Integrate preconv into the groff binary. Changes based on a patch

from Michail Vidiassov.

* src/roff/groff/pipeline.h (MAX_COMMANDS): Increase by 1.

* src/roff/groff/groff.cpp (PRECONV_INDEX): New macro.
(SOELIM_INDEX): Updated.
(help, synopsis): Updated.
(main): Add command line options `-k' and `-K enc' to select
encoding.
Add support for GROFF_ENCODING environment variable.

* src/roff/groff/groff.man: Updated and revised.


Start with autoconf support for preconv.

* configure.ac: Call AC_C_BIGENDIAN.
* configure, src/include/config.hin: Regenerated.

3 files changed, 177 insertions, 52 deletions

diff --git a/src/roff/groff/groff.cpp b/src/roff/groff/groff.cpp
index 5e187573..d8cbea1c 100644
--- a/src/roff/groff/groff.cpp
+++ b/src/roff/groff/groff.cpp
@@ -1,5 +1,5 @@
 // -*- C++ -*-
-/* Copyright (C) 1989-2000, 2001, 2002, 2003, 2004
+/* Copyright (C) 1989-2000, 2001, 2002, 2003, 2004, 2005
    Free Software Foundation, Inc.
      Written by James Clark (jjc@jclark.com)
 
@@ -51,7 +51,8 @@ extern "C" {
 #endif /* NEED_DECLARATION_PUTENV */
 
 // The number of commands must be in sync with MAX_COMMANDS in pipeline.h
-const int SOELIM_INDEX = 0;
+const int PRECONV_INDEX = 0;
+const int SOELIM_INDEX = PRECONV_INDEX + 1;
 const int REFER_INDEX = SOELIM_INDEX + 1;
 const int GRAP_INDEX = REFER_INDEX + 1;
 const int PIC_INDEX = GRAP_INDEX + 1;
@@ -110,6 +111,7 @@ int main(int argc, char **argv)
   setbuf(stderr, stderr_buf);
   assert(NCOMMANDS <= MAX_COMMANDS);
   string Pargs, Largs, Fargs;
+  int Kflag = 0;
   int vflag = 0;
   int Vflag = 0;
   int zflag = 0;
@@ -119,6 +121,7 @@ int main(int argc, char **argv)
   int safer_flag = 1;
   int opt;
   const char *command_prefix = getenv("GROFF_COMMAND_PREFIX");
+  const char *encoding = getenv("GROFF_ENCODING");
   if (!command_prefix)
     command_prefix = PROG_PREFIX;
   commands[TROFF_INDEX].set_name(command_prefix, "troff");
@@ -127,9 +130,10 @@ int main(int argc, char **argv)
     { "version", no_argument, 0, 'v' },
     { NULL, 0, 0, 0 }
   };
-  while ((opt = getopt_long(argc, argv,
-			    "abcCd:eEf:F:gGhiI:lL:m:M:n:No:pP:r:RsStT:UvVw:W:XzZ",
-			    long_options, NULL))
+  while ((opt = getopt_long(
+		  argc, argv,
+		  "abcCd:eEf:F:gGhiI:lkK:L:m:M:n:No:pP:r:RsStT:UvVw:W:XzZ",
+		  long_options, NULL))
 	 != EOF) {
     char buf[3];
     buf[0] = '-';
@@ -149,6 +153,13 @@ int main(int argc, char **argv)
       Pargs += optarg;
       Pargs += '\0';
       break;
+    case 'K':
+      commands[PRECONV_INDEX].append_arg("-e", optarg);
+      Kflag = 1;
+      // fall through
+    case 'k':
+      commands[PRECONV_INDEX].set_name(command_prefix, "preconv");
+      break;
     case 't':
       commands[TBL_INDEX].set_name(command_prefix, "tbl");
       break;
@@ -185,16 +196,15 @@ int main(int argc, char **argv)
       break;
     case 'v':
       vflag = 1;
-      {
-	printf("GNU groff version %s\n", Version_string);
-	printf("Copyright (C) 2004 Free Software Foundation, Inc.\n"
-	       "GNU groff comes with ABSOLUTELY NO WARRANTY.\n"
-	       "You may redistribute copies of groff and its subprograms\n"
-	       "under the terms of the GNU General Public License.\n"
-	       "For more information about these matters, see the file named COPYING.\n");
-	printf("\ncalled subprograms:\n\n");
-        fflush(stdout);
-      }
+      printf("GNU groff version %s\n", Version_string);
+      printf(
+	"Copyright (C) 2005 Free Software Foundation, Inc.\n"
+	"GNU groff comes with ABSOLUTELY NO WARRANTY.\n"
+	"You may redistribute copies of groff and its subprograms\n"
+	"under the terms of the GNU General Public License.\n"
+	"For more information about these matters, see the file named COPYING.\n");
+      printf("\ncalled subprograms:\n\n");
+      fflush(stdout);
       commands[POST_INDEX].append_arg(buf);
       // fall through
     case 'C':
@@ -284,6 +294,11 @@ int main(int argc, char **argv)
       break;
     }
   }
+  if (encoding) {
+    commands[PRECONV_INDEX].set_name(command_prefix, "preconv");
+    if (!Kflag && *encoding)
+      commands[PRECONV_INDEX].append_arg("-e", encoding);
+  }
   if (safer_flag)
     commands[PIC_INDEX].append_arg("-S");
   else
@@ -309,7 +324,8 @@ int main(int argc, char **argv)
   }
   if (postdriver)
     commands[POST_INDEX].set_name(postdriver);
-  int gxditview_flag = postdriver && strcmp(xbasename(postdriver), GXDITVIEW) == 0;
+  int gxditview_flag = postdriver
+		       && strcmp(xbasename(postdriver), GXDITVIEW) == 0;
   if (gxditview_flag && argc - optind == 1) {
     commands[POST_INDEX].append_arg("-title");
     commands[POST_INDEX].append_arg(argv[optind]);
@@ -690,9 +706,9 @@ char **possible_command::get_argv()
 void synopsis(FILE *stream)
 {
   fprintf(stream,
-"usage: %s [-abceghilpstvzCENRSUVXZ] [-Fdir] [-mname] [-Tdev] [-ffam]\n"
+"usage: %s [-abceghiklpstvzCENRSUVXZ] [-Fdir] [-mname] [-Tdev] [-ffam]\n"
 "       [-wname] [-Wname] [-Mdir] [-dcs] [-rcn] [-nnum] [-olist] [-Parg]\n"
-"       [-Larg] [-Idir] [files...]\n",
+"       [-Karg] [-Larg] [-Idir] [files...]\n",
 	  program_name);
 }
 
@@ -701,6 +717,7 @@ void help()
   synopsis(stdout);
   fputs("\n"
 "-h\tprint this message\n"
+"-k\tpreprocess with preconv\n"
 "-t\tpreprocess with tbl\n"
 "-p\tpreprocess with pic\n"
 "-e\tpreprocess with eqn\n"
@@ -737,6 +754,7 @@ void help()
 "-S\tenable safer mode (the default)\n"
 "-U\tenable unsafe mode\n"
 "-Idir\tsearch dir for soelim, troff, and grops.  Implies -s\n"
+"-Karg\tuse arg as input encoding.  Implies -k\n"
 "\n",
 	stdout);
   exit(0);
diff --git a/src/roff/groff/groff.man b/src/roff/groff/groff.man
index 9e6a4ace..714e8670 100644
--- a/src/roff/groff/groff.man
+++ b/src/roff/groff/groff.man
@@ -1,7 +1,7 @@
 .ig
 groff.man
 
-Last update: 01 Jul 2005
+Last update: 30 Dec 2005
 
 Copyright (C) 1989, 2002, 2003, 2004, 2005 Free Software Foundation, Inc.
 Rewritten in 2002 by Bernd Warken <bwarken@mayn.de>
@@ -52,15 +52,15 @@ FDL in the main directory of the groff source package.
 .\" --------------------------------------------------------------------
 .\" Begin of macro definitions
 .de c
-.\" this is like a comment request when escape mechanism is off
+.  \" this is like a comment request when escape mechanism is off
 ..
 .eo
 .
 .c --------------------------------------------------------------------
 .de TP+
-.br
-.ns
-.TP \$1
+.  br
+.  ns
+.  TP \$1
 ..
 .c --------------------------------------------------------------------
 .c Like TP, but if specified indent is more than half
@@ -241,11 +241,12 @@ groff \- front-end for the groff document formatting system
 .
 .ad l
 .Synopsis groff
-.ShortOpt[] abcegilpstzCEGNRSUVXZ
+.ShortOpt[] abcegiklpstzCEGNRSUVXZ
 .ShortOpt[] d cs
 .ShortOpt[] f fam
 .ShortOpt[] F dir
 .ShortOpt[] I dir
+.ShortOpt[] K arg
 .ShortOpt[] L arg
 .ShortOpt[] m name
 .ShortOpt[] M dir
@@ -388,13 +389,35 @@ and
 escapes).
 The current directory is always searched first.
 This option may be specified more than once;
-the directories will be searched in the order specified.
+the directories are searched in the order specified.
 No directory search is performed for files specified using an absolute path.
 This option implies the
 .ShortOpt s
 option.
 .
 .
+.OptDef k
+Preprocess with
+.BR preconv .
+This is run before any other preprocessor.
+.
+Please refer to
+.BR preconv 's
+manual page for its behaviour if no
+.ShortOpt K
+option is specified.
+.
+.
+.OptDef K "" arg
+Set input encoding used by
+.B preconv
+to
+.IR arg .
+.
+Implies
+.ShortOpt k .
+.
+.
 .OptDef l
 Send the output to a spooler program for printing.
 .
@@ -627,7 +650,7 @@ Output the pipeline that would be run by
 .BR groff
 (as a wrapper program) on the standard output, but do not execute it.
 If given more than once,
-the commands will be both printed on the standard error and run.
+the commands are both printed on the standard error and run.
 .
 .
 .OptDef X
@@ -674,7 +697,7 @@ for example
 .OptDef z
 Suppress output generated by
 .BR @g@troff .
-Only error messages will be printed.
+Only error messages are printed.
 .
 .
 .OptDef Z
@@ -873,35 +896,47 @@ program is an allround-viewer for groff files and man pages.
 The groff preprocessors are reimplementations of the classical
 preprocessors with moderate extensions.
 .
-The preprocessors distributed with the
+The standard preprocessors distributed with the
 .I groff
 package are
 .
 .TP
 .BR @g@eqn (@MAN1EXT@)
 for mathematical formul\(ae,
+.
 .TP
 .BR @g@grn (@MAN1EXT@)
 for including
 .BR gremlin (1)
 pictures,
+.
 .TP
 .BR @g@pic (@MAN1EXT@)
 for drawing diagrams,
+.
 .TP
 .BR \%@g@refer (@MAN1EXT@)
 for bibliographic references,
+.
 .TP
 .BR \%@g@soelim (@MAN1EXT@)
 for including macro files from standard locations,
 .
 .P
 and
+.
 .TP
 .BR @g@tbl (@MAN1EXT@)
 for tables.
 .
 .P
+A new preprocessor not available in classical troff is
+.BR \%preconv (@MAN1EXT@)
+which converts various input encodings to something groff can understand.
+.
+It is always run first before any other preprocessor.
+.
+.P
 Besides these, there are some internal preprocessors that are
 automatically run with some devices.
 .
@@ -1062,7 +1097,7 @@ see
 .
 In roff, the output targets are called
 .IR devices .
-A device can be a piece of hardware, e.g. a printer, or a software
+A device can be a piece of hardware, e.g., a printer, or a software
 file format.
 .
 A device is specified by the option
@@ -1077,7 +1112,7 @@ character set.
 .
 .TP
 .B cp1047
-Text output using the EBCDIC code page IBM cp1047 (e.g. OS/390 Unix).
+Text output using the EBCDIC code page IBM cp1047 (e.g., OS/390 Unix).
 .
 .TP
 .B dvi
@@ -1117,6 +1152,7 @@ encoding; see
 .BR \%xditview (1x)
 and
 .BR \%gxditview (@MAN1EXT@).
+.
 A variant for a 12\|pt document base font is
 .BR \%X75-12 .
 .
@@ -1126,6 +1162,7 @@ A variant for a 12\|pt document base font is
 .BR \%xditview (1x)
 and
 .BR \%gxditview (@MAN1EXT@).
+.
 A variant for a 12\|pt document base font is
 .BR \%X100-12 .
 .
@@ -1153,12 +1190,14 @@ groff provides 3\~hardware postprocessors:
 .TP
 .BR \%grolbp (@MAN1EXT@)
 for some Canon printers,
+.
 .TP
 .BR \%grolj4 (@MAN1EXT@)
 for printers compatible to the HP LaserJet\~4 and PCL5,
+.
 .TP
 .BR \%grotty (@MAN1EXT@)
-for text output using various encodings, e.g. on text-oriented
+for text output using various encodings, e.g., on text-oriented
 terminals or line-printers.
 .
 .P
@@ -1176,9 +1215,11 @@ formats are
 .TP
 .BR \%grodvi (@MAN1EXT@)
 for the DVI format,
+.
 .TP
 .BR \%grohtml (@MAN1EXT@)
 for HTML format,
+.
 .TP
 .BR grops (@MAN1EXT@)
 for PostScript.
@@ -1204,6 +1245,22 @@ Add information to troff font description files for use with groff.
 Create font description files for PostScript device.
 .
 .TP
+.BR \%eqn2graph (@MAN1EXT@)
+Convert an
+.B eqn
+image into a cropped image.
+.
+.TP
+.BR \%gdiffmk (@MAN1EXT@)
+Mark differences between groff, nroff, or troff files.
+.
+.TP
+.BR \%grap2graph (@MAN1EXT@)
+Convert a
+.B grap
+diagram into a cropped bitmap image.
+.
+.TP
 .BR \%groffer (@MAN1EXT@)
 General viewer program for groff files and man pages.
 .
@@ -1216,7 +1273,7 @@ The groff X viewer, the GNU version of xditview.
 Create font description files for lj4 device.
 .
 .TP
-.BR \%indxbib (@MAN1EXT@)
+.BR \%@g@indxbib (@MAN1EXT@)
 Make inverted index for bibliographic databases.
 .
 .TP
@@ -1224,14 +1281,25 @@ Make inverted index for bibliographic databases.
 Search bibliographic databases.
 .
 .TP
-.BR \%lookbib (@MAN1EXT@)
+.BR \%@g@lookbib (@MAN1EXT@)
 Interactively search bibliographic databases.
 .
 .TP
+.BR \%pdfroff (@MAN1EXT@)
+Create PDF documents using
+.BR groff .
+.
+.TP
 .BR \%pfbtops (@MAN1EXT@)
 Translate a PostScript font in .pfb format to ASCII.
 .
 .TP
+.BR \%pic2graph (@MAN1EXT@)
+Convert a
+.B pic
+diagram into a cropped image.
+.
+.TP
 .BR \%tfmtodit (@MAN1EXT@)
 Create font description files for TeX DVI device.
 .
@@ -1239,6 +1307,10 @@ Create font description files for TeX DVI device.
 .BR \%xditview (1x)
 roff viewer distributed with X window.
 .
+.TP
+.BR \%xtotroff (@MAN1EXT@)
+Convert X font metrics into GNU troff font metrics.
+.
 .
 .\" --------------------------------------------------------------------
 .SH ENVIRONMENT
@@ -1249,17 +1321,19 @@ colon; this may vary depending on the operating system.
 .
 For example, DOS and Windows use a semicolon instead.
 .
+.
 .TP
 .EnvVar GROFF_BIN_PATH
 This search path, followed by
 .EnvVar $PATH ,
-will be used for commands that are executed by
+is used for commands that are executed by
 .BR groff .
 .
 If it is not set then the directory where the groff binaries were
 installed is prepended to
 .EnvVar PATH .
 .
+.
 .TP
 .EnvVar GROFF_COMMAND_PREFIX
 When there is a need to run different roff implementations at the same
@@ -1292,7 +1366,7 @@ More exactly, if it is set to prefix
 .I xxx
 then
 .B groff
-as a wrapper program will internally call
+as a wrapper program internally calls
 .IB xxx troff
 instead of
 .BR troff .
@@ -1315,6 +1389,37 @@ itself) since they are unique to the groff package.
 .
 .
 .TP
+.EnvVar GROFF_ENCODING
+The value of this environment value is passed to the
+.B preconv
+preprocessor to select the encoding of input files.
+.
+Setting this option implies
+.BR groff 's
+command line option
+.ShortOpt k
+(this is,
+.B groff
+actually always calls
+.BR preconv ).
+.
+If set without a value,
+.B groff
+calls
+.B preconv
+without arguments.
+.
+An explicit
+.ShortOpt K
+command line option overrides the value of 
+.EnvVar GROFF_ENCODING .
+.
+See
+.BR preconv (@MAN1EXT@)
+for details.
+.
+.
+.TP
 .EnvVar GROFF_FONT_PATH
 A list of directories in which to search for the
 .BI dev name
@@ -1341,11 +1446,11 @@ for more details.
 .
 .TP
 .EnvVar GROFF_TMPDIR
-The directory in which temporary files will be created.
+The directory in which temporary files are created.
 .
 If this is not set but the environment variable
 .EnvVar TMPDIR
-instead, temporary files will be created in the directory
+instead, temporary files are created in the directory
 .EnvVar $TMPDIR .
 On MS-DOS and Windows\ 32 platforms, the environment variables
 .EnvVar TMP
@@ -1356,7 +1461,7 @@ and
 and
 .EnvVar TMPDIR .
 .
-Otherwise, temporary files will be created in
+Otherwise, temporary files are created in
 .BR /tmp .
 The
 .BR \%@g@refer (@MAN1EXT@),
@@ -1422,8 +1527,9 @@ before reading the macro sets and any input.
 .
 .TP
 .B troffrc-end
-Final startup file for troff, it is parsed after all macro sets have
-been read.
+Final startup file for troff.
+.
+It is parsed after all macro sets have been read.
 .
 .
 .TP
@@ -1517,7 +1623,7 @@ calling
 .\" --------------------------------------------------------------------
 .
 .P
-On EBCDIC hosts (e.g. OS/390 Unix), output devices
+On EBCDIC hosts (e.g., OS/390 Unix), output devices
 .B ascii
 and
 .B latin1
@@ -1530,7 +1636,7 @@ is not available on ASCII based operating systems.
 .P
 Report bugs to bug-groff@gnu.org.
 .
-Include a complete, self-contained example that will allow the bug to
+Include a complete, self-contained example that allows the bug to
 be reproduced, and say which version of groff you are using.
 .
 .
@@ -1621,10 +1727,8 @@ Free Documentation License (FDL).
 .
 The
 .IR "groff info file"
-contains all information on the groff system within a single document.
-.
-Beneath the detailed documentation of all aspects, it provides
-examples and background information.
+contains all information on the groff system within a single document,
+providing many examples and background information.
 .
 See
 .BR info (1)
@@ -1658,6 +1762,7 @@ Roff preprocessors:
 .BR \%@g@eqn (@MAN1EXT@),
 .BR \%@g@grn (@MAN1EXT@),
 .BR \%@g@pic (@MAN1EXT@),
+.BR \%preconv (@MAN1EXT@),
 .BR \%@g@refer (@MAN1EXT@),
 .BR \%@g@soelim (@MAN1EXT@),
 .BR \%@g@tbl (@MAN1EXT@),
@@ -1677,9 +1782,7 @@ Roff formatter programs:
 .BR ditroff (@MAN7EXT@).
 .
 .TP
-The
-.I intermediate output
-language:
+The intermediate output language:
 .BR \%groff_out (@MAN7EXT@).
 .
 .TP
@@ -1711,15 +1814,19 @@ The following utilities are available:
 .BR \%addftinfo (@MAN1EXT@),
 .BR \%afmtodit (@MAN1EXT@),
 .BR \%eqn2graph (@MAN1EXT@),
+.BR \%gdiffmk (@MAN1EXT@),
 .BR \%grap2graph (@MAN1EXT@),
 .BR \%groffer (@MAN1EXT@),
 .BR \%gxditview (@MAN1EXT@),
 .BR \%hpftodit (@MAN1EXT@),
 .BR \%@g@indxbib (@MAN1EXT@),
+.BR \%lkbib (@MAN1EXT@),
 .BR \%@g@lookbib (@MAN1EXT@),
+.BR \%pdfroff (@MAN1EXT@),
 .BR \%pfbtops (@MAN1EXT@),
 .BR \%pic2graph (@MAN1EXT@),
-.BR \%tfmtodit (@MAN1EXT@).
+.BR \%tfmtodit (@MAN1EXT@),
+.BR \%xtotroff (@MAN1EXT@).
 .
 .cp \n[groff_C]
 .
diff --git a/src/roff/groff/pipeline.h b/src/roff/groff/pipeline.h
index 77bb11a0..81bb0af6 100644
--- a/src/roff/groff/pipeline.h
+++ b/src/roff/groff/pipeline.h
@@ -1,4 +1,4 @@
-/* Copyright (C) 1989, 1990, 1991, 1992, 2000, 2002
+/* Copyright (C) 1989, 1990, 1991, 1992, 2000, 2002, 2005
    Free Software Foundation, Inc.
      Written by James Clark (jjc@jclark.com)
 
@@ -25,7 +25,7 @@ extern "C" {
 #endif
 
 /* run_pipeline can handle at most this many commands */
-#define MAX_COMMANDS 12
+#define MAX_COMMANDS 13
 
 /* Children exit with this status if execvp fails. */
 #define EXEC_FAILED_EXIT_STATUS 0xff