Merge branch 'gix-status'

gitoxide-core/src/query/engine/update.rs

@@ -9,8 +9,8 @@ use anyhow::{anyhow, bail};
 use gix::objs::find::Error;
 use gix::{
     bstr::{BStr, BString, ByteSlice},
+    diff::rewrites::CopySource,
     features::progress,
-    object::tree::diff::rewrites::CopySource,
     parallel::{InOrderIter, SequenceId},
     prelude::ObjectIdExt,
     Count, Progress,
@@ -139,11 +139,10 @@ pub fn update(
         });
 
         let rewrites = {
-            let mut r =
-                gix::object::tree::diff::Rewrites::try_from_config(&repo.config_snapshot(), true)?.unwrap_or_default();
-            r.copies = Some(gix::object::tree::diff::rewrites::Copies {
+            let mut r = gix::diff::new_rewrites(&repo.config_snapshot(), true)?.unwrap_or_default();
+            r.copies = Some(gix::diff::rewrites::Copies {
                 source: if find_copies_harder {
-                    CopySource::FromSetOfModifiedFilesAndSourceTree
+                    CopySource::FromSetOfModifiedFilesAndAllSources
                 } else {
                     CopySource::FromSetOfModifiedFiles
                 },

gix-diff/Cargo.toml

@@ -12,7 +12,7 @@ autotests = false
 
 [features]
 default = ["blob"]
-## Enable diffing of blobs using imara-diff.
+## Enable diffing of blobs using imara-diff, which also allows for a generic rewrite tracking implementation.
 blob = ["dep:imara-diff"]
 ## Data structures implement `serde::Serialize` and `serde::Deserialize`.
 serde = ["dep:serde", "gix-hash/serde", "gix-object/serde"]
@@ -25,10 +25,12 @@ doctest = false
 [dependencies]
 gix-hash = { version = "^0.13.1", path = "../gix-hash" }
 gix-object = { version = "^0.38.0", path = "../gix-object" }
+
 thiserror = "1.0.32"
 imara-diff = { version = "0.1.3", optional = true }
 serde = { version = "1.0.114", optional = true, default-features = false, features = ["derive"]}
 getrandom = { version = "0.2.8", optional = true, default-features = false, features = ["js"] }
+bstr = { version = "1.5.0", default-features = false }
 
 document-features = { version = "0.2.0", optional = true }
 

gix-diff/src/blob.rs

@@ -1,3 +1,18 @@
 //! For using text diffs, please have a look at the [`imara-diff` documentation](https://docs.rs/imara-diff),
 //! maintained by [Pascal Kuthe](https://github.com/pascalkuthe).
+//!
+//!
+/// Information about the diff performed to detect similarity.
+#[derive(Debug, Default, Clone, Copy, Eq, PartialEq)]
+pub struct DiffLineStats {
+    /// The amount of lines to remove from the source to get to the destination.
+    pub removals: u32,
+    /// The amount of lines to add to the source to get to the destination.
+    pub insertions: u32,
+    /// The amount of lines of the previous state, in the source.
+    pub before: u32,
+    /// The amount of lines of the new state, in the destination.
+    pub after: u32,
+}
+
 pub use imara_diff::*;

gix-diff/src/lib.rs

@@ -8,6 +8,34 @@ cfg_attr(doc, doc = ::document_features::document_features!())
 #![deny(missing_docs, rust_2018_idioms)]
 #![forbid(unsafe_code)]
 
+/// A structure to capture how to perform rename and copy tracking, used by the [rewrites::Tracker].
+#[derive(Debug, Copy, Clone, PartialEq)]
+#[cfg(feature = "blob")]
+pub struct Rewrites {
+    /// If `Some(…)`, also find copies. `None` is the default which does not try to detect copies at all.
+    ///
+    /// Note that this is an even more expensive operation than detecting renames stemming from additions and deletions
+    /// as the resulting set to search through is usually larger.
+    pub copies: Option<rewrites::Copies>,
+    /// The percentage of similarity needed for files to be considered renamed, defaulting to `Some(0.5)`.
+    /// This field is similar to `git diff -M50%`.
+    ///
+    /// If `None`, files are only considered equal if their content matches 100%.
+    /// Note that values greater than 1.0 have no different effect than 1.0.
+    pub percentage: Option<f32>,
+    /// The amount of files to consider for fuzzy rename or copy tracking. Defaults to 1000, meaning that only 1000*1000
+    /// combinations can be tested for fuzzy matches, i.e. the ones that try to find matches by comparing similarity.
+    /// If 0, there is no limit.
+    ///
+    /// If the limit would not be enough to test the entire set of combinations, the algorithm will trade in precision and not
+    /// run the fuzzy version of identity tests at all. That way results are never partial.
+    pub limit: usize,
+}
+
+/// Contains a [Tracker](rewrites::Tracker) to detect rewrites.
+#[cfg(feature = "blob")]
+pub mod rewrites;
+
 ///
 pub mod tree;
 

gix-diff/src/rewrites/mod.rs

@@ -0,0 +1,77 @@
+use crate::Rewrites;
+
+/// Types related to the rename tracker for renames, rewrites and copies.
+pub mod tracker;
+
+/// A type to retain state related to an ongoing tracking operation to retain sets of interesting changes
+/// of which some are retained to at a later stage compute the ones that seem to be renames or copies.
+pub struct Tracker<T> {
+    /// The tracked items thus far, which will be used to determine renames/copies and rewrites later.
+    items: Vec<tracker::Item<T>>,
+    /// A place to store all paths in to reduce amount of allocations.
+    path_backing: Vec<u8>,
+    /// A buffer for use when fetching objects for similarity tests.
+    buf1: Vec<u8>,
+    /// Another buffer for use when fetching objects for similarity tests.
+    buf2: Vec<u8>,
+    /// How to track copies and/or rewrites.
+    rewrites: Rewrites,
+    /// The diff algorithm to use when checking for similarity.
+    diff_algo: crate::blob::Algorithm,
+}
+
+/// Determine in which set of files to search for copies.
+#[derive(Default, Debug, Copy, Clone, Eq, PartialEq)]
+pub enum CopySource {
+    /// Find copies from the set of modified files only.
+    #[default]
+    FromSetOfModifiedFiles,
+    /// Find copies from the set of modified files, as well as all files known to the source (i.e. previous state of the tree).
+    ///
+    /// This can be an expensive operation as it scales exponentially with the total amount of files in the set.
+    FromSetOfModifiedFilesAndAllSources,
+}
+
+/// Under which circumstances we consider a file to be a copy.
+#[derive(Debug, Copy, Clone, PartialEq)]
+pub struct Copies {
+    /// The set of files to search when finding the source of copies.
+    pub source: CopySource,
+    /// Equivalent to [`Rewrites::percentage`], but used for copy tracking.
+    ///
+    /// Useful to have similarity-based rename tracking and cheaper copy tracking.
+    pub percentage: Option<f32>,
+}
+
+impl Default for Copies {
+    fn default() -> Self {
+        Copies {
+            source: CopySource::default(),
+            percentage: Some(0.5),
+        }
+    }
+}
+
+/// Information collected while handling rewrites of files which may be tracked.
+#[derive(Default, Clone, Copy, Debug, PartialEq)]
+pub struct Outcome {
+    /// The options used to guide the rewrite tracking. Either fully provided by the caller or retrieved from git configuration.
+    pub options: Rewrites,
+    /// The amount of similarity checks that have been conducted to find renamed files and potentially copies.
+    pub num_similarity_checks: usize,
+    /// Set to the amount of worst-case rename permutations we didn't search as our limit didn't allow it.
+    pub num_similarity_checks_skipped_for_rename_tracking_due_to_limit: usize,
+    /// Set to the amount of worst-case copy permutations we didn't search as our limit didn't allow it.
+    pub num_similarity_checks_skipped_for_copy_tracking_due_to_limit: usize,
+}
+
+/// The default settings for rewrites according to the git configuration defaults.
+impl Default for Rewrites {
+    fn default() -> Self {
+        Rewrites {
+            copies: None,
+            percentage: Some(0.5),
+            limit: 1000,
+        }
+    }
+}