Handle binary files in diffs

This teaches all diff-related operations (i.e. ui-log, ui-diff and ui-patch)
how to handle binary files.

Signed-off-by: Lars Hjemli <hjemli@gmail.com>

5 files changed, 52 insertions, 7 deletions

diff --git a/cgit.h b/cgit.h
index 4fe94c6..8b0defa 100644
--- a/cgit.h
+++ b/cgit.h
@@ -12,12 +12,13 @@
 #include <diff.h>
 #include <diffcore.h>
 #include <refs.h>
 #include <revision.h>
 #include <log-tree.h>
 #include <archive.h>
+#include <xdiff-interface.h>
 #include <xdiff/xdiff.h>
 #include <utf8.h>
 
 
 /*
  * Dateformats used on misc. pages
@@ -218,13 +219,14 @@ extern int cgit_refs_cb(const char *refname, const unsigned char *sha1,
 			int flags, void *cb_data);
 
 extern void *cgit_free_commitinfo(struct commitinfo *info);
 
 extern int cgit_diff_files(const unsigned char *old_sha1,
 			   const unsigned char *new_sha1,
-			   linediff_fn fn);
+			   unsigned long *old_size, unsigned long *new_size,
+			   int *binary, linediff_fn fn);
 
 extern void cgit_diff_tree(const unsigned char *old_sha1,
 			   const unsigned char *new_sha1,
 			   filepair_fn fn, const char *prefix);
 
 extern void cgit_diff_commit(struct commit *commit, filepair_fn fn);
diff --git a/shared.c b/shared.c
index 578a544..dbb84d8 100644
--- a/shared.c
+++ b/shared.c
@@ -254,23 +254,32 @@ int filediff_cb(void *priv, mmbuffer_t *mb, int nbuf)
 		buflen = 0;
 	}
 	return 0;
 }
 
 int cgit_diff_files(const unsigned char *old_sha1,
-		     const unsigned char *new_sha1,
-		     linediff_fn fn)
+		    const unsigned char *new_sha1, unsigned long *old_size,
+		    unsigned long *new_size, int *binary, linediff_fn fn)
 {
 	mmfile_t file1, file2;
 	xpparam_t diff_params;
 	xdemitconf_t emit_params;
 	xdemitcb_t emit_cb;
 
 	if (!load_mmfile(&file1, old_sha1) || !load_mmfile(&file2, new_sha1))
 		return 1;
 
+	*old_size = file1.size;
+	*new_size = file2.size;
+
+	if (buffer_is_binary(file1.ptr, file1.size) ||
+	    buffer_is_binary(file2.ptr, file2.size)) {
+		*binary = 1;
+		return 0;
+	}
+
 	memset(&diff_params, 0, sizeof(diff_params));
 	memset(&emit_params, 0, sizeof(emit_params));
 	memset(&emit_cb, 0, sizeof(emit_cb));
 	diff_params.flags = XDF_NEED_MINIMAL;
 	emit_params.ctxlen = 3;
 	emit_params.flags = XDL_EMIT_FUNCNAMES;
diff --git a/ui-diff.c b/ui-diff.c
index 878496d..2196745 100644
--- a/ui-diff.c
+++ b/ui-diff.c
@@ -24,12 +24,15 @@ static struct fileinfo {
 	unsigned short old_mode;
 	unsigned short new_mode;
 	char *old_path;
 	char *new_path;
 	unsigned int added;
 	unsigned int removed;
+	unsigned long old_size;
+	unsigned long new_size;
+	int binary:1;
 } *items;
 
 
 static void print_fileinfo(struct fileinfo *info)
 {
 	char *class;
@@ -83,12 +86,17 @@ static void print_fileinfo(struct fileinfo *info)
 		       ctx.qry.sha2, info->new_path);
 	if (info->status == DIFF_STATUS_COPIED || info->status == DIFF_STATUS_RENAMED)
 		htmlf(" (%s from %s)",
 		      info->status == DIFF_STATUS_COPIED ? "copied" : "renamed",
 		      info->old_path);
 	html("</td><td class='right'>");
+	if (info->binary) {
+		htmlf("bin</td><td class='graph'>%d -> %d bytes",
+		      info->old_size, info->new_size);
+		return;
+	}
 	htmlf("%d", info->added + info->removed);
 	html("</td><td class='graph'>");
 	htmlf("<table summary='file diffstat' width='%d%%'><tr>", (max_changes > 100 ? 100 : max_changes));
 	htmlf("<td class='add' style='width: %.1f%%;'/>",
 	      info->added * 100.0 / max_changes);
 	htmlf("<td class='rem' style='width: %.1f%%;'/>",
@@ -107,16 +115,20 @@ static void count_diff_lines(char *line, int len)
 			lines_removed++;
 	}
 }
 
 static void inspect_filepair(struct diff_filepair *pair)
 {
+	int binary = 0;
+	unsigned long old_size = 0;
+	unsigned long new_size = 0;
 	files++;
 	lines_added = 0;
 	lines_removed = 0;
-	cgit_diff_files(pair->one->sha1, pair->two->sha1, count_diff_lines);
+	cgit_diff_files(pair->one->sha1, pair->two->sha1, &old_size, &new_size,
+			&binary, count_diff_lines);
 	if (files >= slots) {
 		if (slots == 0)
 			slots = 4;
 		else
 			slots = slots * 2;
 		items = xrealloc(items, slots * sizeof(struct fileinfo));
@@ -127,12 +139,15 @@ static void inspect_filepair(struct diff_filepair *pair)
 	items[files-1].old_mode = pair->one->mode;
 	items[files-1].new_mode = pair->two->mode;
 	items[files-1].old_path = xstrdup(pair->one->path);
 	items[files-1].new_path = xstrdup(pair->two->path);
 	items[files-1].added = lines_added;
 	items[files-1].removed = lines_removed;
+	items[files-1].old_size = old_size;
+	items[files-1].new_size = new_size;
+	items[files-1].binary = binary;
 	if (lines_added + lines_removed > max_changes)
 		max_changes = lines_added + lines_removed;
 	total_adds += lines_added;
 	total_rems += lines_removed;
 }
 
@@ -230,23 +245,30 @@ static void header(unsigned char *sha1, char *path1, int mode1,
 	}
 	html("</div>");
 }
 
 static void filepair_cb(struct diff_filepair *pair)
 {
+	unsigned long old_size = 0;
+	unsigned long new_size = 0;
+	int binary = 0;
+
 	header(pair->one->sha1, pair->one->path, pair->one->mode,
 	       pair->two->sha1, pair->two->path, pair->two->mode);
 	if (S_ISGITLINK(pair->one->mode) || S_ISGITLINK(pair->two->mode)) {
 		if (S_ISGITLINK(pair->one->mode))
 			print_line(fmt("-Subproject %s", sha1_to_hex(pair->one->sha1)), 52);
 		if (S_ISGITLINK(pair->two->mode))
 			print_line(fmt("+Subproject %s", sha1_to_hex(pair->two->sha1)), 52);
 		return;
 	}
-	if (cgit_diff_files(pair->one->sha1, pair->two->sha1, print_line))
+	if (cgit_diff_files(pair->one->sha1, pair->two->sha1, &old_size, 
+			    &new_size, &binary, print_line))
 		cgit_print_error("Error running diff");
+	if (binary)
+		html("Binary files differ");
 }
 
 void cgit_print_diff(const char *new_rev, const char *old_rev, const char *prefix)
 {
 	enum object_type type;
 	unsigned long size;
diff --git a/ui-log.c b/ui-log.c
index 3202848..ba2ab03 100644
--- a/ui-log.c
+++ b/ui-log.c
@@ -23,15 +23,20 @@ void count_lines(char *line, int size)
 	else if (line[0] == '-')
 		rem_lines++;
 }
 
 void inspect_files(struct diff_filepair *pair)
 {
+	unsigned long old_size = 0;
+	unsigned long new_size = 0;
+	int binary = 0;
+
 	files++;
 	if (ctx.repo->enable_log_linecount)
-		cgit_diff_files(pair->one->sha1, pair->two->sha1, count_lines);
+		cgit_diff_files(pair->one->sha1, pair->two->sha1, &old_size,
+				&new_size, &binary, count_lines);
 }
 
 void show_commit_decorations(struct commit *commit)
 {
 	struct name_decoration *deco;
 	static char buf[1024];
diff --git a/ui-patch.c b/ui-patch.c
index 1d77336..5d665d3 100644
--- a/ui-patch.c
+++ b/ui-patch.c
@@ -54,23 +54,30 @@ static void header(unsigned char *sha1, char *path1, int mode1,
 		htmlf("+++ b/%s\n", path2);
 	}
 }
 
 static void filepair_cb(struct diff_filepair *pair)
 {
+	unsigned long old_size = 0;
+	unsigned long new_size = 0;
+	int binary = 0;
+
 	header(pair->one->sha1, pair->one->path, pair->one->mode,
 	       pair->two->sha1, pair->two->path, pair->two->mode);
 	if (S_ISGITLINK(pair->one->mode) || S_ISGITLINK(pair->two->mode)) {
 		if (S_ISGITLINK(pair->one->mode))
 			print_line(fmt("-Subproject %s", sha1_to_hex(pair->one->sha1)), 52);
 		if (S_ISGITLINK(pair->two->mode))
 			print_line(fmt("+Subproject %s", sha1_to_hex(pair->two->sha1)), 52);
 		return;
 	}
-	if (cgit_diff_files(pair->one->sha1, pair->two->sha1, print_line))
+	if (cgit_diff_files(pair->one->sha1, pair->two->sha1, &old_size,
+			    &new_size, &binary, print_line))
 		html("Error running diff");
+	if (binary)
+		html("Binary files differ\n");
 }
 
 void cgit_print_patch(char *hex)
 {
 	struct commit *commit;
 	struct commitinfo *info;