Merge pull request #1837 from GitoxideLabs/improvements

improvements

Author: Byron

gix-index/src/access/mod.rs

@@ -518,7 +518,8 @@ impl State {
 
     /// Physically remove all entries for which `should_remove(idx, path, entry)` returns `true`, traversing them from first to last.
     ///
-    /// Note that the memory used for the removed entries paths is not freed, as it's append-only.
+    /// Note that the memory used for the removed entries paths is not freed, as it's append-only, and
+    /// that some extensions might refer to paths which are now deleted.
     ///
     /// ### Performance
     ///
@@ -534,6 +535,16 @@ impl State {
             res
         });
     }
+
+    /// Physically remove the entry at `index`, or panic if the entry didn't exist.
+    ///
+    /// This call is typically made after looking up `index`, so it's clear that it will not panic.
+    ///
+    /// Note that the memory used for the removed entries paths is not freed, as it's append-only, and
+    /// that some extensions might refer to paths which are now deleted.
+    pub fn remove_entry_at_index(&mut self, index: usize) -> Entry {
+        self.entries.remove(index)
+    }
 }
 
 /// Extensions
@@ -542,6 +553,10 @@ impl State {
     pub fn tree(&self) -> Option<&extension::Tree> {
         self.tree.as_ref()
     }
+    /// Remove the `tree` extension.
+    pub fn remove_tree(&mut self) -> Option<extension::Tree> {
+        self.tree.take()
+    }
     /// Access the `link` extension.
     pub fn link(&self) -> Option<&extension::Link> {
         self.link.as_ref()
@@ -550,6 +565,10 @@ impl State {
     pub fn resolve_undo(&self) -> Option<&extension::resolve_undo::Paths> {
         self.resolve_undo.as_ref()
     }
+    /// Remove the resolve-undo extension.
+    pub fn remove_resolve_undo(&mut self) -> Option<extension::resolve_undo::Paths> {
+        self.resolve_undo.take()
+    }
     /// Obtain the untracked extension.
     pub fn untracked(&self) -> Option<&extension::UntrackedCache> {
         self.untracked.as_ref()

gix-index/src/entry/mod.rs

@@ -107,9 +107,21 @@ mod access {
 mod _impls {
     use std::cmp::Ordering;
 
+    use crate::{entry, Entry, State};
     use bstr::BStr;
-
-    use crate::{Entry, State};
+    use gix_object::tree::EntryKind;
+
+    impl From<EntryKind> for entry::Mode {
+        fn from(value: EntryKind) -> Self {
+            match value {
+                EntryKind::Tree => entry::Mode::DIR,
+                EntryKind::Blob => entry::Mode::FILE,
+                EntryKind::BlobExecutable => entry::Mode::FILE_EXECUTABLE,
+                EntryKind::Link => entry::Mode::SYMLINK,
+                EntryKind::Commit => entry::Mode::COMMIT,
+            }
+        }
+    }
 
     impl Entry {
         /// Compare one entry to another by their path, by comparing only their common path portion byte by byte, then resorting to

gix-index/tests/index/access.rs

@@ -214,6 +214,18 @@ fn remove_entries() {
     file.remove_entries(|_, _, _| unreachable!("should not be called"));
 }
 
+#[test]
+fn remove_entry_at_index() {
+    let mut file = Fixture::Loose("conflicting-file").open();
+
+    file.remove_entry_at_index(0);
+    assert_eq!(file.entries().len(), 2);
+    file.remove_entry_at_index(0);
+    assert_eq!(file.entries().len(), 1);
+    file.remove_entry_at_index(0);
+    assert_eq!(file.entries().len(), 0);
+}
+
 #[test]
 fn sort_entries() {
     let mut file = Fixture::Generated("v4_more_files_IEOT").open();

gix-object/src/blob.rs

@@ -44,4 +44,9 @@ impl BlobRef<'_> {
     pub fn from_bytes(data: &[u8]) -> Result<BlobRef<'_>, Infallible> {
         Ok(BlobRef { data })
     }
+
+    /// Clone the data in this instance by allocating a new vector for a fully owned blob.
+    pub fn into_owned(self) -> Blob {
+        self.into()
+    }
 }

gix-object/src/commit/mod.rs

@@ -5,6 +5,9 @@ use winnow::prelude::*;
 
 use crate::{Commit, CommitRef, TagRef};
 
+/// The well-known field name for gpg signatures.
+pub const SIGNATURE_FIELD_NAME: &str = "gpgsig";
+
 mod decode;
 ///
 pub mod message;
@@ -84,7 +87,7 @@ impl<'a> CommitRef<'a> {
     }
 
     /// Returns a convenient iterator over all extra headers.
-    pub fn extra_headers(&self) -> crate::commit::ExtraHeaders<impl Iterator<Item = (&BStr, &BStr)>> {
+    pub fn extra_headers(&self) -> ExtraHeaders<impl Iterator<Item = (&BStr, &BStr)>> {
         ExtraHeaders::new(self.extra_headers.iter().map(|(k, v)| (*k, v.as_ref())))
     }
 
@@ -113,6 +116,19 @@ impl<'a> CommitRef<'a> {
     }
 }
 
+/// Conversion
+impl CommitRef<'_> {
+    /// Copy all fields of this instance into a fully owned commit, consuming this instance.
+    pub fn into_owned(self) -> Commit {
+        self.into()
+    }
+
+    /// Copy all fields of this instance into a fully owned commit, internally cloning this instance.
+    pub fn to_owned(self) -> Commit {
+        self.clone().into()
+    }
+}
+
 impl Commit {
     /// Returns a convenient iterator over all extra headers.
     pub fn extra_headers(&self) -> ExtraHeaders<impl Iterator<Item = (&BStr, &BStr)>> {
@@ -134,16 +150,26 @@ where
     pub fn new(iter: I) -> Self {
         ExtraHeaders { inner: iter }
     }
+
     /// Find the _value_ of the _first_ header with the given `name`.
     pub fn find(mut self, name: &str) -> Option<&'a BStr> {
         self.inner
             .find_map(move |(k, v)| if k == name.as_bytes().as_bstr() { Some(v) } else { None })
     }
+
+    /// Find the entry index with the given name, or return `None` if unavailable.
+    pub fn find_pos(self, name: &str) -> Option<usize> {
+        self.inner
+            .enumerate()
+            .find_map(|(pos, (field, _value))| (field == name).then_some(pos))
+    }
+
     /// Return an iterator over all _values_ of headers with the given `name`.
     pub fn find_all(self, name: &'a str) -> impl Iterator<Item = &'a BStr> {
         self.inner
             .filter_map(move |(k, v)| if k == name.as_bytes().as_bstr() { Some(v) } else { None })
     }
+
     /// Return an iterator over all git mergetags.
     ///
     /// A merge tag is a tag object embedded within the respective header field of a commit, making
@@ -154,6 +180,6 @@ where
 
     /// Return the cryptographic signature provided by gpg/pgp verbatim.
     pub fn pgp_signature(self) -> Option<&'a BStr> {
-        self.find("gpgsig")
+        self.find(SIGNATURE_FIELD_NAME)
     }
 }

gix-object/src/lib.rs

@@ -58,23 +58,23 @@ pub enum Kind {
     Commit,
     Tag,
 }
-/// A chunk of any [`data`][BlobRef::data].
+/// A chunk of any [`data`](BlobRef::data).
 #[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone, Copy)]
 #[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
 pub struct BlobRef<'a> {
     /// The bytes themselves.
     pub data: &'a [u8],
 }
 
-/// A mutable chunk of any [`data`][Blob::data].
+/// A mutable chunk of any [`data`](Blob::data).
 #[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone)]
 #[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
 pub struct Blob {
     /// The data itself.
     pub data: Vec<u8>,
 }
 
-/// A git commit parsed using [`from_bytes()`][CommitRef::from_bytes()].
+/// A git commit parsed using [`from_bytes()`](CommitRef::from_bytes()).
 ///
 /// A commit encapsulates information about a point in time at which the state of the repository is recorded, usually after a
 /// change which is documented in the commit `message`.
@@ -83,18 +83,18 @@ pub struct Blob {
 pub struct CommitRef<'a> {
     /// HEX hash of tree object we point to. Usually 40 bytes long.
     ///
-    /// Use [`tree()`][CommitRef::tree()] to obtain a decoded version of it.
+    /// Use [`tree()`](CommitRef::tree()) to obtain a decoded version of it.
     #[cfg_attr(feature = "serde", serde(borrow))]
     pub tree: &'a BStr,
     /// HEX hash of each parent commit. Empty for first commit in repository.
     pub parents: SmallVec<[&'a BStr; 1]>,
     /// Who wrote this commit. Name and email might contain whitespace and are not trimmed to ensure round-tripping.
     ///
-    /// Use the [`author()`][CommitRef::author()] method to received a trimmed version of it.
+    /// Use the [`author()`](CommitRef::author()) method to received a trimmed version of it.
     pub author: gix_actor::SignatureRef<'a>,
     /// Who committed this commit. Name and email might contain whitespace and are not trimmed to ensure round-tripping.
     ///
-    /// Use the [`committer()`][CommitRef::committer()] method to received a trimmed version of it.
+    /// Use the [`committer()`](CommitRef::committer()) method to received a trimmed version of it.
     ///
     /// This may be different from the `author` in case the author couldn't write to the repository themselves and
     /// is commonly encountered with contributed commits.
@@ -103,7 +103,7 @@ pub struct CommitRef<'a> {
     pub encoding: Option<&'a BStr>,
     /// The commit message documenting the change.
     pub message: &'a BStr,
-    /// Extra header fields, in order of them being encountered, made accessible with the iterator returned by [`extra_headers()`][CommitRef::extra_headers()].
+    /// Extra header fields, in order of them being encountered, made accessible with the iterator returned by [`extra_headers()`](CommitRef::extra_headers()).
     pub extra_headers: Vec<(&'a BStr, Cow<'a, BStr>)>,
 }
 
@@ -135,15 +135,15 @@ pub struct Commit {
     /// The commit message documenting the change.
     pub message: BString,
     /// Extra header fields, in order of them being encountered, made accessible with the iterator returned
-    /// by [`extra_headers()`][Commit::extra_headers()].
+    /// by [`extra_headers()`](Commit::extra_headers()).
     pub extra_headers: Vec<(BString, BString)>,
 }
 
 /// Represents a git tag, commonly indicating a software release.
-#[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone)]
+#[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone, Copy)]
 #[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
 pub struct TagRef<'a> {
-    /// The hash in hexadecimal being the object this tag points to. Use [`target()`][TagRef::target()] to obtain a byte representation.
+    /// The hash in hexadecimal being the object this tag points to. Use [`target()`](TagRef::target()) to obtain a byte representation.
     #[cfg_attr(feature = "serde", serde(borrow))]
     pub target: &'a BStr,
     /// The kind of object that `target` points to.
@@ -184,13 +184,13 @@ pub struct Tag {
     pub pgp_signature: Option<BString>,
 }
 
-/// Immutable objects are read-only structures referencing most data from [a byte slice][crate::ObjectRef::from_bytes()].
+/// Immutable objects are read-only structures referencing most data from [a byte slice](ObjectRef::from_bytes()).
 ///
 /// Immutable objects are expected to be deserialized from bytes that acts as backing store, and they
-/// cannot be mutated or serialized. Instead, one will [convert][crate::ObjectRef::into_owned()] them into their [`mutable`][Object] counterparts
+/// cannot be mutated or serialized. Instead, one will [convert](ObjectRef::into_owned()) them into their [`mutable`](Object) counterparts
 /// which support mutation and serialization.
 ///
-/// An `ObjectRef` is representing [`Trees`][TreeRef], [`Blobs`][BlobRef], [`Commits`][CommitRef], or [`Tags`][TagRef].
+/// An `ObjectRef` is representing [`Trees`](TreeRef), [`Blobs`](BlobRef), [`Commits`](CommitRef), or [`Tags`](TagRef).
 #[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone)]
 #[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
 #[allow(missing_docs)]
@@ -206,10 +206,10 @@ pub enum ObjectRef<'a> {
 ///
 /// Mutable objects are Commits, Trees, Blobs and Tags that can be changed and serialized.
 ///
-/// They either created using object [construction][Object] or by [deserializing existing objects][ObjectRef::from_bytes()]
-/// and converting these [into mutable copies][ObjectRef::into_owned()] for adjustments.
+/// They either created using object [construction](Object) or by [deserializing existing objects](ObjectRef::from_bytes())
+/// and converting these [into mutable copies](ObjectRef::into_owned()) for adjustments.
 ///
-/// An `Object` is representing [`Trees`][Tree], [`Blobs`][Blob], [`Commits`][Commit] or [`Tags`][Tag].
+/// An `Object` is representing [`Trees`](Tree), [`Blobs`](Blob), [`Commits`](Commit), or [`Tags`](Tag).
 #[derive(PartialEq, Eq, Debug, Hash, Ord, PartialOrd, Clone)]
 #[cfg_attr(feature = "serde", derive(serde::Serialize, serde::Deserialize))]
 #[allow(clippy::large_enum_variant, missing_docs)]

gix-object/src/tag/mod.rs

@@ -23,4 +23,9 @@ impl<'a> TagRef<'a> {
     pub fn target(&self) -> gix_hash::ObjectId {
         gix_hash::ObjectId::from_hex(self.target).expect("prior validation")
     }
+
+    /// Copy all data into a fully-owned instance.
+    pub fn into_owned(self) -> crate::Tag {
+        self.into()
+    }
 }

gix-object/tests/object/commit/from_bytes.rs

@@ -1,12 +1,11 @@
-use gix_actor::SignatureRef;
-use gix_date::{time::Sign, Time};
-use gix_object::{bstr::ByteSlice, commit::message::body::TrailerRef, CommitRef};
-use smallvec::SmallVec;
-
 use crate::{
     commit::{LONG_MESSAGE, MERGE_TAG, SIGNATURE},
     fixture_name, linus_signature, signature,
 };
+use gix_actor::SignatureRef;
+use gix_date::{time::Sign, Time};
+use gix_object::{bstr::ByteSlice, commit::message::body::TrailerRef, CommitRef};
+use smallvec::SmallVec;
 
 #[test]
 fn invalid_timestsamp() {
@@ -354,7 +353,12 @@ fn newline_right_after_signature_multiline_header() -> crate::Result {
     let pgp_sig = crate::commit::OTHER_SIGNATURE.as_bstr();
     assert_eq!(commit.extra_headers[0].1.as_ref(), pgp_sig);
     assert_eq!(commit.extra_headers().pgp_signature(), Some(pgp_sig));
-    assert_eq!(commit.extra_headers().find("gpgsig"), Some(pgp_sig));
+    assert_eq!(
+        commit.extra_headers().find(gix_object::commit::SIGNATURE_FIELD_NAME),
+        Some(pgp_sig)
+    );
+    assert_eq!(commit.extra_headers().find_pos("gpgsig"), Some(0));
+    assert_eq!(commit.extra_headers().find_pos("something else"), None);
     assert!(commit.message.starts_with(b"Rollup"));
     Ok(())
 }

gix-object/tests/object/tag/mod.rs

@@ -4,6 +4,7 @@ use gix_object::{bstr::ByteSlice, Kind, TagRef, TagRefIter};
 use crate::fixture_name;
 
 mod method {
+    use bstr::ByteSlice;
     use gix_object::TagRef;
     use pretty_assertions::assert_eq;
 
@@ -15,6 +16,21 @@ mod method {
         let tag = TagRef::from_bytes(&fixture)?;
         assert_eq!(tag.target(), hex_to_id("ffa700b4aca13b80cb6b98a078e7c96804f8e0ec"));
         assert_eq!(tag.target, "ffa700b4aca13b80cb6b98a078e7c96804f8e0ec".as_bytes());
+
+        let gix_object::Tag {
+            target,
+            target_kind,
+            name,
+            tagger,
+            message,
+            pgp_signature,
+        } = tag.into_owned();
+        assert_eq!(target.to_string(), tag.target);
+        assert_eq!(target_kind, tag.target_kind);
+        assert_eq!(name, tag.name);
+        assert_eq!(tagger.as_ref().map(|s| s.to_ref()), tag.tagger);
+        assert_eq!(message, tag.message);
+        assert_eq!(pgp_signature.as_ref().map(|s| s.as_bstr()), tag.pgp_signature);
         Ok(())
     }
 }

gix/src/repository/object.rs

@@ -342,7 +342,7 @@ impl crate::Repository {
     /// Create a new commit object with `message` referring to `tree` with `parents`, and point `reference`
     /// to it. The commit is written without message encoding field, which can be assumed to be UTF-8.
     /// `author` and `committer` fields are pre-set from the configuration, which can be altered
-    /// [temporarily][crate::Repository::config_snapshot_mut()] before the call if required.
+    /// [temporarily](crate::Repository::config_snapshot_mut()) before the call if required.
     ///
     /// `reference` will be created if it doesn't exist, and can be `"HEAD"` to automatically write-through to the symbolic reference
     /// that `HEAD` points to if it is not detached. For this reason, detached head states cannot be created unless the `HEAD` is detached
@@ -352,6 +352,11 @@ impl crate::Repository {
     /// If there is no parent, the `reference` is expected to not exist yet.
     ///
     /// The method fails immediately if a `reference` lock can't be acquired.
+    ///
+    /// ### Writing a commit without `reference` update
+    ///
+    /// If the reference shouldn't be updated, use [`Self::write_object()`] along with a newly created [`crate::objs::Object`] whose fields
+    /// can be fully defined.
     pub fn commit<Name, E>(
         &self,
         reference: Name,