[bug-diffutils] [PATCH] Escape file names that contain special characters

[Andreas Gruenbacher]

On Don, 2012-08-02 at 23:13 +0200, Jim Meyering wrote:
> Paul Eggert wrote:
> > On 08/02/2012 01:02 PM, Andreas Grünbacher wrote:
> >> Paul and Jim, any thoughts about adding filename quoting to GNU diff 
> >> anytime
> >> soon, or about the format?
> >
> > The git format looks OK.  I don't have time right now to add it to GNU 
> > diff, tho.
> 
> I agree, and have no more free time for the task than Paul does.
> Would someone care to contribute a complete patch?

Here's an attempt.  Needs to be mentioned in NEWS still at least.  Can
one of you give this a close look?

Thanks,
Andreas


From: Andreas Gruenbacher <address@hidden>
Date: Tue, 7 Aug 2012 02:34:24 +0200
Subject: [PATCH] diff: Escape file names that contain special characters

If a file name contains space, '"', '\\', or control characters, it is enclosed
in double quotes, and all characters other than space are backslash escaped as
in C strings.
* src/util.c (c_escape_char): New function.
(c_escape): New function.
(begin_output): Escape file names when needed.
* src/context.c (print_context_header): New names parameter.
(print_context_label): New name parameter.
* src/diff.h (print_context_header): Change prototype.
---
 src/context.c |   13 ++++---
 src/diff.h    |    2 +-
 src/util.c    |  100 ++++++++++++++++++++++++++++++++++++++++++++++++++++++---
 3 files changed, 103 insertions(+), 12 deletions(-)

diff --git a/src/context.c b/src/context.c
index b73d5c3..53aa203 100644
--- a/src/context.c
+++ b/src/context.c
@@ -40,6 +40,7 @@ static lin find_function_last_match;
 static void
 print_context_label (char const *mark,
                     struct file_data *inf,
+                    char const *name,
                     char const *label)
 {
   if (label)
@@ -70,24 +71,24 @@ print_context_label (char const *mark,
              sprintf (buf, "%"PRIuMAX".%.9d", sec, nsec);
            }
        }
-      fprintf (outfile, "%s %s\t%s\n", mark, inf->name, buf);
+      fprintf (outfile, "%s %s\t%s\n", mark, name, buf);
     }
 }
 
 /* Print a header for a context diff, with the file names and dates.  */
 
 void
-print_context_header (struct file_data inf[], bool unidiff)
+print_context_header (struct file_data inf[], char const *const *names, bool 
unidiff)
 {
   if (unidiff)
     {
-      print_context_label ("---", &inf[0], file_label[0]);
-      print_context_label ("+++", &inf[1], file_label[1]);
+      print_context_label ("---", &inf[0], names[0], file_label[0]);
+      print_context_label ("+++", &inf[1], names[1], file_label[1]);
     }
   else
     {
-      print_context_label ("***", &inf[0], file_label[0]);
-      print_context_label ("---", &inf[1], file_label[1]);
+      print_context_label ("***", &inf[0], names[0], file_label[0]);
+      print_context_label ("---", &inf[1], names[1], file_label[1]);
     }
 }
 
diff --git a/src/diff.h b/src/diff.h
index 4b6595c..212a8d0 100644
--- a/src/diff.h
+++ b/src/diff.h
@@ -332,7 +332,7 @@ XTERN FILE *outfile;
 extern int diff_2_files (struct comparison *);
 
 /* context.c */
-extern void print_context_header (struct file_data[], bool);
+extern void print_context_header (struct file_data[], char const * const *, 
bool);
 extern void print_context_script (struct change *, bool);
 
 /* dir.c */
diff --git a/src/util.c b/src/util.c
index 38b33ef..ca6ef2a 100644
--- a/src/util.c
+++ b/src/util.c
@@ -166,24 +166,109 @@ setup_output (char const *name0, char const *name1, bool 
recursive)
 static pid_t pr_pid;
 #endif
 
+static char c_escape_char (char c)
+{
+  switch (c) {
+    case '\a': return 'a';
+    case '\b': return 'b';
+    case '\t': return 't';
+    case '\n': return 'n';
+    case '\v': return 'v';
+    case '\f': return 'f';
+    case '\r': return 'r';
+    case '"': return '"';
+    case '\\': return '\\';
+    default:
+      return c < 32;
+  }
+}
+
+char *
+c_escape (char const *str)
+{
+  char const *s;
+  unsigned int plus = 0;
+  bool must_quote = false;
+
+  for (s = str; *s; s++)
+    {
+      char c = *s;
+
+      if (c == ' ')
+       {
+         must_quote = true;
+         continue;
+       }
+      switch (c_escape_char (*s))
+       {
+         case 1:
+           plus += 3;
+           /* fall through */
+         case 0:
+           break;
+         default:
+           plus++;
+           break;
+       }
+    }
+
+  if (must_quote || plus)
+    {
+      char *buffer, *b;
+
+      b = buffer = xmalloc (strlen (str) + plus + 3);
+      *b++ = '"';
+      for (s = str; *s; s++)
+       {
+         char c = *s;
+         char escape = c_escape_char (c);
+
+         switch (escape)
+           {
+             case 0:
+               *b++ = c;
+               break;
+             case 1:
+               *b++ = '\\';
+               *b++ = ((c >> 6) & 03) + '0';
+               *b++ = ((c >> 3) & 07) + '0';
+               *b++ = ((c >> 0) & 07) + '0';
+               break;
+             default:
+               *b++ = '\\';
+               *b++ = escape;
+               break;
+           }
+       }
+      *b++ = '"';
+      *b = 0;
+      return buffer;
+    }
+  else
+    return (char *) str;
+}
+
 void
 begin_output (void)
 {
+  char *names[2];
   char *name;
 
   if (outfile != 0)
     return;
 
+  names[0] = c_escape (current_name0);
+  names[1] = c_escape (current_name1);
+
   /* Construct the header of this piece of diff.  */
-  name = xmalloc (strlen (current_name0) + strlen (current_name1)
-                 + strlen (switch_string) + 7);
+  name = xmalloc (strlen (names[0]) + strlen (names[1]) + strlen 
(switch_string) + 7);
 
   /* POSIX 1003.1-2001 specifies this format.  But there are some bugs in
      the standard: it says that we must print only the last component
      of the pathnames, and it requires two spaces after "diff" if
      there are no options.  These requirements are silly and do not
      match historical practice.  */
-  sprintf (name, "diff%s %s %s", switch_string, current_name0, current_name1);
+  sprintf (name, "diff%s %s %s", switch_string, names[0], names[1]);
 
   if (paginate)
     {
@@ -258,16 +343,21 @@ begin_output (void)
   switch (output_style)
     {
     case OUTPUT_CONTEXT:
-      print_context_header (files, false);
+      print_context_header (files, (char const *const *)names, false);
       break;
 
     case OUTPUT_UNIFIED:
-      print_context_header (files, true);
+      print_context_header (files, (char const *const *)names, true);
       break;
 
     default:
       break;
     }
+
+  if (names[0] != current_name0)
+    free (names[0]);
+  if (names[1] != current_name1)
+    free (names[1]);
 }
 
 /* Call after the end of output of diffs for one file.
-- 
1.7.7.6