gix_diff/blob/
mod.rs

1//! For using text diffs, please have a look at the [`imara-diff` documentation](https://docs.rs/imara-diff),
2//! maintained by [Pascal Kuthe](https://github.com/pascalkuthe).
3use std::{collections::HashMap, path::PathBuf};
4
5use bstr::BString;
6pub use imara_diff::*;
7
8///
9pub mod pipeline;
10
11///
12pub mod platform;
13
14pub mod unified_diff;
15pub use unified_diff::_impl::UnifiedDiff;
16
17/// Information about the diff performed to detect similarity.
18#[derive(Debug, Default, Clone, Copy, PartialEq, PartialOrd)]
19pub struct DiffLineStats {
20    /// The amount of lines to remove from the source to get to the destination.
21    pub removals: u32,
22    /// The amount of lines to add to the source to get to the destination.
23    pub insertions: u32,
24    /// The amount of lines of the previous state, in the source.
25    pub before: u32,
26    /// The amount of lines of the new state, in the destination.
27    pub after: u32,
28    /// A range from 0 to 1.0, where 1.0 is a perfect match and 0.5 is a similarity of 50%.
29    /// Similarity is the ratio between all lines in the previous blob and the current blob,
30    /// calculated as `(old_lines_count - new_lines_count) as f32 / old_lines_count.max(new_lines_count) as f32`.
31    pub similarity: f32,
32}
33
34/// A way to classify a resource suitable for diffing.
35#[derive(Copy, Clone, Debug, Ord, PartialOrd, Eq, PartialEq, Hash)]
36pub enum ResourceKind {
37    /// The source of a rewrite, rename or copy operation, or generally the old version of a resource.
38    OldOrSource,
39    /// The destination of a rewrite, rename or copy operation, or generally the new version of a resource.
40    NewOrDestination,
41}
42
43/// A set of values to define how to diff something that is associated with it using `git-attributes`, relevant for regular files.
44///
45/// Some values are related to diffing, some are related to conversions.
46#[derive(Default, Debug, Clone, PartialEq, Eq)]
47pub struct Driver {
48    /// The name of the driver, as referred to by `[diff "name"]` in the git configuration.
49    pub name: BString,
50    /// The command to execute to perform the diff entirely like `<command> old-file old-hex old-mode new-file new-hex new-mode`.
51    ///
52    /// Please note that we don't make this call ourselves, but use it to determine that we should not run the our standard
53    /// built-in algorithm but bail instead as the output of such a program isn't standardized.
54    pub command: Option<BString>,
55    /// The per-driver algorithm to use.
56    pub algorithm: Option<Algorithm>,
57    /// The external filter program to call like `<binary_to_text_command> /path/to/blob` which outputs a textual version of the provided
58    /// binary file.
59    /// Note that it's invoked with a shell if arguments are given.
60    /// Further, if present, it will always be executed, whether `is_binary` is set or not.
61    pub binary_to_text_command: Option<BString>,
62    /// `Some(true)` if this driver deals with binary files, which means that a `binary_to_text_command` should be used to convert binary
63    /// into a textual representation.
64    /// Without such a command, anything that is considered binary is not diffed, but only the size of its data is made available.
65    /// If `Some(false)`, it won't be considered binary, and the its data will not be sampled for the null-byte either.
66    /// Leaving it to `None` means binary detection is automatic, and is based on the presence of the `0` byte in the first 8kB of the buffer.
67    pub is_binary: Option<bool>,
68}
69
70/// A conversion pipeline to take an object or path from what's stored in `git` to what can be diffed, while
71/// following the guidance of git-attributes at the respective path to learn if diffing should happen or if
72/// the content is considered binary.
73///
74/// There are two different conversion flows, where the target of the flow is a buffer with diffable content:
75// TODO: update this with information about possible directions.
76///
77/// * `worktree on disk` -> `text conversion`
78/// * `object` -> `worktree-filters` -> `text conversion`
79#[derive(Clone)]
80pub struct Pipeline {
81    /// A way to read data directly from the worktree.
82    pub roots: pipeline::WorktreeRoots,
83    /// A pipeline to convert objects from what's stored in `git` to its worktree version.
84    pub worktree_filter: gix_filter::Pipeline,
85    /// Options affecting the way we read files.
86    pub options: pipeline::Options,
87    /// Drivers to help customize the conversion behaviour depending on the location of items.
88    drivers: Vec<Driver>,
89    /// Pre-configured attributes to obtain additional diff-related information.
90    attrs: gix_filter::attributes::search::Outcome,
91    /// A buffer to manipulate paths
92    path: PathBuf,
93}
94
95/// A utility for performing a diff of two blobs, including flexible conversions, conversion-caching
96/// acquisition of diff information.
97/// Note that this instance will not call external filters as their output can't be known programmatically,
98/// but it allows to prepare their input if the caller wishes to perform this task.
99///
100/// Optimized for NxM lookups with built-in caching.
101#[derive(Clone)]
102pub struct Platform {
103    /// The old version of a diff-able blob, if set.
104    old: Option<platform::CacheKey>,
105    /// The new version of a diff-able blob, if set.
106    new: Option<platform::CacheKey>,
107
108    /// Options to alter how diffs should be performed.
109    pub options: platform::Options,
110    /// A way to convert objects into a diff-able format.
111    pub filter: Pipeline,
112    /// A way to access .gitattributes
113    pub attr_stack: gix_worktree::Stack,
114
115    /// The way we convert resources into diffable states.
116    pub filter_mode: pipeline::Mode,
117    /// A continuously growing cache keeping ready-for-diff blobs by their path in the worktree,
118    /// as that is what affects their final diff-able state.
119    ///
120    /// That way, expensive rewrite-checks with NxM matrix checks would be as fast as possible,
121    /// avoiding duplicate work.
122    diff_cache: HashMap<platform::CacheKey, platform::CacheValue>,
123    /// A list of previously used buffers, ready for re-use.
124    free_list: Vec<Vec<u8>>,
125}
126
127mod impls {
128    use crate::blob::ResourceKind;
129
130    impl std::fmt::Display for ResourceKind {
131        fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
132            f.write_str(match self {
133                ResourceKind::OldOrSource => "old",
134                ResourceKind::NewOrDestination => "new",
135            })
136        }
137    }
138}
gix_diff/blob/mod.rs

gix_diff/blob/
mod.rs
