Json output upgrades (#3)

* Sort JSON lines and enrich them * Fix release.yml * Slimmer testing * Cargo fmt * actionlint * Less testing * Replace map with list to preserve order --------- Co-authored-by: Nimrod <[email protected]> Replace map with list to preserve order Replace map with list to preserve order
baz-scm · May 20, 2024 · e67ef93 · e67ef93
1 parent a9bcee7
commit e67ef93
Show file tree

Hide file tree

Showing 7 changed files with 36 additions and 114 deletions.
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
@@ -5,65 +5,39 @@ on:
     tags:
       - "[0-9]+.*"
 
+permissions:
+  contents: write
+  packages: write
+
 jobs:
   create-release:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-22.04
     steps:
-      - uses: actions/checkout@v3
-      - uses: taiki-e/create-gh-release-action@v1
+      - uses: actions/checkout@v4
+      - name: Create a Release
+        uses: elgohr/Github-Release-Action@v5
         with:
-          # Create a draft release so we can add the changelog before publishing it
-          # so the changelog is included in the release email
-          draft: true
+          title: ${{ github.ref }}
         env:
-          # (required)
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
 
   upload-assets:
     strategy:
       fail-fast: false
       matrix:
         include:
-          # Targets that match the OS of the runner.
-          - target: x86_64-unknown-linux-gnu
-            os: ubuntu-20.04
-          - target: x86_64-apple-darwin
-            os: macos-12
-          - target: x86_64-pc-windows-msvc
-            os: windows-latest
-
-          # Targets using cross-compilation (upload-rust-binary-action
-          # detects that they need cross).
-
-          # TODO: fix musl so it statically links during release, see
-          # https://github.com/Wilfred/difftastic/issues/563
-          #
-          # - target: x86_64-unknown-linux-musl
-          #   os: ubuntu-20.04
           - target: aarch64-unknown-linux-gnu
             os: ubuntu-20.04
           - target: aarch64-apple-darwin
             os: macos-latest
     runs-on: ${{ matrix.os }}
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - uses: taiki-e/upload-rust-binary-action@v1
         with:
           bin: difft
           # (optional) On which platform to distribute the `.tar.gz` file.
           tar: unix
-          # (optional) On which platform to distribute the `.zip` file.
-          zip: windows
           # (optional) Target triple, default is host triple.
           target: ${{ matrix.target }}
-        env:
-          # (required)
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-
-  push_crates_io:
-    runs-on: ubuntu-20.04
-    steps:
-      - uses: actions/checkout@v3
-      - uses: katyo/publish-crates@v1
-        with:
-          registry-token: ${{ secrets.CARGO_REGISTRY_TOKEN }}
+          token: ${{ secrets.GITHUB_TOKEN }}
diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
@@ -1,4 +1,4 @@
-on: [push, pull_request]
+on: [ push, pull_request ]
 
 name: Continuous integration
 
@@ -10,65 +10,24 @@ jobs:
       fail-fast: false
       matrix:
         job:
-          # Operating systems available: https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners/about-github-hosted-runners#standard-github-hosted-runners-for-public-repositories
-          - { target: x86_64-unknown-linux-gnu, os: ubuntu-20.04 }
-          # mac-latest (currently mac-14) is an ARM device, so use macos-12 to get Intel.
-          # Update annoucement: https://github.blog/changelog/2024-04-01-macos-14-sonoma-is-generally-available-and-the-latest-macos-runner-image/
-          - { target: x86_64-apple-darwin, os: macos-12 }
-          - { target: x86_64-pc-windows-msvc, os: windows-latest }
-
-          - { target: aarch64-unknown-linux-gnu, os: ubuntu-20.04, use-cross: true }
-
-          # mac-14 is an M1 ARM device.
           - { target: aarch64-apple-darwin, os: macos-14 }
-
-          # musl binaries produced by GitHub actions segfault when run, see
-          # https://github.com/Wilfred/difftastic/issues/563
-          #
-          # musl builds seem to work fine elsewhere, so just verify
-          # that compilation works for now.
-          - { target: x86_64-unknown-linux-musl, os: ubuntu-20.04, use-cross: true, build-only: true }
     env:
       BUILD_CMD: cargo
       SUBCOMMAND: test
 
     steps:
-    - name: Checkout source code
-      uses: actions/checkout@v3
-
-    - name: Install Rust toolchain
-      uses: dtolnay/[email protected]
-      with:
-        targets: ${{ matrix.job.target }}
-
-    - name: Install cross
-      if: matrix.job.use-cross
-      uses: taiki-e/install-action@v2
-      with:
-        tool: [email protected]
-
-    - name: Overwrite build command env variable
-      if: matrix.job.use-cross
-      shell: bash
-      run: echo "BUILD_CMD=cross" >> $GITHUB_ENV
+      - name: Checkout source code
+        uses: actions/checkout@v3
 
-    - name: Overwrite build subcommand env variable
-      if: matrix.job.build-only
-      shell: bash
-      run: echo "SUBCOMMAND=build" >> $GITHUB_ENV
+      - name: Install Rust toolchain
+        uses: dtolnay/[email protected]
+        with:
+          targets: ${{ matrix.job.target }}
 
-    - name: Test
-      shell: bash
-      run: $BUILD_CMD $SUBCOMMAND --target ${{ matrix.job.target }}
 
-  test_mime_db:
-    name: Test with MIME database
-    runs-on: ubuntu-20.04
-    steps:
-      - uses: actions/checkout@v3
-      - uses: dtolnay/[email protected]
-      # This runs tests that rely on the MIME database being present.
-      - run: cargo test -- --ignored
+      - name: Test
+        shell: bash
+        run: $BUILD_CMD $SUBCOMMAND --target ${{ matrix.job.target }}
 
   test_linux_latest_stable:
     name: Test Linux (latest stable Rust)

diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -78,7 +78,7 @@ strum = { version = "0.25", features = ["derive"] }
 hashbrown = "0.14.0"
 humansize = "2.1.3"
 serde = { version = "1.0", features = ["derive"] }
-serde_json = "1.0"
+serde_json = { version = "1.0", features = ["preserve_order"] }
 line-numbers = "0.3.0"
 smallvec = "1.13.2"
 

diff --git a/sample_files/compare.expected b/sample_files/compare.expected
@@ -68,7 +68,7 @@ sample_files/erlang_1.erl sample_files/erlang_2.erl
 dccdb8f65d2f099ab1a8cb66011376a2  -
 
 sample_files/f_sharp_1.fs sample_files/f_sharp_2.fs
-6ba65e62a804124f56bdfb19afc8cd40  -
+1a9173c15a42c1ebb9522109df172faf  -
 
 sample_files/hack_1.php sample_files/hack_2.php
 c2bb0aa7d7b07d6ced79f6a5363e878b  -

diff --git a/src/display/json.rs b/src/display/json.rs
@@ -1,13 +1,10 @@
-use std::collections::HashMap;
-
 use line_numbers::LineNumber;
 use serde::{ser::SerializeStruct, Serialize, Serializer};
 
 use crate::{
     display::{
         context::{all_matched_lines_filled, opposite_positions},
         hunks::{matched_lines_indexes_for_hunk, matched_pos_to_hunks, merge_adjacent},
-        side_by_side::lines_with_novel,
     },
     lines::MaxLine,
     parse::syntax::{self, MatchedPos, StringKind},
@@ -99,9 +96,6 @@ impl<'f> From<&'f DiffResult> for File<'f> {
                 let lhs_lines = lhs_src.split('\n').collect::<Vec<&str>>();
                 let rhs_lines = rhs_src.split('\n').collect::<Vec<&str>>();
 
-                let (lhs_lines_with_novel, rhs_lines_with_novel) =
-                    lines_with_novel(&summary.lhs_positions, &summary.rhs_positions);
-
                 let matched_lines = all_matched_lines_filled(
                     &summary.lhs_positions,
                     &summary.rhs_positions,
@@ -112,25 +106,15 @@ impl<'f> From<&'f DiffResult> for File<'f> {
 
                 let mut chunks = Vec::with_capacity(hunks.len());
                 for hunk in &hunks {
-                    let mut lines = HashMap::with_capacity(hunk.lines.len());
+                    let mut lines = Vec::with_capacity(hunk.lines.len());
 
-                    let (start_i, end_i) = matched_lines_indexes_for_hunk(matched_lines, hunk, 0);
+                    let (start_i, end_i) = matched_lines_indexes_for_hunk(matched_lines, hunk, 3);
                     let aligned_lines = &matched_lines[start_i..end_i];
                     matched_lines = &matched_lines[start_i..];
 
                     for (lhs_line_num, rhs_line_num) in aligned_lines {
-                        if !lhs_lines_with_novel.contains(&lhs_line_num.unwrap_or(LineNumber(0)))
-                            && !rhs_lines_with_novel
-                                .contains(&rhs_line_num.unwrap_or(LineNumber(0)))
-                        {
-                            continue;
-                        }
-
-                        let line = lines
-                            .entry((lhs_line_num.map(|l| l.0), rhs_line_num.map(|l| l.0)))
-                            .or_insert_with(|| {
-                                Line::new(lhs_line_num.map(|l| l.0), rhs_line_num.map(|l| l.0))
-                            });
+                        let mut line =
+                            Line::new(lhs_line_num.map(|l| l.0), rhs_line_num.map(|l| l.0));
 
                         if let Some(line_num) = lhs_line_num {
                             add_changes_to_side(
@@ -148,9 +132,10 @@ impl<'f> From<&'f DiffResult> for File<'f> {
                                 &summary.rhs_positions,
                             );
                         }
+                        lines.push(line);
                     }
 
-                    chunks.push(lines.into_values().collect());
+                    chunks.push(lines);
                 }
 
                 File::with_sections(&summary.file_format, &summary.display_path, chunks)
@@ -230,6 +215,7 @@ struct Change<'c> {
     end: u32,
     content: &'c str,
     highlight: Highlight,
+    highlight_type: &'c str,
 }
 
 #[derive(Debug, Serialize)]
@@ -297,7 +283,7 @@ fn add_changes_to_side<'s>(
     side: &mut Side<'s>,
     line_num: LineNumber,
     src_lines: &[&'s str],
-    all_matches: &[MatchedPos],
+    all_matches: &'s [MatchedPos],
 ) {
     let src_line = src_lines[line_num.0 as usize];
 
@@ -308,6 +294,7 @@ fn add_changes_to_side<'s>(
             end: m.pos.end_col,
             content: &src_line[(m.pos.start_col as usize)..(m.pos.end_col as usize)],
             highlight: Highlight::from_match(&m.kind),
+            highlight_type: m.kind.as_ref(),
         })
     }
 }

diff --git a/src/parse/syntax.rs b/src/parse/syntax.rs
@@ -6,6 +6,7 @@ use std::{cell::Cell, env, fmt, hash::Hash, num::NonZeroU32};
 
 use line_numbers::LinePositions;
 use line_numbers::SingleLineSpan;
+use strum::AsRefStr;
 use typed_arena::Arena;
 
 use self::Syntax::*;
@@ -589,7 +590,7 @@ pub(crate) enum TokenKind {
 }
 
 /// A matched token (an atom, a delimiter, or a comment word).
-#[derive(PartialEq, Eq, Debug, Clone)]
+#[derive(PartialEq, Eq, Debug, Clone, AsRefStr)]
 pub(crate) enum MatchKind {
     UnchangedToken {
         highlight: TokenKind,