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
@@ -6,24 +6,25 @@
 #include <cache.h>
 #include <grep.h>
 #include <object.h>
 #include <tree.h>
 #include <commit.h>
 #include <tag.h>
 #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
  */
 #define FMT_LONGDATE "%Y-%m-%d %H:%M:%S (%Z)"
 #define FMT_SHORTDATE "%Y-%m-%d"
 #define FMT_ATOMDATE "%Y-%m-%dT%H:%M:%SZ"
 
 
@@ -212,25 +213,26 @@ extern int chk_non_negative(int result, char *msg);
 extern char *trim_end(const char *str, char c);
 extern char *strlpart(char *txt, int maxlen);
 extern char *strrpart(char *txt, int maxlen);
 
 extern void cgit_add_ref(struct reflist *list, struct refinfo *ref);
 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);
 
 extern char *fmt(const char *format,...);
 
 extern struct commitinfo *cgit_parse_commit(struct commit *commit);
 extern struct taginfo *cgit_parse_tag(struct tag *tag);
 extern void cgit_parse_url(const char *url);
diff --git a/shared.c b/shared.c
index 578a544..dbb84d8 100644
--- a/shared.c
+++ b/shared.c
@@ -248,35 +248,44 @@ int filediff_cb(void *priv, mmbuffer_t *mb, int nbuf)
                 buflen = 0;
         }
         if (diffbuf) {
                 ((linediff_fn)priv)(diffbuf, buflen);
                 free(diffbuf);
                 diffbuf = NULL;
                 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;
         emit_cb.outf = filediff_cb;
         emit_cb.priv = fn;
         xdl_diff(&file1, &file2, &diff_params, &emit_params, &emit_cb);
         return 0;
 }
 
diff --git a/ui-diff.c b/ui-diff.c
index 878496d..2196745 100644
--- a/ui-diff.c
+++ b/ui-diff.c
@@ -18,24 +18,27 @@ static int total_adds, total_rems, max_changes;
 static int lines_added, lines_removed;
 
 static struct fileinfo {
         char status;
         unsigned char old_sha1[20];
         unsigned char new_sha1[20];
         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;
 
         switch (info->status) {
         case DIFF_STATUS_ADDED:
                 class = "add";
                 break;
         case DIFF_STATUS_COPIED:
@@ -77,68 +80,80 @@ static void print_fileinfo(struct fileinfo *info)
                 html("<span class='modechange'>[");
                 cgit_print_filemode(info->old_mode);
                 html("]</span>");
         }
         htmlf("</td><td class='%s'>", class);
         cgit_diff_link(info->new_path, NULL, NULL, ctx.qry.head, ctx.qry.sha1,
                        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%%;'/>",
               info->removed * 100.0 / max_changes);
         htmlf("<td class='none' style='width: %.1f%%;'/>",
               (max_changes - info->removed - info->added) * 100.0 / max_changes);
         html("</tr></table></td></tr>\n");
 }
 
 static void count_diff_lines(char *line, int len)
 {
         if (line && (len > 0)) {
                 if (line[0] == '+')
                         lines_added++;
                 else if (line[0] == '-')
                         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));
         }
         items[files-1].status = pair->status;
         hashcpy(items[files-1].old_sha1, pair->one->sha1);
         hashcpy(items[files-1].new_sha1, pair->two->sha1);
         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;
 }
 
 void cgit_print_diffstat(const unsigned char *old_sha1,
                          const unsigned char *new_sha1)
 {
         int i;
 
         html("<div class='diffstat-header'>");
@@ -224,35 +239,42 @@ static void header(unsigned char *sha1, char *path1, int mode1,
                 html("<br/>+++ b/");
                 if (mode2 != 0)
                         cgit_tree_link(path2, NULL, NULL, ctx.qry.head,
                                        sha1_to_hex(new_rev_sha1), path2);
                 else
                         html_txt(path2);
         }
         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;
         struct commit *commit, *commit2;
 
         if (!new_rev)
                 new_rev = ctx.qry.head;
         get_sha1(new_rev, new_rev_sha1);
         type = sha1_object_info(new_rev_sha1, &size);
diff --git a/ui-log.c b/ui-log.c
index 3202848..ba2ab03 100644
--- a/ui-log.c
+++ b/ui-log.c
@@ -17,27 +17,32 @@ void count_lines(char *line, int size)
         if (size <= 0)
                 return;
 
         if (line[0] == '+')
                 add_lines++;
 
         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];
 
         buf[sizeof(buf) - 1] = 0;
         deco = lookup_decoration(&name_decoration, &commit->object);
         while (deco) {
                 if (!prefixcmp(deco->name, "refs/heads/")) {
                         strncpy(buf, deco->name + 11, sizeof(buf) - 1);
diff --git a/ui-patch.c b/ui-patch.c
index 1d77336..5d665d3 100644
--- a/ui-patch.c
+++ b/ui-patch.c
@@ -48,35 +48,42 @@ static void header(unsigned char *sha1, char *path1, int mode1,
                 if (mode1 != 0 && mode2 != 0) {
                         htmlf(" %.6o", mode1);
                         if (mode2 != mode1)
                                 htmlf("..%.6o", mode2);
                 }
                 htmlf("\n--- a/%s\n", path1);
                 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;
         unsigned char sha1[20], old_sha1[20];
         char *patchname;
 
         if (!hex)
                 hex = ctx.qry.head;