFileWarning>), StatusError>
     {
         let map = self.get_map_mut();
         super::status::status(map, matcher, root_dir, ignore_files, options)
     }
 
     pub fn copy_map_len(&self) -> usize {
         let map = self.get_map();
         map.nodes_with_copy_source_count as usize
     }
 
     pub fn copy_map_iter(&self) -> CopyMapIter<'_> {
         let map = self.get_map();
         Box::new(filter_map_results(map.iter_nodes(), move |node| {
             Ok(if let Some(source) = node.copy_source(map.on_disk)? {
                 Some((node.full_path(map.on_disk)?, source))
             } else {
                 None
             })
         }))
     }
 
     pub fn copy_map_contains_key(
         &self,
         key: &HgPath,
     ) -> Result<bool, DirstateV2ParseError> {
         let map = self.get_map();
         Ok(if let Some(node) = map.get_node(key)? {
             node.has_copy_source()
         } else {
             false
         })
     }
 
     pub fn copy_map_get(
         &self,
         key: &HgPath,
     ) -> Result<Option<&HgPath>, DirstateV2ParseError> {
         let map = self.get_map();
         if let Some(node) = map.get_node(key)? {
             if let Some(source) = node.copy_source(map.on_disk)? {
                 return Ok(Some(source));
             }
         }
         Ok(None)
     }
 
     pub fn copy_map_remove(
         &mut self,
         key: &HgPath,
     ) -> Result<Option<HgPathBuf>, DirstateV2ParseError> {
         let map = self.get_map_mut();
         let count = &mut map.nodes_with_copy_source_count;
         let unreachable_bytes = &mut map.unreachable_bytes;
         Ok(DirstateMap::get_node_mut(
             map.on_disk,
             unreachable_bytes,
             &mut map.root,
             key,
         )?
         .and_then(|node| {
             if let Some(source) = &node.copy_source {
                 *count -= 1;
                 DirstateMap::count_dropped_path(unreachable_bytes, source);
             }
             node.copy_source.take().map(Cow::into_owned)
         }))
     }
 
     pub fn copy_map_insert(
         &mut self,
         key: HgPathBuf,
         value: HgPathBuf,
     ) -> Result<Option<HgPathBuf>, DirstateV2ParseError> {
         let map = self.get_map_mut();
         let node = DirstateMap::get_or_insert_node(
             map.on_disk,
             &mut map.unreachable_bytes,
             &mut map.root,
             &key,
             WithBasename::to_cow_owned,
             |_ancestor| {},
         )?;
         if node.copy_source.is_none() {
             map.nodes_with_copy_source_count += 1
         }
         Ok(node.copy_source.replace(value.into()).map(Cow::into_owned))
     }
 
     pub fn len(&self) -> usize {
         let map = self.get_map();
         map.nodes_with_entry_count as usize
     }
 
     pub fn contains_key(
         &self,
         key: &HgPath,
     ) -> Result<bool, DirstateV2ParseError> {
         Ok(self.get(key)?.is_some())
     }
 
     pub fn get(
         &self,
         key: &HgPath,
     ) -> Result<Option<DirstateEntry>, DirstateV2ParseError> {
         let map = self.get_map();
         Ok(if let Some(node) = map.get_node(key)? {
             node.entry()?
         } else {
             None
         })
     }
 
     pub fn iter(&self) -> StateMapIter<'_> {
         let map = self.get_map();
         Box::new(filter_map_results(map.iter_nodes(), move |node| {
             Ok(if let Some(entry) = node.entry()? {
                 Some((node.full_path(map.on_disk)?, entry))
             } else {
                 None
             })
         }))
     }
 
     pub fn iter_tracked_dirs(
         &mut self,
     ) -> Result<
         Box<
             dyn Iterator<Item = Result<&HgPath, DirstateV2ParseError>>
                 + Send
                 + '_,
         >,
         DirstateError,
     > {
         let map = self.get_map_mut();
         let on_disk = map.on_disk;
         Ok(Box::new(filter_map_results(
             map.iter_nodes(),
             move |node| {
                 Ok(if node.tracked_descendants_count() > 0 {
                     Some(node.full_path(on_disk)?)
                 } else {
                     None
                 })
             },
         )))
     }
 
     pub fn debug_iter(
         &self,
         all: bool,
     ) -> Box<
         dyn Iterator<
                 Item = Result<
                     (&HgPath, (u8, i32, i32, i32)),
                     DirstateV2ParseError,
                 >,
             > + Send
             + '_,
     > {
         let map = self.get_map();
         Box::new(filter_map_results(map.iter_nodes(), move |node| {
             let debug_tuple = if let Some(entry) = node.entry()? {
                 entry.debug_tuple()
             } else if !all {
                 return Ok(None);
             } else if let Some(mtime) = node.cached_directory_mtime()? {
                 (b' ', 0, -1, mtime.truncated_seconds() as i32)
             } else {
                 (b' ', 0, -1, -1)
             };
             Ok(Some((node.full_path(map.on_disk)?, debug_tuple)))
         }))
     }
 }

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

@@ -1,791 +1,837 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 rand::Rng;
 use std::borrow::Cow;
 use std::convert::{TryFrom, TryInto};
+use std::fmt::Write;
 
 /// 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,
     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 {
+pub 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 TreeMetadata {
+    pub fn as_bytes(&self) -> &[u8] {
+        BytesCast::as_bytes(self)
+    }
+}
+
 impl<'on_disk> Docket<'on_disk> {
+    /// Generate the identifier for a new data file
+    ///
+    /// TODO: support the `HGTEST_UUIDFILE` environment variable.
+    /// See `mercurial/revlogutils/docket.py`
+    pub fn new_uid() -> String {
+        const ID_LENGTH: usize = 8;
+        let mut id = String::with_capacity(ID_LENGTH);
+        let mut rng = rand::thread_rng();
+        for _ in 0..ID_LENGTH {
+            // One random hexadecimal digit.
+            // `unwrap` never panics because `impl Write for String`
+            // never returns an error.
+            write!(&mut id, "{:x}", rng.gen_range(0, 16)).unwrap();
+        }
+        id
+    }
+
+    pub fn serialize(
+        parents: DirstateParents,
+        tree_metadata: TreeMetadata,
+        data_size: u64,
+        uuid: &[u8],
+    ) -> Result<Vec<u8>, std::num::TryFromIntError> {
+        let header = DocketHeader {
+            marker: *V2_FORMAT_MARKER,
+            parent_1: parents.p1.pad_to_256_bits(),
+            parent_2: parents.p2.pad_to_256_bits(),
+            metadata: tree_metadata,
+            data_size: u32::try_from(data_size)?.into(),
+            uuid_size: uuid.len().try_into()?,
+        };
+        let header = header.as_bytes();
+        let mut docket = Vec::with_capacity(header.len() + uuid.len());
+        docket.extend_from_slice(header);
+        docket.extend_from_slice(uuid);
+        Ok(docket)
+    }
+
     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> {
+) -> Result<(Vec<u8>, TreeMetadata, 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))
+    Ok((writer.out, meta, 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

@@ -1,464 +1,532 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::on_disk::Docket as DirstateDocket;
 use crate::dirstate_tree::owning::OwningDirstateMap;
-use crate::errors::HgError;
 use crate::errors::HgResultExt;
+use crate::errors::{HgError, IoResultExt};
 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::io::Seek;
+use std::io::SeekFrom;
+use std::io::Write as IoWrite;
 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() {
             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)
     }
+
+    /// Write to disk any updates that were made through `dirstate_map_mut`.
+    ///
+    /// The "wlock" must be held while calling this.
+    /// See for example `try_with_wlock_no_wait`.
+    ///
+    /// TODO: have a `WritableRepo` type only accessible while holding the
+    /// lock?
+    pub fn write_dirstate(&self) -> Result<(), DirstateError> {
+        let map = self.dirstate_map()?;
+        // TODO: Maintain a `DirstateMap::dirty` flag, and return early here if
+        // it’s unset
+        let parents = self.dirstate_parents()?;
+        let packed_dirstate = if self.has_dirstate_v2() {
+            let uuid = self.dirstate_data_file_uuid.get_or_init(self)?;
+            let mut uuid = uuid.as_ref();
+            let can_append = uuid.is_some();
+            let (data, tree_metadata, append) = map.pack_v2(can_append)?;
+            if !append {
+                uuid = None
+            }
+            let uuid = if let Some(uuid) = uuid {
+                std::str::from_utf8(uuid)
+                    .map_err(|_| {
+                        HgError::corrupted("non-UTF-8 dirstate data file ID")
+                    })?
+                    .to_owned()
+            } else {
+                DirstateDocket::new_uid()
+            };
+            let data_filename = format!("dirstate.{}", uuid);
+            let data_filename = self.hg_vfs().join(data_filename);
+            let mut options = std::fs::OpenOptions::new();
+            if append {
+                options.append(true);
+            } else {
+                options.write(true).create_new(true);
+            }
+            let data_size = (|| {
+                // TODO: loop and try another random ID if !append and this
+                // returns `ErrorKind::AlreadyExists`? Collision chance of two
+                // random IDs is one in 2**32
+                let mut file = options.open(&data_filename)?;
+                file.write_all(&data)?;
+                file.flush()?;
+                // TODO: use https://doc.rust-lang.org/std/io/trait.Seek.html#method.stream_position when we require Rust 1.51+
+                file.seek(SeekFrom::Current(0))
+            })()
+            .when_writing_file(&data_filename)?;
+            DirstateDocket::serialize(
+                parents,
+                tree_metadata,
+                data_size,
+                uuid.as_bytes(),
+            )
+            .map_err(|_: std::num::TryFromIntError| {
+                HgError::corrupted("overflow in dirstate docket serialization")
+            })?
+        } else {
+            map.pack_v1(parents)?
+        };
+        self.hg_vfs().atomic_write("dirstate", &packed_dirstate)?;
+        Ok(())
+    }
 }
 
 /// 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()))
     }
 }

rust/hg-core/src/revlog/node.rs

@@ -1,415 +1,421 b''
 // Copyright 2019-2020 Georges Racinet <[e-mail]>
 //
 // This software may be used and distributed according to the terms of the
 // GNU General Public License version 2 or any later version.
 
 //! Definitions and utilities for Revision nodes
 //!
 //! In Mercurial code base, it is customary to call "a node" the binary SHA
 //! of a revision.
 
 use crate::errors::HgError;
 use bytes_cast::BytesCast;
 use std::convert::{TryFrom, TryInto};
 use std::fmt;
 
 /// The length in bytes of a `Node`
 ///
 /// This constant is meant to ease refactors of this module, and
 /// are private so that calling code does not expect all nodes have
 /// the same size, should we support several formats concurrently in
 /// the future.
 pub const NODE_BYTES_LENGTH: usize = 20;
 
 /// Id of the null node.
 ///
 /// Used to indicate the absence of node.
 pub const NULL_NODE_ID: [u8; NODE_BYTES_LENGTH] = [0u8; NODE_BYTES_LENGTH];
 
 /// The length in bytes of a `Node`
 ///
 /// see also `NODES_BYTES_LENGTH` about it being private.
 const NODE_NYBBLES_LENGTH: usize = 2 * NODE_BYTES_LENGTH;
 
 /// Default for UI presentation
 const SHORT_PREFIX_DEFAULT_NYBBLES_LENGTH: u8 = 12;
 
 /// Private alias for readability and to ease future change
 type NodeData = [u8; NODE_BYTES_LENGTH];
 
 /// Binary revision SHA
 ///
 /// ## Future changes of hash size
 ///
 /// To accomodate future changes of hash size, Rust callers
 /// should use the conversion methods at the boundaries (FFI, actual
 /// computation of hashes and I/O) only, and only if required.
 ///
 /// All other callers outside of unit tests should just handle `Node` values
 /// and never make any assumption on the actual length, using [`nybbles_len`]
 /// if they need a loop boundary.
 ///
 /// All methods that create a `Node` either take a type that enforces
 /// the size or return an error at runtime.
 ///
 /// [`nybbles_len`]: #method.nybbles_len
 #[derive(Copy, Clone, Debug, PartialEq, BytesCast, derive_more::From)]
 #[repr(transparent)]
 pub struct Node {
     data: NodeData,
 }
 
 /// The node value for NULL_REVISION
 pub const NULL_NODE: Node = Node {
     data: [0; NODE_BYTES_LENGTH],
 };
 
 /// Return an error if the slice has an unexpected length
 impl<'a> TryFrom<&'a [u8]> for &'a Node {
     type Error = ();
 
     #[inline]
     fn try_from(bytes: &'a [u8]) -> Result<Self, Self::Error> {
         match Node::from_bytes(bytes) {
             Ok((node, rest)) if rest.is_empty() => Ok(node),
             _ => Err(()),
         }
     }
 }
 
 /// Return an error if the slice has an unexpected length
 impl TryFrom<&'_ [u8]> for Node {
     type Error = std::array::TryFromSliceError;
 
     #[inline]
     fn try_from(bytes: &'_ [u8]) -> Result<Self, Self::Error> {
         let data = bytes.try_into()?;
         Ok(Self { data })
     }
 }
 
 impl From<&'_ NodeData> for Node {
     #[inline]
     fn from(data: &'_ NodeData) -> Self {
         Self { data: *data }
     }
 }
 
 impl fmt::LowerHex for Node {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         for &byte in &self.data {
             write!(f, "{:02x}", byte)?
         }
         Ok(())
     }
 }
 
 #[derive(Debug)]
 pub struct FromHexError;
 
 /// Low level utility function, also for prefixes
 fn get_nybble(s: &[u8], i: usize) -> u8 {
     if i % 2 == 0 {
         s[i / 2] >> 4
     } else {
         s[i / 2] & 0x0f
     }
 }
 
 impl Node {
     /// Retrieve the `i`th half-byte of the binary data.
     ///
     /// This is also the `i`th hexadecimal digit in numeric form,
     /// also called a [nybble](https://en.wikipedia.org/wiki/Nibble).
     pub fn get_nybble(&self, i: usize) -> u8 {
         get_nybble(&self.data, i)
     }
 
     /// Length of the data, in nybbles
     pub fn nybbles_len(&self) -> usize {
         // public exposure as an instance method only, so that we can
         // easily support several sizes of hashes if needed in the future.
         NODE_NYBBLES_LENGTH
     }
 
     /// Convert from hexadecimal string representation
     ///
     /// Exact length is required.
     ///
     /// To be used in FFI and I/O only, in order to facilitate future
     /// changes of hash format.
     pub fn from_hex(hex: impl AsRef<[u8]>) -> Result<Node, FromHexError> {
         let prefix = NodePrefix::from_hex(hex)?;
         if prefix.nybbles_len() == NODE_NYBBLES_LENGTH {
             Ok(Self { data: prefix.data })
         } else {
             Err(FromHexError)
         }
     }
 
     /// `from_hex`, but for input from an internal file of the repository such
     /// as a changelog or manifest entry.
     ///
     /// An error is treated as repository corruption.
     pub fn from_hex_for_repo(hex: impl AsRef<[u8]>) -> Result<Node, HgError> {
         Self::from_hex(hex.as_ref()).map_err(|FromHexError| {
             HgError::CorruptedRepository(format!(
                 "Expected a full hexadecimal node ID, found {}",
                 String::from_utf8_lossy(hex.as_ref())
             ))
         })
     }
 
     /// Provide access to binary data
     ///
     /// This is needed by FFI layers, for instance to return expected
     /// binary values to Python.
     pub fn as_bytes(&self) -> &[u8] {
         &self.data
     }
 
     pub fn short(&self) -> NodePrefix {
         NodePrefix {
             nybbles_len: SHORT_PREFIX_DEFAULT_NYBBLES_LENGTH,
             data: self.data,
         }
     }
+
+    pub fn pad_to_256_bits(&self) -> [u8; 32] {
+        let mut bits = [0; 32];
+        bits[..NODE_BYTES_LENGTH].copy_from_slice(&self.data);
+        bits
+    }
 }
 
 /// The beginning of a binary revision SHA.
 ///
 /// Since it can potentially come from an hexadecimal representation with
 /// odd length, it needs to carry around whether the last 4 bits are relevant
 /// or not.
 #[derive(Debug, PartialEq, Copy, Clone)]
 pub struct NodePrefix {
     /// In `1..=NODE_NYBBLES_LENGTH`
     nybbles_len: u8,
     /// The first `4 * length_in_nybbles` bits are used (considering bits
     /// within a bytes in big-endian: most significant first), the rest
     /// are zero.
     data: NodeData,
 }
 
 impl NodePrefix {
     /// Convert from hexadecimal string representation
     ///
     /// Similarly to `hex::decode`, can be used with Unicode string types
     /// (`String`, `&str`) as well as bytes.
     ///
     /// To be used in FFI and I/O only, in order to facilitate future
     /// changes of hash format.
     pub fn from_hex(hex: impl AsRef<[u8]>) -> Result<Self, FromHexError> {
         let hex = hex.as_ref();
         let len = hex.len();
         if len > NODE_NYBBLES_LENGTH || len == 0 {
             return Err(FromHexError);
         }
 
         let mut data = [0; NODE_BYTES_LENGTH];
         let mut nybbles_len = 0;
         for &ascii_byte in hex {
             let nybble = match char::from(ascii_byte).to_digit(16) {
                 Some(digit) => digit as u8,
                 None => return Err(FromHexError),
             };
             // Fill in the upper half of a byte first, then the lower half.
             let shift = if nybbles_len % 2 == 0 { 4 } else { 0 };
             data[nybbles_len as usize / 2] |= nybble << shift;
             nybbles_len += 1;
         }
         Ok(Self { data, nybbles_len })
     }
 
     pub fn nybbles_len(&self) -> usize {
         self.nybbles_len as _
     }
 
     pub fn is_prefix_of(&self, node: &Node) -> bool {
         let full_bytes = self.nybbles_len() / 2;
         if self.data[..full_bytes] != node.data[..full_bytes] {
             return false;
         }
         if self.nybbles_len() % 2 == 0 {
             return true;
         }
         let last = self.nybbles_len() - 1;
         self.get_nybble(last) == node.get_nybble(last)
     }
 
     /// Retrieve the `i`th half-byte from the prefix.
     ///
     /// This is also the `i`th hexadecimal digit in numeric form,
     /// also called a [nybble](https://en.wikipedia.org/wiki/Nibble).
     pub fn get_nybble(&self, i: usize) -> u8 {
         assert!(i < self.nybbles_len());
         get_nybble(&self.data, i)
     }
 
     fn iter_nybbles(&self) -> impl Iterator<Item = u8> + '_ {
         (0..self.nybbles_len()).map(move |i| get_nybble(&self.data, i))
     }
 
     /// Return the index first nybble that's different from `node`
     ///
     /// If the return value is `None` that means that `self` is
     /// a prefix of `node`, but the current method is a bit slower
     /// than `is_prefix_of`.
     ///
     /// Returned index is as in `get_nybble`, i.e., starting at 0.
     pub fn first_different_nybble(&self, node: &Node) -> Option<usize> {
         self.iter_nybbles()
             .zip(NodePrefix::from(*node).iter_nybbles())
             .position(|(a, b)| a != b)
     }
 }
 
 impl fmt::LowerHex for NodePrefix {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         let full_bytes = self.nybbles_len() / 2;
         for &byte in &self.data[..full_bytes] {
             write!(f, "{:02x}", byte)?
         }
         if self.nybbles_len() % 2 == 1 {
             let last = self.nybbles_len() - 1;
             write!(f, "{:x}", self.get_nybble(last))?
         }
         Ok(())
     }
 }
 
 /// A shortcut for full `Node` references
 impl From<&'_ Node> for NodePrefix {
     fn from(node: &'_ Node) -> Self {
         NodePrefix {
             nybbles_len: node.nybbles_len() as _,
             data: node.data,
         }
     }
 }
 
 /// A shortcut for full `Node` references
 impl From<Node> for NodePrefix {
     fn from(node: Node) -> Self {
         NodePrefix {
             nybbles_len: node.nybbles_len() as _,
             data: node.data,
         }
     }
 }
 
 impl PartialEq<Node> for NodePrefix {
     fn eq(&self, other: &Node) -> bool {
         Self::from(*other) == *self
     }
 }
 
 #[cfg(test)]
 mod tests {
     use super::*;
 
     const SAMPLE_NODE_HEX: &str = "0123456789abcdeffedcba9876543210deadbeef";
     const SAMPLE_NODE: Node = Node {
         data: [
             0x01, 0x23, 0x45, 0x67, 0x89, 0xab, 0xcd, 0xef, 0xfe, 0xdc, 0xba,
             0x98, 0x76, 0x54, 0x32, 0x10, 0xde, 0xad, 0xbe, 0xef,
         ],
     };
 
     /// Pad an hexadecimal string to reach `NODE_NYBBLES_LENGTH`
     /// The padding is made with zeros.
     pub fn hex_pad_right(hex: &str) -> String {
         let mut res = hex.to_string();
         while res.len() < NODE_NYBBLES_LENGTH {
             res.push('0');
         }
         res
     }
 
     #[test]
     fn test_node_from_hex() {
         let not_hex = "012... oops";
         let too_short = "0123";
         let too_long = format!("{}0", SAMPLE_NODE_HEX);
         assert_eq!(Node::from_hex(SAMPLE_NODE_HEX).unwrap(), SAMPLE_NODE);
         assert!(Node::from_hex(not_hex).is_err());
         assert!(Node::from_hex(too_short).is_err());
         assert!(Node::from_hex(&too_long).is_err());
     }
 
     #[test]
     fn test_node_encode_hex() {
         assert_eq!(format!("{:x}", SAMPLE_NODE), SAMPLE_NODE_HEX);
     }
 
     #[test]
     fn test_prefix_from_to_hex() -> Result<(), FromHexError> {
         assert_eq!(format!("{:x}", NodePrefix::from_hex("0e1")?), "0e1");
         assert_eq!(format!("{:x}", NodePrefix::from_hex("0e1a")?), "0e1a");
         assert_eq!(
             format!("{:x}", NodePrefix::from_hex(SAMPLE_NODE_HEX)?),
             SAMPLE_NODE_HEX
         );
         Ok(())
     }
 
     #[test]
     fn test_prefix_from_hex_errors() {
         assert!(NodePrefix::from_hex("testgr").is_err());
         let mut long = format!("{:x}", NULL_NODE);
         long.push('c');
         assert!(NodePrefix::from_hex(&long).is_err())
     }
 
     #[test]
     fn test_is_prefix_of() -> Result<(), FromHexError> {
         let mut node_data = [0; NODE_BYTES_LENGTH];
         node_data[0] = 0x12;
         node_data[1] = 0xca;
         let node = Node::from(node_data);
         assert!(NodePrefix::from_hex("12")?.is_prefix_of(&node));
         assert!(!NodePrefix::from_hex("1a")?.is_prefix_of(&node));
         assert!(NodePrefix::from_hex("12c")?.is_prefix_of(&node));
         assert!(!NodePrefix::from_hex("12d")?.is_prefix_of(&node));
         Ok(())
     }
 
     #[test]
     fn test_get_nybble() -> Result<(), FromHexError> {
         let prefix = NodePrefix::from_hex("dead6789cafe")?;
         assert_eq!(prefix.get_nybble(0), 13);
         assert_eq!(prefix.get_nybble(7), 9);
         Ok(())
     }
 
     #[test]
     fn test_first_different_nybble_even_prefix() {
         let prefix = NodePrefix::from_hex("12ca").unwrap();
         let mut node = Node::from([0; NODE_BYTES_LENGTH]);
         assert_eq!(prefix.first_different_nybble(&node), Some(0));
         node.data[0] = 0x13;
         assert_eq!(prefix.first_different_nybble(&node), Some(1));
         node.data[0] = 0x12;
         assert_eq!(prefix.first_different_nybble(&node), Some(2));
         node.data[1] = 0xca;
         // now it is a prefix
         assert_eq!(prefix.first_different_nybble(&node), None);
     }
 
     #[test]
     fn test_first_different_nybble_odd_prefix() {
         let prefix = NodePrefix::from_hex("12c").unwrap();
         let mut node = Node::from([0; NODE_BYTES_LENGTH]);
         assert_eq!(prefix.first_different_nybble(&node), Some(0));
         node.data[0] = 0x13;
         assert_eq!(prefix.first_different_nybble(&node), Some(1));
         node.data[0] = 0x12;
         assert_eq!(prefix.first_different_nybble(&node), Some(2));
         node.data[1] = 0xca;
         // now it is a prefix
         assert_eq!(prefix.first_different_nybble(&node), None);
     }
 }
 
 #[cfg(test)]
 pub use tests::hex_pad_right;

rust/hg-cpython/src/dirstate/dirstate_map.rs

@@ -1,499 +1,499 b''
 // dirstate_map.rs
 //
 // Copyright 2019 Raphaël Gomès <[e-mail]>
 //
 // This software may be used and distributed according to the terms of the
 // GNU General Public License version 2 or any later version.
 
 //! Bindings for the `hg::dirstate::dirstate_map` file provided by the
 //! `hg-core` package.
 
 use std::cell::{RefCell, RefMut};
 use std::convert::TryInto;
 
 use cpython::{
     exc, PyBool, PyBytes, PyClone, PyDict, PyErr, PyList, PyNone, PyObject,
     PyResult, Python, PythonObject, ToPyObject, UnsafePyLeaked,
 };
 
 use crate::{
     dirstate::copymap::{CopyMap, CopyMapItemsIterator, CopyMapKeysIterator},
     dirstate::item::DirstateItem,
     pybytes_deref::PyBytesDeref,
 };
 use hg::{
     dirstate::StateMapIter,
     dirstate_tree::dirstate_map::DirstateMap as TreeDirstateMap,
     dirstate_tree::on_disk::DirstateV2ParseError,
     dirstate_tree::owning::OwningDirstateMap,
     revlog::Node,
     utils::files::normalize_case,
     utils::hg_path::{HgPath, HgPathBuf},
     DirstateEntry, DirstateError, DirstateParents, EntryState,
 };
 
 // TODO
 //     This object needs to share references to multiple members of its Rust
 //     inner struct, namely `copy_map`, `dirs` and `all_dirs`.
 //     Right now `CopyMap` is done, but it needs to have an explicit reference
 //     to `RustDirstateMap` which itself needs to have an encapsulation for
 //     every method in `CopyMap` (copymapcopy, etc.).
 //     This is ugly and hard to maintain.
 //     The same logic applies to `dirs` and `all_dirs`, however the `Dirs`
 //     `py_class!` is already implemented and does not mention
 //     `RustDirstateMap`, rightfully so.
 //     All attributes also have to have a separate refcount data attribute for
 //     leaks, with all methods that go along for reference sharing.
 py_class!(pub class DirstateMap |py| {
     @shared data inner: OwningDirstateMap;
 
     /// Returns a `(dirstate_map, parents)` tuple
     @staticmethod
     def new_v1(
         on_disk: PyBytes,
     ) -> PyResult<PyObject> {
         let on_disk = PyBytesDeref::new(py, on_disk);
         let mut map = OwningDirstateMap::new_empty(on_disk);
         let (on_disk, map_placeholder) = map.get_pair_mut();
 
         let (actual_map, parents) = TreeDirstateMap::new_v1(on_disk)
             .map_err(|e| dirstate_error(py, e))?;
         *map_placeholder = actual_map;
         let map = Self::create_instance(py, map)?;
         let parents = parents.map(|p| {
             let p1 = PyBytes::new(py, p.p1.as_bytes());
             let p2 = PyBytes::new(py, p.p2.as_bytes());
             (p1, p2)
         });
         Ok((map, parents).to_py_object(py).into_object())
     }
 
     /// Returns a DirstateMap
     @staticmethod
     def new_v2(
         on_disk: PyBytes,
         data_size: usize,
         tree_metadata: PyBytes,
     ) -> PyResult<PyObject> {
         let dirstate_error = |e: DirstateError| {
             PyErr::new::<exc::OSError, _>(py, format!("Dirstate error: {:?}", e))
         };
         let on_disk = PyBytesDeref::new(py, on_disk);
         let mut map = OwningDirstateMap::new_empty(on_disk);
         let (on_disk, map_placeholder) = map.get_pair_mut();
         *map_placeholder = TreeDirstateMap::new_v2(
             on_disk, data_size, tree_metadata.data(py),
         ).map_err(dirstate_error)?;
         let map = Self::create_instance(py, map)?;
         Ok(map.into_object())
     }
 
     def clear(&self) -> PyResult<PyObject> {
         self.inner(py).borrow_mut().clear();
         Ok(py.None())
     }
 
     def get(
         &self,
         key: PyObject,
         default: Option<PyObject> = None
     ) -> PyResult<Option<PyObject>> {
         let key = key.extract::<PyBytes>(py)?;
         match self
             .inner(py)
             .borrow()
             .get(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))?
         {
             Some(entry) => {
                 Ok(Some(DirstateItem::new_as_pyobject(py, entry)?))
             },
             None => Ok(default)
         }
     }
 
     def set_dirstate_item(
         &self,
         path: PyObject,
         item: DirstateItem
     ) -> PyResult<PyObject> {
         let f = path.extract::<PyBytes>(py)?;
         let filename = HgPath::new(f.data(py));
         self.inner(py)
             .borrow_mut()
             .set_entry(filename, item.get_entry(py))
             .map_err(|e| v2_error(py, e))?;
         Ok(py.None())
     }
 
     def addfile(
         &self,
         f: PyBytes,
         item: DirstateItem,
     ) -> PyResult<PyNone> {
         let filename = HgPath::new(f.data(py));
         let entry = item.get_entry(py);
         self.inner(py)
             .borrow_mut()
             .add_file(filename, entry)
             .map_err(|e |dirstate_error(py, e))?;
         Ok(PyNone)
     }
 
     def removefile(
         &self,
         f: PyObject,
         in_merge: PyObject
     ) -> PyResult<PyObject> {
         self.inner(py).borrow_mut()
             .remove_file(
                 HgPath::new(f.extract::<PyBytes>(py)?.data(py)),
                 in_merge.extract::<PyBool>(py)?.is_true(),
             )
             .or_else(|_| {
                 Err(PyErr::new::<exc::OSError, _>(
                     py,
                     "Dirstate error".to_string(),
                 ))
             })?;
         Ok(py.None())
     }
 
     def drop_item_and_copy_source(
         &self,
         f: PyBytes,
     ) -> PyResult<PyNone> {
         self.inner(py)
             .borrow_mut()
             .drop_entry_and_copy_source(HgPath::new(f.data(py)))
             .map_err(|e |dirstate_error(py, e))?;
         Ok(PyNone)
     }
 
     def hastrackeddir(&self, d: PyObject) -> PyResult<PyBool> {
         let d = d.extract::<PyBytes>(py)?;
         Ok(self.inner(py).borrow_mut()
             .has_tracked_dir(HgPath::new(d.data(py)))
             .map_err(|e| {
                 PyErr::new::<exc::ValueError, _>(py, e.to_string())
             })?
             .to_py_object(py))
     }
 
     def hasdir(&self, d: PyObject) -> PyResult<PyBool> {
         let d = d.extract::<PyBytes>(py)?;
         Ok(self.inner(py).borrow_mut()
             .has_dir(HgPath::new(d.data(py)))
             .map_err(|e| {
                 PyErr::new::<exc::ValueError, _>(py, e.to_string())
             })?
             .to_py_object(py))
     }
 
     def write_v1(
         &self,
         p1: PyObject,
         p2: PyObject,
     ) -> PyResult<PyBytes> {
         let inner = self.inner(py).borrow();
         let parents = DirstateParents {
             p1: extract_node_id(py, &p1)?,
             p2: extract_node_id(py, &p2)?,
         };
         let result = inner.pack_v1(parents);
         match result {
             Ok(packed) => Ok(PyBytes::new(py, &packed)),
             Err(_) => Err(PyErr::new::<exc::OSError, _>(
                 py,
                 "Dirstate error".to_string(),
             )),
         }
     }
 
     /// Returns new data together with whether that data should be appended to
     /// the existing data file whose content is at `self.on_disk` (True),
     /// instead of written to a new data file (False).
     def write_v2(
         &self,
         can_append: bool,
     ) -> PyResult<PyObject> {
         let inner = self.inner(py).borrow();
         let result = inner.pack_v2(can_append);
         match result {
             Ok((packed, tree_metadata, append)) => {
                 let packed = PyBytes::new(py, &packed);
-                let tree_metadata = PyBytes::new(py, &tree_metadata);
+                let tree_metadata = PyBytes::new(py, tree_metadata.as_bytes());
                 let tuple = (packed, tree_metadata, append);
                 Ok(tuple.to_py_object(py).into_object())
             },
             Err(_) => Err(PyErr::new::<exc::OSError, _>(
                 py,
                 "Dirstate error".to_string(),
             )),
         }
     }
 
     def filefoldmapasdict(&self) -> PyResult<PyDict> {
         let dict = PyDict::new(py);
         for item in self.inner(py).borrow_mut().iter() {
             let (path, entry) = item.map_err(|e| v2_error(py, e))?;
             if entry.state() != EntryState::Removed {
                 let key = normalize_case(path);
                 let value = path;
                 dict.set_item(
                     py,
                     PyBytes::new(py, key.as_bytes()).into_object(),
                     PyBytes::new(py, value.as_bytes()).into_object(),
                 )?;
             }
         }
         Ok(dict)
     }
 
     def __len__(&self) -> PyResult<usize> {
         Ok(self.inner(py).borrow().len())
     }
 
     def __contains__(&self, key: PyObject) -> PyResult<bool> {
         let key = key.extract::<PyBytes>(py)?;
         self.inner(py)
             .borrow()
             .contains_key(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))
     }
 
     def __getitem__(&self, key: PyObject) -> PyResult<PyObject> {
         let key = key.extract::<PyBytes>(py)?;
         let key = HgPath::new(key.data(py));
         match self
             .inner(py)
             .borrow()
             .get(key)
             .map_err(|e| v2_error(py, e))?
         {
             Some(entry) => {
                 Ok(DirstateItem::new_as_pyobject(py, entry)?)
             },
             None => Err(PyErr::new::<exc::KeyError, _>(
                 py,
                 String::from_utf8_lossy(key.as_bytes()),
             )),
         }
     }
 
     def keys(&self) -> PyResult<DirstateMapKeysIterator> {
         let leaked_ref = self.inner(py).leak_immutable();
         DirstateMapKeysIterator::from_inner(
             py,
             unsafe { leaked_ref.map(py, |o| o.iter()) },
         )
     }
 
     def items(&self) -> PyResult<DirstateMapItemsIterator> {
         let leaked_ref = self.inner(py).leak_immutable();
         DirstateMapItemsIterator::from_inner(
             py,
             unsafe { leaked_ref.map(py, |o| o.iter()) },
         )
     }
 
     def __iter__(&self) -> PyResult<DirstateMapKeysIterator> {
         let leaked_ref = self.inner(py).leak_immutable();
         DirstateMapKeysIterator::from_inner(
             py,
             unsafe { leaked_ref.map(py, |o| o.iter()) },
         )
     }
 
     // TODO all copymap* methods, see docstring above
     def copymapcopy(&self) -> PyResult<PyDict> {
         let dict = PyDict::new(py);
         for item in self.inner(py).borrow().copy_map_iter() {
             let (key, value) = item.map_err(|e| v2_error(py, e))?;
             dict.set_item(
                 py,
                 PyBytes::new(py, key.as_bytes()),
                 PyBytes::new(py, value.as_bytes()),
             )?;
         }
         Ok(dict)
     }
 
     def copymapgetitem(&self, key: PyObject) -> PyResult<PyBytes> {
         let key = key.extract::<PyBytes>(py)?;
         match self
             .inner(py)
             .borrow()
             .copy_map_get(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))?
         {
             Some(copy) => Ok(PyBytes::new(py, copy.as_bytes())),
             None => Err(PyErr::new::<exc::KeyError, _>(
                 py,
                 String::from_utf8_lossy(key.data(py)),
             )),
         }
     }
     def copymap(&self) -> PyResult<CopyMap> {
         CopyMap::from_inner(py, self.clone_ref(py))
     }
 
     def copymaplen(&self) -> PyResult<usize> {
         Ok(self.inner(py).borrow().copy_map_len())
     }
     def copymapcontains(&self, key: PyObject) -> PyResult<bool> {
         let key = key.extract::<PyBytes>(py)?;
         self.inner(py)
             .borrow()
             .copy_map_contains_key(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))
     }
     def copymapget(
         &self,
         key: PyObject,
         default: Option<PyObject>
     ) -> PyResult<Option<PyObject>> {
         let key = key.extract::<PyBytes>(py)?;
         match self
             .inner(py)
             .borrow()
             .copy_map_get(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))?
         {
             Some(copy) => Ok(Some(
                 PyBytes::new(py, copy.as_bytes()).into_object(),
             )),
             None => Ok(default),
         }
     }
     def copymapsetitem(
         &self,
         key: PyObject,
         value: PyObject
     ) -> PyResult<PyObject> {
         let key = key.extract::<PyBytes>(py)?;
         let value = value.extract::<PyBytes>(py)?;
         self.inner(py)
             .borrow_mut()
             .copy_map_insert(
                 HgPathBuf::from_bytes(key.data(py)),
                 HgPathBuf::from_bytes(value.data(py)),
             )
             .map_err(|e| v2_error(py, e))?;
         Ok(py.None())
     }
     def copymappop(
         &self,
         key: PyObject,
         default: Option<PyObject>
     ) -> PyResult<Option<PyObject>> {
         let key = key.extract::<PyBytes>(py)?;
         match self
             .inner(py)
             .borrow_mut()
             .copy_map_remove(HgPath::new(key.data(py)))
             .map_err(|e| v2_error(py, e))?
         {
             Some(copy) => Ok(Some(
                 PyBytes::new(py, copy.as_bytes()).into_object(),
             )),
             None => Ok(default),
         }
     }
 
     def copymapiter(&self) -> PyResult<CopyMapKeysIterator> {
         let leaked_ref = self.inner(py).leak_immutable();
         CopyMapKeysIterator::from_inner(
             py,
             unsafe { leaked_ref.map(py, |o| o.copy_map_iter()) },
         )
     }
 
     def copymapitemsiter(&self) -> PyResult<CopyMapItemsIterator> {
         let leaked_ref = self.inner(py).leak_immutable();
         CopyMapItemsIterator::from_inner(
             py,
             unsafe { leaked_ref.map(py, |o| o.copy_map_iter()) },
         )
     }
 
     def tracked_dirs(&self) -> PyResult<PyList> {
         let dirs = PyList::new(py, &[]);
         for path in self.inner(py).borrow_mut().iter_tracked_dirs()
             .map_err(|e |dirstate_error(py, e))?
         {
             let path = path.map_err(|e| v2_error(py, e))?;
             let path = PyBytes::new(py, path.as_bytes());
             dirs.append(py, path.into_object())
         }
         Ok(dirs)
     }
 
     def debug_iter(&self, all: bool) -> PyResult<PyList> {
         let dirs = PyList::new(py, &[]);
         for item in self.inner(py).borrow().debug_iter(all) {
             let (path, (state, mode, size, mtime)) =
                 item.map_err(|e| v2_error(py, e))?;
             let path = PyBytes::new(py, path.as_bytes());
             let item = (path, state, mode, size, mtime);
             dirs.append(py, item.to_py_object(py).into_object())
         }
         Ok(dirs)
     }
 });
 
 impl DirstateMap {
     pub fn get_inner_mut<'a>(
         &'a self,
         py: Python<'a>,
     ) -> RefMut<'a, OwningDirstateMap> {
         self.inner(py).borrow_mut()
     }
     fn translate_key(
         py: Python,
         res: Result<(&HgPath, DirstateEntry), DirstateV2ParseError>,
     ) -> PyResult<Option<PyBytes>> {
         let (f, _entry) = res.map_err(|e| v2_error(py, e))?;
         Ok(Some(PyBytes::new(py, f.as_bytes())))
     }
     fn translate_key_value(
         py: Python,
         res: Result<(&HgPath, DirstateEntry), DirstateV2ParseError>,
     ) -> PyResult<Option<(PyBytes, PyObject)>> {
         let (f, entry) = res.map_err(|e| v2_error(py, e))?;
         Ok(Some((
             PyBytes::new(py, f.as_bytes()),
             DirstateItem::new_as_pyobject(py, entry)?,
         )))
     }
 }
 
 py_shared_iterator!(
     DirstateMapKeysIterator,
     UnsafePyLeaked<StateMapIter<'static>>,
     DirstateMap::translate_key,
     Option<PyBytes>
 );
 
 py_shared_iterator!(
     DirstateMapItemsIterator,
     UnsafePyLeaked<StateMapIter<'static>>,
     DirstateMap::translate_key_value,
     Option<(PyBytes, PyObject)>
 );
 
 fn extract_node_id(py: Python, obj: &PyObject) -> PyResult<Node> {
     let bytes = obj.extract::<PyBytes>(py)?;
     match bytes.data(py).try_into() {
         Ok(s) => Ok(s),
         Err(e) => Err(PyErr::new::<exc::ValueError, _>(py, e.to_string())),
     }
 }
 
 pub(super) fn v2_error(py: Python<'_>, _: DirstateV2ParseError) -> PyErr {
     PyErr::new::<exc::ValueError, _>(py, "corrupted dirstate-v2")
 }
 
 fn dirstate_error(py: Python<'_>, e: DirstateError) -> PyErr {
     PyErr::new::<exc::OSError, _>(py, format!("Dirstate error: {:?}", e))
 }