Add unified reject format and the --reject-format=FORMAT option

ChangeLog

@@ -1,3 +1,15 @@
+2009-03-20  Andreas Gruenbacher  <agruen@suse.de>
+
+	* patch.c: New option --reject-format=FORMAT.
+	(abort_hunk_context): Rename from abort_hunk().
+	(abort_hunk_unified, abort_hunk, mangled_patch,
+	print_unidiff_range): New functions.
+	* patch.man: Document this.
+	* pch.c, pch.h (pch_normalize): New function.
+	* tests/reject-format: New test case.
+	* Makefile.in (TESTS): Add test case.
+	* tests/preserve-c-function-names: Update.
+
 2009-03-19  Andreas Gruenbacher  <agruen@suse.de>
 
 	* Makefile.in (DISTFILES): Add the files from DISTFILES_M4,

Makefile.in

@@ -88,9 +88,9 @@ HDRS = argmatch.h backupfile.h common.h dirname.h \
 MISC = AUTHORS COPYING ChangeLog INSTALL Makefile.in NEWS README VERSION \
 	aclocal.m4 config.hin configure configure.ac install-sh \
 	mkinstalldirs patch.man stdbool_.h timespec.h
-TESTS = tests/corrupt-reject-files tests/no-newline-triggers-assert \
-	tests/preserve-c-function-names \
-	tests/remember-backup-files tests/crlf-handling
+TESTS = tests/corrupt-reject-files tests/crlf-handling \
+	tests/no-newline-triggers-assert tests/preserve-c-function-names \
+	tests/reject-format tests/remember-backup-files
 DISTFILES = $(MISC) $(SRCS) $(HDRS) $(M4FILES) $(TESTS) \
 	pc/chdirsaf.c pc/djgpp/config.sed pc/djgpp/configure.bat \
 	pc/djgpp/configure.sed pc/djgpp/README

NEWS

@@ -1,3 +1,11 @@
+Changes in version 2.6:
+
+* If the input patch is in unified format, any .rej output is now
+  in unified format as well.  Formerly it was in context format.
+  This can be overridden with --reject-format=FORMAT.
+* Changed lines in reject files in context format will now be indicated
+  with '!' as per the definition of the format instead of '+' and '-'.
+
 Changes in versions 2.5.8 and 2.5.9: bug fixes only.
 
 Changes in version 2.5.7:

patch.c

@@ -68,7 +68,6 @@ static bool similar (char const *, size_t, char const *, size_t);
 static bool spew_output (struct outstate *, struct stat *);
 static char const *make_temp (char);
 static int numeric_string (char const *, bool, char const *);
-static void abort_hunk (void);
 static void cleanup (void);
 static void get_some_switches (void);
 static void init_output (char const *, int, struct outstate *);
@@ -77,6 +76,11 @@ static void reinitialize_almost_everything (void);
 static void remove_if_needed (char const *, int volatile *);
 static void usage (FILE *, int) __attribute__((noreturn));
 
+static void abort_hunk (void);
+static void abort_hunk_context (void);
+static void abort_hunk_unified (void);
+
+static enum diff reject_format = NO_DIFF;  /* automatic */
 static bool make_backups;
 static bool backup_if_mismatch;
 static char const *version_control;
@@ -527,6 +531,7 @@ static struct option const longopts[] =
   {"no-backup-if-mismatch", no_argument, NULL, CHAR_MAX + 6},
   {"posix", no_argument, NULL, CHAR_MAX + 7},
   {"quoting-style", required_argument, NULL, CHAR_MAX + 8},
+  {"reject-format", required_argument, NULL, CHAR_MAX + 9},
   {NULL, no_argument, NULL, 0}
 };
 
@@ -587,6 +592,7 @@ static char const *const option_help[] =
 "  --posix  Conform to the POSIX standard.",
 "",
 "  -d DIR  --directory=DIR  Change the working directory to DIR first.",
+"  --reject-format=FORMAT  Create 'context' or 'unified' rejects.",
 #if HAVE_SETMODE_DOS
 "  --binary  Read and write data in binary mode.",
 #else
@@ -784,6 +790,14 @@ get_some_switches (void)
 				     (enum quoting_style) i);
 		}
 		break;
+	    case CHAR_MAX + 9:
+		if (strcmp (optarg, "context") == 0)
+		  reject_format = NEW_CONTEXT_DIFF;
+		else if (strcmp (optarg, "unified") == 0)
+		  reject_format = UNI_DIFF;
+		else
+		  usage (stderr, 2);
+		break;
 	    default:
 		usage (stderr, 2);
 	}
@@ -932,10 +946,99 @@ locate_hunk (LINENUM fuzz)
     return 0;
 }
 
-/* We did not find the pattern, dump out the hunk so they can handle it. */
+static void
+mangled_patch (LINENUM old, LINENUM new)
+{
+  char numbuf0[LINENUM_LENGTH_BOUND + 1];
+  char numbuf1[LINENUM_LENGTH_BOUND + 1];
+  if (debug & 1)
+    say ("oldchar = '%c', newchar = '%c'\n",
+        pch_char (old), pch_char (new));
+  fatal ("Out-of-sync patch, lines %s,%s -- mangled text or line numbers, "
+        "maybe?",
+        format_linenum (numbuf0, pch_hunk_beg () + old),
+        format_linenum (numbuf1, pch_hunk_beg () + new));
+}
+
+/* Output a line number range in unified format.  */
 
 static void
-abort_hunk (void)
+print_unidiff_range (FILE *fp, LINENUM start, LINENUM count)
+{
+  char numbuf0[LINENUM_LENGTH_BOUND + 1];
+  char numbuf1[LINENUM_LENGTH_BOUND + 1];
+
+  switch (count)
+    {
+    case 0:
+      fprintf (fp, "%s,0", format_linenum (numbuf0, start - 1));
+      break;
+
+    case 1:
+      fprintf (fp, "%s", format_linenum (numbuf0, start));
+      break;
+
+    default:
+      fprintf (fp, "%s,%s",
+              format_linenum (numbuf0, start),
+              format_linenum (numbuf1, count));
+      break;
+    }
+}
+
+/* Produce unified reject files */
+
+static void
+abort_hunk_unified (void)
+{
+  FILE *fp = rejfp;
+  register LINENUM old = 1;
+  register LINENUM lastline = pch_ptrn_lines ();
+  register LINENUM new = lastline + 1;
+
+  /* Add last_offset to guess the same as the previous successful hunk.  */
+  fprintf (fp, "@@ -");
+  print_unidiff_range (fp, pch_first () + last_offset, lastline);
+  fprintf (fp, " +");
+  print_unidiff_range (fp, pch_newfirst () + last_offset, pch_repl_lines ());
+  fprintf (fp, " @@\n");
+
+  while (pch_char (new) == '=' || pch_char (new) == '\n')
+    new++;
+
+  if (diff_type != UNI_DIFF)
+    pch_normalize (UNI_DIFF);
+
+  for (; ; old++, new++)
+    {
+      for (;  pch_char (old) == '-';  old++)
+	{
+	  fputc ('-', fp);
+	  pch_write_line (old, fp);
+	}
+      for (;  pch_char (new) == '+';  new++)
+	{
+	  fputc ('+', fp);
+	  pch_write_line (new, fp);
+	}
+
+      if (old > lastline)
+	  break;
+
+      if (pch_char (new) != pch_char (old))
+	mangled_patch (old, new);
+
+      fputc (' ', fp);
+      pch_write_line (old, fp);
+    }
+  if (pch_char (new) != '^')
+    mangled_patch (old, new);
+}
+
+/* Output the rejected patch in context format.  */
+
+static void
+abort_hunk_context (void)
 {
     register LINENUM i;
     register LINENUM pat_end = pch_end ();
@@ -950,6 +1053,9 @@ abort_hunk (void)
       (int) NEW_CONTEXT_DIFF <= (int) diff_type ? " ----" : " -----";
     char const *c_function = pch_c_function();
 
+    if (diff_type == UNI_DIFF)
+      pch_normalize (NEW_CONTEXT_DIFF);
+
     fprintf(rejfp, "***************%s\n", c_function ? c_function : "");
     for (i=0; i<=pat_end; i++) {
 	char numbuf0[LINENUM_LENGTH_BOUND + 1];
@@ -984,13 +1090,25 @@ abort_hunk (void)
 	    pch_write_line (i, rejfp);
 	    break;
 	default:
-	    fatal ("fatal internal error in abort_hunk");
+	    fatal ("fatal internal error in abort_hunk_context");
 	}
 	if (ferror (rejfp))
 	  write_fatal ();
     }
 }
 
+/* Output the rejected hunk.  */
+
+static void
+abort_hunk (void)
+{
+  if (reject_format == UNI_DIFF
+      || (reject_format == NO_DIFF && diff_type == UNI_DIFF))
+    abort_hunk_unified ();
+  else
+    abort_hunk_context ();
+}
+
 /* We found where to apply it (we hope), so do it. */
 
 static bool
@@ -1054,16 +1172,8 @@ apply_hunk (struct outstate *outstate, LINENUM where)
 	    outstate->zero_output = false;
 	    new++;
 	}
-	else if (pch_char(new) != pch_char(old)) {
-	    char numbuf0[LINENUM_LENGTH_BOUND + 1];
-	    char numbuf1[LINENUM_LENGTH_BOUND + 1];
-	    if (debug & 1)
-	      say ("oldchar = '%c', newchar = '%c'\n",
-		   pch_char (old), pch_char (new));
-	    fatal ("Out-of-sync patch, lines %s,%s -- mangled text or line numbers, maybe?",
-		   format_linenum (numbuf0, pch_hunk_beg() + old),
-		   format_linenum (numbuf1, pch_hunk_beg() + new));
-	}
+	else if (pch_char(new) != pch_char(old))
+	  mangled_patch (old, new);
 	else if (pch_char(new) == '!') {
 	    assert (outstate->after_newline);
 	    if (! copy_till (outstate, where + old - 1))

patch.man

@@ -111,8 +111,7 @@ would generate a file name that is too long
 makes the file name too long, then
 .B #
 replaces the file name's last character).
-(The rejected hunk comes out in ordinary context diff form regardless of
-the input patch's form.
+(The rejected hunk comes out in unified or context diff format.
 If the input was a normal diff, many of the contexts are simply null.)
 The line numbers on the hunks in the reject file may be different than
 in the patch file: they reflect the approximate location patch thinks the
@@ -546,6 +545,12 @@ Luckily, most patches add or change lines rather than delete them, so most
 reversed normal diffs begin with a delete, which fails, triggering
 the heuristic.)
 .TP
+\fB\*=reject\-format=\fP\fIformat\fP
+Produce reject files in the specified \fIformat\fP (either \fBcontext\fP or
+\fBunified\fP).  Without this option, rejected hunks come out in unified diff
+format if the input patch was of that format, otherwise in ordinary context
+diff form.
+.TP
 \fB\-s\fP  or  \fB\*=silent\fP  or  \fB\*=quiet\fP
 Work silently, unless an error occurs.
 .TP

pch.c

@@ -2054,3 +2054,70 @@ do_ed_script (FILE *ofp)
 	  read_fatal ();
       }
 }
+
+void
+pch_normalize (enum diff format)
+{
+  LINENUM old = 1;
+  LINENUM new = p_ptrn_lines + 1;
+
+  while (p_Char[new] == '=' || p_Char[new] == '\n')
+    new++;
+
+  if (format == UNI_DIFF)
+    {
+      /* Convert '!' markers into '-' and '+' as defined by the Unified
+         Format.  */
+
+      for (; old <= p_ptrn_lines; old++)
+	if (p_Char[old] == '!')
+	  p_Char[old] = '-';
+      for (; new <= p_end; new++)
+	if (p_Char[new] == '!')
+	  p_Char[new] = '+';
+    }
+  else
+    {
+      /* Convert '-' and '+' markers which are part of a group into '!' as
+	 defined by the Context Format.  */
+
+      while (old <= p_ptrn_lines)
+	{
+	  if (p_Char[old] == '-')
+	    {
+	      if (new <= p_end && p_Char[new] == '+')
+		{
+		  do
+		    {
+		      p_Char[old] = '!';
+		      old++;
+		    }
+		  while (old <= p_ptrn_lines && p_Char[old] == '-');
+		  do
+		    {
+		      p_Char[new] = '!';
+		      new++;
+		    }
+		  while (new <= p_end && p_Char[new] == '+');
+		}
+	      else
+		{
+		  do
+		    old++;
+		  while (old <= p_ptrn_lines && p_Char[old] == '-');
+		}
+	    }
+	  else if (new <= p_end && p_Char[new] == '+')
+	    {
+	      do
+		new++;
+	      while (new <= p_end && p_Char[new] == '+');
+	    }
+	  else
+	    {
+	      old++;
+	      new++;
+	    }
+	}
+    }
+}

pch.h

@@ -44,3 +44,4 @@ void do_ed_script (FILE *);
 void open_patch_file (char const *);
 void re_patch (void);
 void set_hunkmax (void);
+void pch_normalize (enum diff);

tests/preserve-c-function-names

@@ -33,7 +33,7 @@ int foo()
   }
 EOF
 
-diff -p -u a b > ab.diff
+diff -p -c a b > ab.diff
 printf "" > c
 check 'patch c < ab.diff || cat c.rej' <<EOF
 patching file c
@@ -44,12 +44,12 @@ Hunk #1 FAILED at 2.
     {
       /* waste a line */
   
--     return 1;
+!     return 1;
     }
 --- 2,6 ----
     {
       /* waste a line */
   
-+     return 2;
+!     return 2;
     }
 EOF

tests/reject-format

@@ -0,0 +1,126 @@
+#! /bin/bash
+
+# Copyright (C) 2009 Free Software Foundation, Inc.
+#
+# Copying and distribution of this file, with or without modification,
+# in any medium, are permitted without royalty provided the copyright
+# notice and this notice are preserved.
+
+# Test the --reject-format=FORMAT option
+
+. $srcdir/test-lib.sh
+
+require_cat
+require_diff
+use_local_patch
+use_tmpdir
+
+# ==============================================================
+
+cat > f.orig <<EOF
+a() {
+2
+3
+
+5
+6
+}
+EOF
+
+sed -e 's/5/5a/' f.orig > f
+diff -p -c f.orig f > f.diff
+
+check 'patch -f -F0 --no-backup-if-mismatch f < f.diff || echo "Status: $?"' <<EOF
+patching file f
+Hunk #1 FAILED at 2.
+1 out of 1 hunk FAILED -- saving rejects to file f.rej
+Status: 1
+EOF
+
+check 'cat f.rej' <<EOF
+*************** a() {
+*** 2,7 ****
+  2
+  3
+  
+! 5
+  6
+  }
+--- 2,7 ----
+  2
+  3
+  
+! 5a
+  6
+  }
+EOF
+
+# --------------------------------------------------------------
+
+diff -p -u f.orig f > f.diff
+check 'patch -f -F0 --no-backup-if-mismatch f < f.diff || echo "Status: $?"' <<EOF
+patching file f
+Hunk #1 FAILED at 2.
+1 out of 1 hunk FAILED -- saving rejects to file f.rej
+Status: 1
+EOF
+
+check 'cat f.rej' <<EOF
+@@ -2,6 +2,6 @@
+ 2
+ 3
+ 
+-5
++5a
+ 6
+ }
+EOF
+
+# --------------------------------------------------------------
+
+diff -p -c f.orig f > f.diff
+check 'patch -f -F0 --no-backup-if-mismatch --reject-format=unified f < f.diff || echo "Status: $?"' <<EOF
+patching file f
+Hunk #1 FAILED at 2.
+1 out of 1 hunk FAILED -- saving rejects to file f.rej
+Status: 1
+EOF
+
+check 'cat f.rej' <<EOF
+@@ -2,6 +2,6 @@
+ 2
+ 3
+ 
+-5
++5a
+ 6
+ }
+EOF
+
+# --------------------------------------------------------------
+
+diff -p -u f.orig f > f.diff
+check 'patch -f -F0 --no-backup-if-mismatch --reject-format=context f < f.diff || echo "Status: $?"' <<EOF
+patching file f
+Hunk #1 FAILED at 2.
+1 out of 1 hunk FAILED -- saving rejects to file f.rej
+Status: 1
+EOF
+
+check 'cat f.rej' <<EOF
+*************** a() {
+*** 2,7 ****
+  2
+  3
+  
+! 5
+  6
+  }
+--- 2,7 ----
+  2
+  3
+  
+! 5a
+  6
+  }
+EOF