diff.c: color moved lines differently

When a patch consists mostly of moving blocks of code around, it can
be quite tedious to ensure that the blocks are moved verbatim, and not
undesirably modified in the move. To that end, color blocks that are
moved within the same patch differently. For example (OM, del, add,
and NM are different colors):

    [OM]  -void sensitive_stuff(void)
    [OM]  -{
    [OM]  -        if (!is_authorized_user())
    [OM]  -                die("unauthorized");
    [OM]  -        sensitive_stuff(spanning,
    [OM]  -                        multiple,
    [OM]  -                        lines);
    [OM]  -}

           void another_function()
           {
    [del] -        printf("foo");
    [add] +        printf("bar");
           }

    [NM]  +void sensitive_stuff(void)
    [NM]  +{
    [NM]  +        if (!is_authorized_user())
    [NM]  +                die("unauthorized");
    [NM]  +        sensitive_stuff(spanning,
    [NM]  +                        multiple,
    [NM]  +                        lines);
    [NM]  +}

However adjacent blocks may be problematic. For example, in this
potentially malicious patch, the swapping of blocks can be spotted:

    [OM]  -void sensitive_stuff(void)
    [OM]  -{
    [OMA] -        if (!is_authorized_user())
    [OMA] -                die("unauthorized");
    [OM]  -        sensitive_stuff(spanning,
    [OM]  -                        multiple,
    [OM]  -                        lines);
    [OMA] -}

           void another_function()
           {
    [del] -        printf("foo");
    [add] +        printf("bar");
           }

    [NM]  +void sensitive_stuff(void)
    [NM]  +{
    [NMA] +        sensitive_stuff(spanning,
    [NMA] +                        multiple,
    [NMA] +                        lines);
    [NM]  +        if (!is_authorized_user())
    [NM]  +                die("unauthorized");
    [NMA] +}

If the moved code is larger, it is easier to hide some permutation in the
code, which is why some alternative coloring is needed.

This patch implements the first mode:
* basic alternating 'Zebra' mode
  This conveys all information needed to the user.  Defer customization to
  later patches.

First I implemented an alternative design, which would try to fingerprint
a line by its neighbors to detect if we are in a block or at the boundary.
This idea iss error prone as it inspected each line and its neighboring
lines to determine if the line was (a) moved and (b) if was deep inside
a hunk by having matching neighboring lines. This is unreliable as the
we can construct hunks which have equal neighbors that just exceed the
number of lines inspected. (Think of 'AXYZBXYZCXYZD..' with each letter
as a line, that is permutated to AXYZCXYZBXYZD..').

Instead this provides a dynamic programming greedy algorithm that finds
the largest moved hunk and then has several modes on highlighting bounds.

A note on the options '--submodule=diff' and '--color-words/--word-diff':
In the conversion to use emit_line in the prior patches both submodules
as well as word diff output carefully chose to call emit_line with sign=0.
All output with sign=0 is ignored for move detection purposes in this
patch, such that no weird looking output will be generated for these
cases. This leads to another thought: We could pass on '--color-moved' to
submodules such that they color up moved lines for themselves. If we'd do
so only line moves within a repository boundary are marked up.

It is useful to have moved lines colored, but there are annoying corner
cases, such as a single line moved, that is very common. For example
in a typical patch of C code, we have closing braces that end statement
blocks or functions.

While it is technically true that these lines are moved as they show up
elsewhere, it is harmful for the review as the reviewers attention is
drawn to such a minor side annoyance.

For now let's have a simple solution of hardcoding the number of
moved lines to be at least 3 before coloring them. Note, that the
length is applied across all blocks to find the 'lonely' blocks
that pollute new code, but do not interfere with a permutated
block where each permutation has less lines than 3.

Helped-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Stefan Beller <sbeller@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

1 files changed, 330 insertions, 14 deletions

diff --git a/diff.c b/diff.c
index 31720abf8f..084ecc5678 100644
--- a/diff.c
+++ b/diff.c
@@ -16,6 +16,7 @@
 #include "userdiff.h"
 #include "submodule-config.h"
 #include "submodule.h"
+#include "hashmap.h"
 #include "ll-merge.h"
 #include "string-list.h"
 #include "argv-array.h"
@@ -32,6 +33,7 @@ static int diff_indent_heuristic = 1;
 static int diff_rename_limit_default = 400;
 static int diff_suppress_blank_empty;
 static int diff_use_color_default = -1;
+static int diff_color_moved_default;
 static int diff_context_default = 3;
 static int diff_interhunk_context_default;
 static const char *diff_word_regex_cfg;
@@ -56,6 +58,10 @@ static char diff_colors[][COLOR_MAXLEN] = {
 	GIT_COLOR_YELLOW,	/* COMMIT */
 	GIT_COLOR_BG_RED,	/* WHITESPACE */
 	GIT_COLOR_NORMAL,	/* FUNCINFO */
+	GIT_COLOR_MAGENTA,	/* OLD_MOVED */
+	GIT_COLOR_BLUE,		/* OLD_MOVED ALTERNATIVE */
+	GIT_COLOR_CYAN,		/* NEW_MOVED */
+	GIT_COLOR_YELLOW,	/* NEW_MOVED ALTERNATIVE */
 };
 
 static NORETURN void die_want_option(const char *option_name)
@@ -81,6 +87,14 @@ static int parse_diff_color_slot(const char *var)
 		return DIFF_WHITESPACE;
 	if (!strcasecmp(var, "func"))
 		return DIFF_FUNCINFO;
+	if (!strcasecmp(var, "oldmoved"))
+		return DIFF_FILE_OLD_MOVED;
+	if (!strcasecmp(var, "oldmovedalternative"))
+		return DIFF_FILE_OLD_MOVED_ALT;
+	if (!strcasecmp(var, "newmoved"))
+		return DIFF_FILE_NEW_MOVED;
+	if (!strcasecmp(var, "newmovedalternative"))
+		return DIFF_FILE_NEW_MOVED_ALT;
 	return -1;
 }
 
@@ -229,12 +243,40 @@ int git_diff_heuristic_config(const char *var, const char *value, void *cb)
 	return 0;
 }
 
+static int parse_color_moved(const char *arg)
+{
+	switch (git_parse_maybe_bool(arg)) {
+	case 0:
+		return COLOR_MOVED_NO;
+	case 1:
+		return COLOR_MOVED_DEFAULT;
+	default:
+		break;
+	}
+
+	if (!strcmp(arg, "no"))
+		return COLOR_MOVED_NO;
+	else if (!strcmp(arg, "zebra"))
+		return COLOR_MOVED_ZEBRA;
+	else if (!strcmp(arg, "default"))
+		return COLOR_MOVED_DEFAULT;
+	else
+		return error(_("color moved setting must be one of 'no', 'default', 'zebra'"));
+}
+
 int git_diff_ui_config(const char *var, const char *value, void *cb)
 {
 	if (!strcmp(var, "diff.color") || !strcmp(var, "color.diff")) {
 		diff_use_color_default = git_config_colorbool(var, value);
 		return 0;
 	}
+	if (!strcmp(var, "diff.colormoved")) {
+		int cm = parse_color_moved(value);
+		if (cm < 0)
+			return -1;
+		diff_color_moved_default = cm;
+		return 0;
+	}
 	if (!strcmp(var, "diff.context")) {
 		diff_context_default = git_config_int(var, value);
 		if (diff_context_default < 0)
@@ -602,7 +644,9 @@ enum diff_symbol {
  * 13-15 are WSEH_NEW | WSEH_OLD | WSEH_CONTEXT
  * 16 is marking if the line is blank at EOF
  */
-#define DIFF_SYMBOL_CONTENT_BLANK_LINE_EOF (1<<16)
+#define DIFF_SYMBOL_CONTENT_BLANK_LINE_EOF	(1<<16)
+#define DIFF_SYMBOL_MOVED_LINE			(1<<17)
+#define DIFF_SYMBOL_MOVED_LINE_ALT		(1<<18)
 #define DIFF_SYMBOL_CONTENT_WS_MASK (WSEH_NEW | WSEH_OLD | WSEH_CONTEXT | WS_RULE_MASK)
 
 /*
@@ -645,6 +689,243 @@ static void append_emitted_diff_symbol(struct diff_options *o,
 	f->line = e->line ? xmemdupz(e->line, e->len) : NULL;
 }
 
+struct moved_entry {
+	struct hashmap_entry ent;
+	const struct emitted_diff_symbol *es;
+	struct moved_entry *next_line;
+};
+
+static int next_byte(const char **cp, const char **endp,
+		     const struct diff_options *diffopt)
+{
+	int retval;
+
+	if (*cp > *endp)
+		return -1;
+
+	if (DIFF_XDL_TST(diffopt, IGNORE_WHITESPACE_CHANGE)) {
+		while (*cp < *endp && isspace(**cp))
+			(*cp)++;
+		/*
+		 * After skipping a couple of whitespaces, we still have to
+		 * account for one space.
+		 */
+		return (int)' ';
+	}
+
+	if (DIFF_XDL_TST(diffopt, IGNORE_WHITESPACE)) {
+		while (*cp < *endp && isspace(**cp))
+			(*cp)++;
+		/* return the first non-ws character via the usual below */
+	}
+
+	retval = (unsigned char)(**cp);
+	(*cp)++;
+	return retval;
+}
+
+static int moved_entry_cmp(const struct diff_options *diffopt,
+			   const struct moved_entry *a,
+			   const struct moved_entry *b,
+			   const void *keydata)
+{
+	const char *ap = a->es->line, *ae = a->es->line + a->es->len;
+	const char *bp = b->es->line, *be = b->es->line + b->es->len;
+
+	if (!(diffopt->xdl_opts & XDF_WHITESPACE_FLAGS))
+		return a->es->len != b->es->len  || memcmp(ap, bp, a->es->len);
+
+	if (DIFF_XDL_TST(diffopt, IGNORE_WHITESPACE_AT_EOL)) {
+		while (ae > ap && isspace(*ae))
+			ae--;
+		while (be > bp && isspace(*be))
+			be--;
+	}
+
+	while (1) {
+		int ca, cb;
+		ca = next_byte(&ap, &ae, diffopt);
+		cb = next_byte(&bp, &be, diffopt);
+		if (ca != cb)
+			return 1;
+		if (ca < 0)
+			return 0;
+	}
+}
+
+static unsigned get_string_hash(struct emitted_diff_symbol *es, struct diff_options *o)
+{
+	if (o->xdl_opts & XDF_WHITESPACE_FLAGS) {
+		static struct strbuf sb = STRBUF_INIT;
+		const char *ap = es->line, *ae = es->line + es->len;
+		int c;
+
+		strbuf_reset(&sb);
+		while (ae > ap && isspace(*ae))
+			ae--;
+		while ((c = next_byte(&ap, &ae, o)) > 0)
+			strbuf_addch(&sb, c);
+
+		return memhash(sb.buf, sb.len);
+	} else {
+		return memhash(es->line, es->len);
+	}
+}
+
+static struct moved_entry *prepare_entry(struct diff_options *o,
+					 int line_no)
+{
+	struct moved_entry *ret = xmalloc(sizeof(*ret));
+	struct emitted_diff_symbol *l = &o->emitted_symbols->buf[line_no];
+
+	ret->ent.hash = get_string_hash(l, o);
+	ret->es = l;
+	ret->next_line = NULL;
+
+	return ret;
+}
+
+static void add_lines_to_move_detection(struct diff_options *o,
+					struct hashmap *add_lines,
+					struct hashmap *del_lines)
+{
+	struct moved_entry *prev_line = NULL;
+
+	int n;
+	for (n = 0; n < o->emitted_symbols->nr; n++) {
+		struct hashmap *hm;
+		struct moved_entry *key;
+
+		switch (o->emitted_symbols->buf[n].s) {
+		case DIFF_SYMBOL_PLUS:
+			hm = add_lines;
+			break;
+		case DIFF_SYMBOL_MINUS:
+			hm = del_lines;
+			break;
+		default:
+			prev_line = NULL;
+			continue;
+		}
+
+		key = prepare_entry(o, n);
+		if (prev_line && prev_line->es->s == o->emitted_symbols->buf[n].s)
+			prev_line->next_line = key;
+
+		hashmap_add(hm, key);
+		prev_line = key;
+	}
+}
+
+static int shrink_potential_moved_blocks(struct moved_entry **pmb,
+					 int pmb_nr)
+{
+	int lp, rp;
+
+	/* Shrink the set of potential block to the remaining running */
+	for (lp = 0, rp = pmb_nr - 1; lp <= rp;) {
+		while (lp < pmb_nr && pmb[lp])
+			lp++;
+		/* lp points at the first NULL now */
+
+		while (rp > -1 && !pmb[rp])
+			rp--;
+		/* rp points at the last non-NULL */
+
+		if (lp < pmb_nr && rp > -1 && lp < rp) {
+			pmb[lp] = pmb[rp];
+			pmb[rp] = NULL;
+			rp--;
+			lp++;
+		}
+	}
+
+	/* Remember the number of running sets */
+	return rp + 1;
+}
+
+/* Find blocks of moved code, delegate actual coloring decision to helper */
+static void mark_color_as_moved(struct diff_options *o,
+				struct hashmap *add_lines,
+				struct hashmap *del_lines)
+{
+	struct moved_entry **pmb = NULL; /* potentially moved blocks */
+	int pmb_nr = 0, pmb_alloc = 0;
+	int n, flipped_block = 1, block_length = 0;
+
+
+	for (n = 0; n < o->emitted_symbols->nr; n++) {
+		struct hashmap *hm = NULL;
+		struct moved_entry *key;
+		struct moved_entry *match = NULL;
+		struct emitted_diff_symbol *l = &o->emitted_symbols->buf[n];
+		int i;
+
+		switch (l->s) {
+		case DIFF_SYMBOL_PLUS:
+			hm = del_lines;
+			key = prepare_entry(o, n);
+			match = hashmap_get(hm, key, o);
+			free(key);
+			break;
+		case DIFF_SYMBOL_MINUS:
+			hm = add_lines;
+			key = prepare_entry(o, n);
+			match = hashmap_get(hm, key, o);
+			free(key);
+			break;
+		default:
+			flipped_block = 1;
+		}
+
+		if (!match) {
+			if (block_length < COLOR_MOVED_MIN_BLOCK_LENGTH) {
+				for (i = 0; i < block_length + 1; i++) {
+					l = &o->emitted_symbols->buf[n - i];
+					l->flags &= ~DIFF_SYMBOL_MOVED_LINE;
+				}
+			}
+			pmb_nr = 0;
+			block_length = 0;
+			continue;
+		}
+
+		l->flags |= DIFF_SYMBOL_MOVED_LINE;
+		block_length++;
+
+		/* Check any potential block runs, advance each or nullify */
+		for (i = 0; i < pmb_nr; i++) {
+			struct moved_entry *p = pmb[i];
+			struct moved_entry *pnext = (p && p->next_line) ?
+					p->next_line : NULL;
+			if (pnext && !hm->cmpfn(o, pnext, match, NULL)) {
+				pmb[i] = p->next_line;
+			} else {
+				pmb[i] = NULL;
+			}
+		}
+
+		pmb_nr = shrink_potential_moved_blocks(pmb, pmb_nr);
+
+		if (pmb_nr == 0) {
+			/*
+			 * The current line is the start of a new block.
+			 * Setup the set of potential blocks.
+			 */
+			for (; match; match = hashmap_get_next(hm, match)) {
+				ALLOC_GROW(pmb, pmb_nr + 1, pmb_alloc);
+				pmb[pmb_nr++] = match;
+			}
+
+			flipped_block = (flipped_block + 1) % 2;
+		}
+
+		if (flipped_block)
+			l->flags |= DIFF_SYMBOL_MOVED_LINE_ALT;
+	}
+
+	free(pmb);
+}
 
 static void emit_line_ws_markup(struct diff_options *o,
 				const char *set, const char *reset,
@@ -720,14 +1001,24 @@ static void emit_diff_symbol_from_struct(struct diff_options *o,
 				    flags & (DIFF_SYMBOL_CONTENT_WS_MASK), 0);
 		break;
 	case DIFF_SYMBOL_PLUS:
-		set = diff_get_color_opt(o, DIFF_FILE_NEW);
+		if (flags & DIFF_SYMBOL_MOVED_LINE_ALT)
+			set = diff_get_color_opt(o, DIFF_FILE_NEW_MOVED_ALT);
+		else if (flags & DIFF_SYMBOL_MOVED_LINE)
+			set = diff_get_color_opt(o, DIFF_FILE_NEW_MOVED);
+		else
+			set = diff_get_color_opt(o, DIFF_FILE_NEW);
 		reset = diff_get_color_opt(o, DIFF_RESET);
 		emit_line_ws_markup(o, set, reset, line, len, '+',
 				    flags & DIFF_SYMBOL_CONTENT_WS_MASK,
 				    flags & DIFF_SYMBOL_CONTENT_BLANK_LINE_EOF);
 		break;
 	case DIFF_SYMBOL_MINUS:
-		set = diff_get_color_opt(o, DIFF_FILE_OLD);
+		if (flags & DIFF_SYMBOL_MOVED_LINE_ALT)
+			set = diff_get_color_opt(o, DIFF_FILE_OLD_MOVED_ALT);
+		else if (flags & DIFF_SYMBOL_MOVED_LINE)
+			set = diff_get_color_opt(o, DIFF_FILE_OLD_MOVED);
+		else
+			set = diff_get_color_opt(o, DIFF_FILE_OLD);
 		reset = diff_get_color_opt(o, DIFF_RESET);
 		emit_line_ws_markup(o, set, reset, line, len, '-',
 				    flags & DIFF_SYMBOL_CONTENT_WS_MASK, 0);
@@ -3740,6 +4031,8 @@ void diff_setup(struct diff_options *options)
 		options->a_prefix = "a/";
 		options->b_prefix = "b/";
 	}
+
+	options->color_moved = diff_color_moved_default;
 }
 
 void diff_setup_done(struct diff_options *options)
@@ -3849,6 +4142,9 @@ void diff_setup_done(struct diff_options *options)
 
 	if (DIFF_OPT_TST(options, FOLLOW_RENAMES) && options->pathspec.nr != 1)
 		die(_("--follow requires exactly one pathspec"));
+
+	if (!options->use_color || external_diff())
+		options->color_moved = 0;
 }
 
 static int opt_arg(const char *arg, int arg_short, const char *arg_long, int *val)
@@ -4273,7 +4569,19 @@ int diff_opt_parse(struct diff_options *options,
 	}
 	else if (!strcmp(arg, "--no-color"))
 		options->use_color = 0;
-	else if (!strcmp(arg, "--color-words")) {
+	else if (!strcmp(arg, "--color-moved")) {
+		if (diff_color_moved_default)
+			options->color_moved = diff_color_moved_default;
+		if (options->color_moved == COLOR_MOVED_NO)
+			options->color_moved = COLOR_MOVED_DEFAULT;
+	} else if (!strcmp(arg, "--no-color-moved"))
+		options->color_moved = COLOR_MOVED_NO;
+	else if (skip_prefix(arg, "--color-moved=", &arg)) {
+		int cm = parse_color_moved(arg);
+		if (cm < 0)
+			die("bad --color-moved argument: %s", arg);
+		options->color_moved = cm;
+	} else if (!strcmp(arg, "--color-words")) {
 		options->use_color = 1;
 		options->word_diff = DIFF_WORDS_COLOR;
 	}
@@ -5086,16 +5394,8 @@ static void diff_flush_patch_all_file_pairs(struct diff_options *o)
 	if (WSEH_NEW & WS_RULE_MASK)
 		die("BUG: WS rules bit mask overlaps with diff symbol flags");
 
-	/*
-	 * For testing purposes we want to make sure the diff machinery
-	 * works completely with the buffer. If there is anything emitted
-	 * outside the emit_string, then the order is screwed
-	 * up and the tests will fail.
-	 *
-	 * TODO (later in this series):
-	 * We'll unset this pointer in a later patch.
-	 */
-	o->emitted_symbols = &esm;
+	if (o->color_moved)
+		o->emitted_symbols = &esm;
 
 	for (i = 0; i < q->nr; i++) {
 		struct diff_filepair *p = q->queue[i];
@@ -5104,6 +5404,21 @@ static void diff_flush_patch_all_file_pairs(struct diff_options *o)
 	}
 
 	if (o->emitted_symbols) {
+		if (o->color_moved) {
+			struct hashmap add_lines, del_lines;
+
+			hashmap_init(&del_lines,
+				     (hashmap_cmp_fn)moved_entry_cmp, o, 0);
+			hashmap_init(&add_lines,
+				     (hashmap_cmp_fn)moved_entry_cmp, o, 0);
+
+			add_lines_to_move_detection(o, &add_lines, &del_lines);
+			mark_color_as_moved(o, &add_lines, &del_lines);
+
+			hashmap_free(&add_lines, 0);
+			hashmap_free(&del_lines, 0);
+		}
+
 		for (i = 0; i < esm.nr; i++)
 			emit_diff_symbol_from_struct(o, &esm.buf[i]);
 
@@ -5185,6 +5500,7 @@ void diff_flush(struct diff_options *options)
 			fclose(options->file);
 		options->file = xfopen("/dev/null", "w");
 		options->close_file = 1;
+		options->color_moved = 0;
 		for (i = 0; i < q->nr; i++) {
 			struct diff_filepair *p = q->queue[i];
 			if (check_pair_status(p))