From 0dd000d2366bf4c8ad7f290ef2250a6be544739c Mon Sep 17 00:00:00 2001
From: Martin von Zweigbergk <martinvonz@google.com>
Date: Wed, 7 Apr 2021 10:07:37 -0700
Subject: [PATCH] diff: do final refinement at byte-level for non-word bytes

This results in significantly more readable diffs on commits like
659393bec219 in this repo.


Before:
test bench_diff_10k_lines_reversed  ... bench:  38,122,998 ns/iter (+/- 557,688)
test bench_diff_10k_modified_lines  ... bench:  32,556,563 ns/iter (+/- 548,114)
test bench_diff_10k_unchanged_lines ... bench:       4,231 ns/iter (+/- 15)
test bench_diff_1k_lines_reversed   ... bench:     958,296 ns/iter (+/- 46,963)
test bench_diff_1k_modified_lines   ... bench:   3,014,723 ns/iter (+/- 15,830)
test bench_diff_1k_unchanged_lines  ... bench:         249 ns/iter (+/- 2)
test bench_diff_git_git_read_tree_c ... bench:      78,599 ns/iter (+/- 1,079)

After:
test bench_diff_10k_lines_reversed  ... bench:  38,289,493 ns/iter (+/- 413,712)
test bench_diff_10k_modified_lines  ... bench:  37,352,516 ns/iter (+/- 1,293,950)
test bench_diff_10k_unchanged_lines ... bench:       4,238 ns/iter (+/- 13)
test bench_diff_1k_lines_reversed   ... bench:     967,253 ns/iter (+/- 8,506)
test bench_diff_1k_modified_lines   ... bench:   3,358,028 ns/iter (+/- 37,154)
test bench_diff_1k_unchanged_lines  ... bench:         233 ns/iter (+/- 1)
test bench_diff_git_git_read_tree_c ... bench:      95,787 ns/iter (+/- 740)


So the biggest slowdown is when there are modified lines.
---
 lib/src/diff.rs | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

diff --git a/lib/src/diff.rs b/lib/src/diff.rs
index c114164ef..f297e6ebf 100644
--- a/lib/src/diff.rs
+++ b/lib/src/diff.rs
@@ -37,18 +37,21 @@ pub fn find_line_ranges(text: &[u8]) -> Vec<Range<usize>> {
     ranges
 }
 
+fn is_word_byte(b: u8) -> bool {
+    // TODO: Make this configurable (probably higher up in the call stack)
+    matches!(b, b'A'..=b'Z' | b'a'..=b'z' | b'0'..=b'9' | b'_')
+}
+
 pub fn find_word_ranges(text: &[u8]) -> Vec<Range<usize>> {
     let mut word_ranges = vec![];
     let mut word_start_pos = 0;
     let mut in_word = false;
     for (i, b) in text.iter().enumerate() {
-        // TODO: Make this configurable (probably higher up in the call stack)
-        let is_word_byte = matches!(*b, b'A'..=b'Z' | b'a'..=b'z' | b'0'..=b'9' | b'_');
-        if in_word && !is_word_byte {
+        if in_word && !is_word_byte(*b) {
             in_word = false;
             word_ranges.push(word_start_pos..i);
             word_start_pos = i;
-        } else if !in_word && is_word_byte {
+        } else if !in_word && is_word_byte(*b) {
             in_word = true;
             word_start_pos = i;
         }
@@ -59,10 +62,10 @@ pub fn find_word_ranges(text: &[u8]) -> Vec<Range<usize>> {
     word_ranges
 }
 
-pub fn find_newline_ranges(text: &[u8]) -> Vec<Range<usize>> {
+pub fn find_nonword_ranges(text: &[u8]) -> Vec<Range<usize>> {
     let mut ranges = vec![];
     for (i, b) in text.iter().enumerate() {
-        if *b == b'\n' {
+        if !is_word_byte(*b) {
             ranges.push(i..i + 1);
         }
     }
@@ -472,7 +475,7 @@ pub fn diff<'a>(left: &'a [u8], right: &'a [u8]) -> Vec<SliceDiff<'a>> {
     let range_diffs = vec![RangeDiff::Replaced(0..left.len(), 0..right.len())];
     let range_diffs = refine_changed_ranges(left, right, &range_diffs, &find_line_ranges);
     let range_diffs = refine_changed_ranges(left, right, &range_diffs, &find_word_ranges);
-    let range_diffs = refine_changed_ranges(left, right, &range_diffs, &find_newline_ranges);
+    let range_diffs = refine_changed_ranges(left, right, &range_diffs, &find_nonword_ranges);
     range_diffs_to_slice_diffs(left, right, &range_diffs)
 }