Attachment #72589 for bug #104593




  done
  troff=""
  nroff=""
  jnroff=""
  eqn=""
  neqn=""
  jneqn=""
  tbl=""
  col=""
  vgrind=""

  then
    if test $Fnroff = "missing"
    then
      nroff="nroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
    else
      nroff="$Fnroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
    fi
    troff="troff -mandoc"
    echo "Warning: could not find groff"
  else
    if test $Fnroff = "missing"
    then
      nroff="$Fgroff -Tutf8 -mandoc"
    else
      nroff="$Fnroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
    fi
    troff="$Fgroff -Tps -mandoc"
    jnroff="$Fgroff -Tnippon -mandocj"
  fi
  eqn="$Fgeqn -Tps"
  neqn="$Fgeqn -Tutf8"
  jneqn="$Fgeqn -Tnippon"
  tbl="$Fgtbl"
  col="$Fcol"
  vgrind="$Fvgrind"

  cat="$Fcat"
  awk="$Fawk"

  FILTERS="troff nroff eqn neqn tbl col vgrind refer grap pic pager browser htmlpager cmp cat awk"
fi

# Note: older nroff gives an error message for -c

if [ "x$set_compatibility_mode_for_colored_groff" = "xtrue" ]; then
    troff="$troff -c"
    nroff="$nroff -c"
    jnroff="$jnroff -c"
fi

if [ x$default = x ]; then

s,@LIBOBJS@,$LIBOBJS,
s,@troff@,$troff,
s,@nroff@,$nroff,
s,@jnroff@,$jnroff,
s,@eqn@,$eqn,
s,@neqn@,$neqn,
s,@jneqn@,$jneqn,
s,@tbl@,$tbl,
s,@nocol@,$nocol,
s,@pcol@,$pcol,




int fsstnd;
int noautopath;
int nocache;

static char *language;
static char **section_list;

#ifdef DO_COMPRESS

}

static int
is_lang_page (const char *lang, const char *file) {
	char lang_path[16] = "";

	snprintf(lang_path, sizeof(lang_path), "/%s/", lang);

	return 0;
}

/* we need to pass, as a parameter, what the character set of a man page
 * is likely to be if it is not utf-8, so that nroff can iconv it to utf-8.
 *
 * the character set/encoding is "guessed" based on the most common non-Unicode
 * encoding used for man pages.
 */

typedef struct {
    const char *lang;
    const char *encoding;
} lc2enc_t;

static const lc2enc_t lc2enc[] = {
    { "C",     "ASCII"      },
    { "de",    "ISO-8859-1" },
    { "en",    "ISO-8859-1" },
    { "es",    "ISO-8859-1" },
    { "fr",    "ISO-8859-1" },
    { "it",    "ISO-8859-1" },
    { "pt",    "ISO-8859-1" },
    { "hr",    "ISO-8859-2" },
    { "cs",    "ISO-8859-2" },
    { "et",    "ISO-8859-2" },
    { "hu",    "ISO-8859-2" },
    { "lv",    "ISO-8859-2" },
    { "lt",    "ISO-8859-2" },
    { "pl",    "ISO-8859-2" },
    { "ro",    "ISO-8859-2" },
    { "sk",    "ISO-8859-2" },
    { "sl",    "ISO-8859-2" },
    { "bg",    "KOI8-R"     },
    { "be",    "KOI8-R"     },
    { "mk",    "KOI8-R"     },
    { "ru",    "KOI8-R"     },
    { "sh",    "KOI8-R"     },
    { "uk",    "KOI8-R"     },
    { "el",    "ISO-8859-7" },
    { "tr",    "ISO-8859-9" },
    { "ja",    "EUC-JP"     },
    { "ko",    "EUC-KR"     },
    { "zh_CN", "GB2312"     },
    { "zh_HK", "BIG5"       },
    { "zh_TW", "BIG5"       },
    { NULL }
};

static const char *
get_legacy_encoding(const char *file) {
    const lc2enc_t *ptr;

    for (ptr = lc2enc; ptr->lang != NULL; ptr++) {
        if (is_lang_page(ptr->lang, file)) {
            return ptr->encoding;
        }
    }
    return "ISO-8859-1";
}

static int
parse_roff_directive (char *cp, const char *file, char *buf, int buflen) {
     char c;
     int tbl_found = 0;
     int use_jroff;

     use_jroff = (is_japanese &&
		   (strstr(file, "/jman/") || is_lang_page(language, file)));

     while ((c = *cp++) != '\0') {
	  switch (c) {
	  case 'e':
	       if (debug)
		    gripe (FOUND_EQN);
	       add_directive((do_troff ? "EQN" : "NEQN"), file, buf, buflen);
			     file, buf, buflen);
	       break;

	  case 'g':

     if (*buf == 0)
	  return 1;

     add_directive (do_troff ? "TROFF" : "NROFF", "", buf, buflen);
		    "", buf, buflen);

     if (!do_troff && strstr(buf, "NROFF_OLD_CHARSET") != NULL) {
         const char *encoding = NULL;
         size_t len = strlen("NROFF_OLD_CHARSET");
         char *p = strstr(buf, "NROFF_OLD_CHARSET");

         if (debug) {
             fprintf(stderr, "\nfound '%s' in path\n", "NROFF_OLD_CHARSET");
         }
         encoding = get_legacy_encoding(file);
         if (debug) {
             fprintf(stderr, "\nold charset of '%s' is '%s'\n", file, encoding);
         }
         if (strlen(encoding) < len) {
             memmove(p, p + len, strlen(p + len) + 1);
             memmove(p + strlen(encoding), p, strlen(p) + 1);
             memcpy(p, encoding, strlen(encoding));
         }
     }
     
     if (tbl_found && !do_troff && *getval("COL"))
	  add_directive ("COL", "", buf, buflen);


     return status;
}

/* Special code for Japanese (to pick jnroff instead of nroff, etc.) */
static void
setlang(void) {
	char *lang;

	/* We use getenv() instead of setlocale(), because of
	   glibc 2.1.x security policy for SetUID/SetGID binary. */
	if ((lang = getenv("LANG")) == NULL &&
	    (lang = getenv("LC_ALL")) == NULL &&
	    (lang = getenv("LC_CTYPE")) == NULL)
		/* nothing */;

	language = lang;
	is_japanese = (lang && strncmp(lang, "ja", 2) == 0);
}

/*
 * Handle the apropos option.  Cheat by using another program.
 */

     setlocale(LC_MESSAGES, "");
#endif

     /* No doubt we'll need some generic language code here later.
	For the moment only Japanese support. */
     setlang();

     /* Handle /usr/man/man1.Z/name.1 nonsense from HP */
     dohp = getenv("MAN_HP_DIREXT");		/* .Z */





# (Maybe - but today I need -Tlatin1 to prevent double conversion to utf8.)
#
# If you have a new troff (version 1.18.1?) and its colored output
# causes problems, add the -c option to TROFF, NROFF.
#
TROFF		@troff@
NROFF		@nroff@
JNROFF		@jnroff@
EQN		@eqn@
NEQN		@neqn@
JNEQN		@jneqn@
TBL		@tbl@
@nocol@COL		@col@
REFER		@refer@

Lines 16-25 Link Here

(-)man-1.6b/src/paths.h.in.nroff (-2 lines)
16	{ "WHATIS", "@whatis@" },	16	{ "WHATIS", "@whatis@" },
17	{ "TROFF", "@troff@" },	17	{ "TROFF", "@troff@" },
18	{ "NROFF", "@nroff@" },	18	{ "NROFF", "@nroff@" },
19	{ "JNROFF", "@jnroff@" },
20	{ "EQN", "@eqn@" },	19	{ "EQN", "@eqn@" },
21	{ "NEQN", "@neqn@" },	20	{ "NEQN", "@neqn@" },
22	{ "JNEQN", "@jneqn@" },
23	{ "TBL", "@tbl@" },	21	{ "TBL", "@tbl@" },
24	{ "COL", "@pcol@" },	22	{ "COL", "@pcol@" },
25	{ "REFER", "@refer@" },	23	{ "REFER", "@refer@" },

Return to bug 104593

Lines 405-414 Link Here

(-)man-1.6b/configure.nroff (-13 / +6 lines)
405	done	405	done
406	troff=""	406	troff=""
407	nroff=""	407	nroff=""
408	jnroff=""
409	eqn=""	408	eqn=""
410	neqn=""	409	neqn=""
411	jneqn=""
412	tbl=""	410	tbl=""
413	col=""	411	col=""
414	vgrind=""	412	vgrind=""
Lines 467-491 Link Here
467	then	465	then
468	if test $Fnroff = "missing"	466	if test $Fnroff = "missing"
469	then	467	then
470	nroff="nroff -Tlatin1 -mandoc"	468	nroff="nroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
471	else	469	else
472	nroff="$Fnroff -Tlatin1 -mandoc"	470	nroff="$Fnroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
473	fi	471	fi
474	troff="troff -mandoc"	472	troff="troff -mandoc"
475	echo "Warning: could not find groff"	473	echo "Warning: could not find groff"
476	else	474	else
477	if test $Fnroff = "missing"	475	if test $Fnroff = "missing"
478	then	476	then
479	nroff="$Fgroff -Tlatin1 -mandoc"	477	nroff="$Fgroff -Tutf8 -mandoc"
480	else	478	else
481	nroff="$Fnroff -Tlatin1 -mandoc"	479	nroff="$Fnroff --legacy NROFF_OLD_CHARSET -mandoc 2>/dev/null"
482	fi	480	fi
483	troff="$Fgroff -Tps -mandoc"	481	troff="$Fgroff -Tps -mandoc"
484	jnroff="$Fgroff -Tnippon -mandocj"
485	fi	482	fi
486	eqn="$Fgeqn -Tps"	483	eqn="$Fgeqn -Tps"
487	neqn="$Fgeqn -Tlatin1"	484	neqn="$Fgeqn -Tutf8"
488	jneqn="$Fgeqn -Tnippon"
489	tbl="$Fgtbl"	485	tbl="$Fgtbl"
490	col="$Fcol"	486	col="$Fcol"
491	vgrind="$Fvgrind"	487	vgrind="$Fvgrind"
Lines 518-524 Link Here
518	cat="$Fcat"	514	cat="$Fcat"
519	awk="$Fawk"	515	awk="$Fawk"
520		516
521	FILTERS="troff nroff jnroff eqn neqn jneqn tbl col vgrind refer grap pic pager browser htmlpager cmp cat awk"	517	FILTERS="troff nroff eqn neqn tbl col vgrind refer grap pic pager browser htmlpager cmp cat awk"
522	fi	518	fi
523		519
524	# Note: older nroff gives an error message for -c	520	# Note: older nroff gives an error message for -c
Lines 526-532 Link Here
526	if [ "x$set_compatibility_mode_for_colored_groff" = "xtrue" ]; then	522	if [ "x$set_compatibility_mode_for_colored_groff" = "xtrue" ]; then
527	troff="$troff -c"	523	troff="$troff -c"
528	nroff="$nroff -c"	524	nroff="$nroff -c"
529	jnroff="$jnroff -c"
530	fi	525	fi
531		526
532	if [ x$default = x ]; then	527	if [ x$default = x ]; then
Lines 1281-1290 Link Here
1281	s,@LIBOBJS@,$LIBOBJS,	1276	s,@LIBOBJS@,$LIBOBJS,
1282	s,@troff@,$troff,	1277	s,@troff@,$troff,
1283	s,@nroff@,$nroff,	1278	s,@nroff@,$nroff,
1284	s,@jnroff@,$jnroff,
1285	s,@eqn@,$eqn,	1279	s,@eqn@,$eqn,
1286	s,@neqn@,$neqn,	1280	s,@neqn@,$neqn,
1287	s,@jneqn@,$jneqn,
1288	s,@tbl@,$tbl,	1281	s,@tbl@,$tbl,
1289	s,@nocol@,$nocol,	1282	s,@nocol@,$nocol,
1290	s,@pcol@,$pcol,	1283	s,@pcol@,$pcol,

Lines 80-87 Link Here

(-)man-1.6b/src/man.c.nroff (-31 / +81 lines)
80	int fsstnd;	80	int fsstnd;
81	int noautopath;	81	int noautopath;
82	int nocache;	82	int nocache;
83	static int is_japanese;	83
84	static char *language;
85	static char **section_list;	84	static char **section_list;
86		85
87	#ifdef DO_COMPRESS	86	#ifdef DO_COMPRESS
Lines 442-448 Link Here
442	}	441	}
443		442
444	static int	443	static int
445	is_lang_page (char lang, const char file) {	444	is_lang_page (const char lang, const char file) {
446	char lang_path[16] = "";	445	char lang_path[16] = "";
447		446
448	snprintf(lang_path, sizeof(lang_path), "/%s/", lang);	447	snprintf(lang_path, sizeof(lang_path), "/%s/", lang);
Lines 457-478 Link Here
457	return 0;	456	return 0;
458	}	457	}
459		458
		459	/* we need to pass, as a parameter, what the character set of a man page
		460	* is likely to be if it is not utf-8, so that nroff can iconv it to utf-8.
		461	*
		462	* the character set/encoding is "guessed" based on the most common non-Unicode
		463	* encoding used for man pages.
		464	*/
		465
		466	typedef struct {
		467	const char *lang;
		468	const char *encoding;
		469	} lc2enc_t;
		470
		471	static const lc2enc_t lc2enc[] = {
		472	{ "C", "ASCII" },
		473	{ "de", "ISO-8859-1" },
		474	{ "en", "ISO-8859-1" },
		475	{ "es", "ISO-8859-1" },
		476	{ "fr", "ISO-8859-1" },
		477	{ "it", "ISO-8859-1" },
		478	{ "pt", "ISO-8859-1" },
		479	{ "hr", "ISO-8859-2" },
		480	{ "cs", "ISO-8859-2" },
		481	{ "et", "ISO-8859-2" },
		482	{ "hu", "ISO-8859-2" },
		483	{ "lv", "ISO-8859-2" },
		484	{ "lt", "ISO-8859-2" },
		485	{ "pl", "ISO-8859-2" },
		486	{ "ro", "ISO-8859-2" },
		487	{ "sk", "ISO-8859-2" },
		488	{ "sl", "ISO-8859-2" },
		489	{ "bg", "KOI8-R" },
		490	{ "be", "KOI8-R" },
		491	{ "mk", "KOI8-R" },
		492	{ "ru", "KOI8-R" },
		493	{ "sh", "KOI8-R" },
		494	{ "uk", "KOI8-R" },
		495	{ "el", "ISO-8859-7" },
		496	{ "tr", "ISO-8859-9" },
		497	{ "ja", "EUC-JP" },
		498	{ "ko", "EUC-KR" },
		499	{ "zh_CN", "GB2312" },
		500	{ "zh_HK", "BIG5" },
		501	{ "zh_TW", "BIG5" },
		502	{ NULL }
		503	};
		504
		505	static const char *
		506	get_legacy_encoding(const char *file) {
		507	const lc2enc_t *ptr;
		508
		509	for (ptr = lc2enc; ptr->lang != NULL; ptr++) {
		510	if (is_lang_page(ptr->lang, file)) {
		511	return ptr->encoding;
		512	}
		513	}
		514	return "ISO-8859-1";
		515	}
		516
460	static int	517	static int
461	parse_roff_directive (char cp, const char file, char *buf, int buflen) {	518	parse_roff_directive (char cp, const char file, char *buf, int buflen) {
462	char c;	519	char c;
463	int tbl_found = 0;	520	int tbl_found = 0;
464	int use_jroff;
465
466	use_jroff = (is_japanese &&
467	(strstr(file, "/jman/") \|\| is_lang_page(language, file)));
468		521
469	while ((c = *cp++) != '\0') {	522	while ((c = *cp++) != '\0') {
470	switch (c) {	523	switch (c) {
471	case 'e':	524	case 'e':
472	if (debug)	525	if (debug)
473	gripe (FOUND_EQN);	526	gripe (FOUND_EQN);
474	add_directive((do_troff ? "EQN" : use_jroff ? "JNEQN": "NEQN"),	527	add_directive((do_troff ? "EQN" : "NEQN"), file, buf, buflen);
475	file, buf, buflen);
476	break;	528	break;
477		529
478	case 'g':	530	case 'g':
Lines 520-528 Link Here
520	if (*buf == 0)	572	if (*buf == 0)
521	return 1;	573	return 1;
522		574
523	add_directive (do_troff ? "TROFF" : use_jroff ? "JNROFF" : "NROFF",	575	add_directive (do_troff ? "TROFF" : "NROFF", "", buf, buflen);
524	"", buf, buflen);
525		576
		577	if (!do_troff && strstr(buf, "NROFF_OLD_CHARSET") != NULL) {
		578	const char *encoding = NULL;
		579	size_t len = strlen("NROFF_OLD_CHARSET");
		580	char *p = strstr(buf, "NROFF_OLD_CHARSET");
		581
		582	if (debug) {
		583	fprintf(stderr, "\nfound '%s' in path\n", "NROFF_OLD_CHARSET");
		584	}
		585	encoding = get_legacy_encoding(file);
		586	if (debug) {
		587	fprintf(stderr, "\nold charset of '%s' is '%s'\n", file, encoding);
		588	}
		589	if (strlen(encoding) < len) {
		590	memmove(p, p + len, strlen(p + len) + 1);
		591	memmove(p + strlen(encoding), p, strlen(p) + 1);
		592	memcpy(p, encoding, strlen(encoding));
		593	}
		594	}
		595
526	if (tbl_found && !do_troff && *getval("COL"))	596	if (tbl_found && !do_troff && *getval("COL"))
527	add_directive ("COL", "", buf, buflen);	597	add_directive ("COL", "", buf, buflen);
528		598
Lines 1181-1202 Link Here
1181	return status;	1251	return status;
1182	}	1252	}
1183		1253
1184	/* Special code for Japanese (to pick jnroff instead of nroff, etc.) */
1185	static void
1186	setlang(void) {
1187	char *lang;
1188
1189	/* We use getenv() instead of setlocale(), because of
1190	glibc 2.1.x security policy for SetUID/SetGID binary. */
1191	if ((lang = getenv("LANG")) == NULL &&
1192	(lang = getenv("LC_ALL")) == NULL &&
1193	(lang = getenv("LC_CTYPE")) == NULL)
1194	/* nothing */;
1195
1196	language = lang;
1197	is_japanese = (lang && strncmp(lang, "ja", 2) == 0);
1198	}
1199
1200	/*	1254	/*
1201	* Handle the apropos option. Cheat by using another program.	1255	* Handle the apropos option. Cheat by using another program.
1202	*/	1256	*/
Lines 1263-1272 Link Here
1263	setlocale(LC_MESSAGES, "");	1317	setlocale(LC_MESSAGES, "");
1264	#endif	1318	#endif
1265		1319
1266	/* No doubt we'll need some generic language code here later.
1267	For the moment only Japanese support. */
1268	setlang();
1269
1270	/* Handle /usr/man/man1.Z/name.1 nonsense from HP */	1320	/* Handle /usr/man/man1.Z/name.1 nonsense from HP */
1271	dohp = getenv("MAN_HP_DIREXT"); /* .Z */	1321	dohp = getenv("MAN_HP_DIREXT"); /* .Z */
1272		1322

Lines 86-99 Link Here

(-)man-1.6b/src/man.conf.in.nroff (-3 / +1 lines)
86	# (Maybe - but today I need -Tlatin1 to prevent double conversion to utf8.)	86	# (Maybe - but today I need -Tlatin1 to prevent double conversion to utf8.)
87	#	87	#
88	# If you have a new troff (version 1.18.1?) and its colored output	88	# If you have a new troff (version 1.18.1?) and its colored output
89	# causes problems, add the -c option to TROFF, NROFF, JNROFF.	89	# causes problems, add the -c option to TROFF, NROFF.
90	#	90	#
91	TROFF @troff@	91	TROFF @troff@
92	NROFF @nroff@	92	NROFF @nroff@
93	JNROFF @jnroff@
94	EQN @eqn@	93	EQN @eqn@
95	NEQN @neqn@	94	NEQN @neqn@
96	JNEQN @jneqn@
97	TBL @tbl@	95	TBL @tbl@
98	@nocol@COL @col@	96	@nocol@COL @col@
99	REFER @refer@	97	REFER @refer@