diff: define block by number of alphanumeric chars

The existing behavior of diff --color-moved=zebra does not define the
minimum size of a block at all, instead relying on a heuristic applied
later to filter out sets of adjacent moved lines that are shorter than 3
lines long. This can be confusing, because a block could thus be colored
as moved at the source but not at the destination (or vice versa),
depending on its neighbors.

Instead, teach diff that the minimum size of a block is 20 alphanumeric
characters, the same heuristic used by "git blame". This allows diff to
still exclude uninteresting lines appearing on their own (such as those
solely consisting of one or a few closing braces), as was the intention
of the adjacent-moved-line heuristic.

This requires a change in some tests in that some of their lines are no
longer considered to be part of a block, because they are too short.

Signed-off-by: Jonathan Tan <[email protected]>
Signed-off-by: Junio C Hamano <[email protected]>

Author: jonathantanmy

Documentation/diff-options.txt

@@ -254,13 +254,11 @@ plain::
 	moved line, but it is not very useful in a review to determine
 	if a block of code was moved without permutation.
 zebra::
-	Blocks of moved code are detected greedily. The detected blocks are
+	Blocks of moved text of at least 20 alphanumeric characters
+	are detected greedily. The detected blocks are
 	painted using either the 'color.diff.{old,new}Moved' color or
 	'color.diff.{old,new}MovedAlternative'. The change between
-	the two colors indicates that a new block was detected. If there
-	are fewer than 3 adjacent moved lines, they are not marked up
-	as moved, but the regular colors 'color.diff.{old,new}' will be
-	used.
+	the two colors indicates that a new block was detected.
 dimmed_zebra::
 	Similar to 'zebra', but additional dimming of uninteresting parts
 	of moved code is performed. The bordering lines of two adjacent

diff.c

@@ -864,19 +864,31 @@ static int shrink_potential_moved_blocks(struct moved_entry **pmb,
 /*
  * If o->color_moved is COLOR_MOVED_PLAIN, this function does nothing.
  *
- * Otherwise, if the last block has fewer lines than
- * COLOR_MOVED_MIN_BLOCK_LENGTH, unset DIFF_SYMBOL_MOVED_LINE on all lines in
+ * Otherwise, if the last block has fewer alphanumeric characters than
+ * COLOR_MOVED_MIN_ALNUM_COUNT, unset DIFF_SYMBOL_MOVED_LINE on all lines in
  * that block.
  *
  * The last block consists of the (n - block_length)'th line up to but not
  * including the nth line.
+ *
+ * NEEDSWORK: This uses the same heuristic as blame_entry_score() in blame.c.
+ * Think of a way to unify them.
  */
 static void adjust_last_block(struct diff_options *o, int n, int block_length)
 {
-	int i;
-	if (block_length >= COLOR_MOVED_MIN_BLOCK_LENGTH ||
-	    o->color_moved == COLOR_MOVED_PLAIN)
+	int i, alnum_count = 0;
+	if (o->color_moved == COLOR_MOVED_PLAIN)
 		return;
+	for (i = 1; i < block_length + 1; i++) {
+		const char *c = o->emitted_symbols->buf[n - i].line;
+		for (; *c; c++) {
+			if (!isalnum(*c))
+				continue;
+			alnum_count++;
+			if (alnum_count >= COLOR_MOVED_MIN_ALNUM_COUNT)
+				return;
+		}
+	}
 	for (i = 1; i < block_length + 1; i++)
 		o->emitted_symbols->buf[n - i].flags &= ~DIFF_SYMBOL_MOVED_LINE;
 }
@@ -923,7 +935,6 @@ static void mark_color_as_moved(struct diff_options *o,
 		}
 
 		l->flags |= DIFF_SYMBOL_MOVED_LINE;
-		block_length++;
 
 		if (o->color_moved == COLOR_MOVED_PLAIN)
 			continue;
@@ -953,8 +964,13 @@ static void mark_color_as_moved(struct diff_options *o,
 			}
 
 			flipped_block = (flipped_block + 1) % 2;
+
+			adjust_last_block(o, n, block_length);
+			block_length = 0;
 		}
 
+		block_length++;
+
 		if (flipped_block)
 			l->flags |= DIFF_SYMBOL_MOVED_LINE_ALT;
 	}

diff.h

@@ -195,7 +195,7 @@ struct diff_options {
 		COLOR_MOVED_ZEBRA_DIM = 3,
 	} color_moved;
 	#define COLOR_MOVED_DEFAULT COLOR_MOVED_ZEBRA
-	#define COLOR_MOVED_MIN_BLOCK_LENGTH 3
+	#define COLOR_MOVED_MIN_ALNUM_COUNT 20
 };
 
 void diff_emit_submodule_del(struct diff_options *o, const char *line);