Simplify and improve parser.

Clarify manpage.

3 files changed, 38 insertions, 244 deletions

diff --git a/usr.bin/colldef/colldef.1 b/usr.bin/colldef/colldef.1
index 4c44cf7..d7e7998 100644
--- a/usr.bin/colldef/colldef.1
+++ b/usr.bin/colldef/colldef.1
@@ -124,11 +124,7 @@ representation, and can be only one character in length.
 .Pp
 Symbol names cannot be specified in
 .Ar substitute
-fields. Symbol names also cannot be combined with
-any other representation, such as, <c>h, c<h>,
-<c>\ex68, or <c><h>. Symbol names can be used with
-primary and secondary ordering as in the following
-example.
+fields.
 .Pp
 The
 .Ar charmap
@@ -153,14 +149,15 @@ statement is optional.
 .Ar order order_list
 .Pp
 .Ar order_list
-is a list of symbols, separated by semi colons, that defines the collating sequence. The
+is a list of symbols, separated by semi colons, that defines the
+collating sequence. The
 special symbol
 .Ar ...
 specifies, in a short-hand
 form, symbols that are sequential in machine code
 order.
 .Pp
-A symbol can be up to two characters in length and
+An order list element
 can be represented in any one of the following
 ways:
 .Bl -bullet
@@ -208,18 +205,33 @@ Symbols
 .Ar \er ,
 .Ar \ev
 are permitted in its usual C-language meaning
+.It
+The symbol range (for example,
+.Ar a;...;z )
+.It
+Comma-separated symbols, ranges and chains enclosed in parenthesis (for example
+.Ar \&(
+.Ar sym1 ,
+.Ar sym2 ,
+.Ar ...
+.Ar \&) )
+are assigned the
+same primary ordering but different secondary
+ordering.
+.It
+Comma-separated symbols, ranges and chains enclosed in curly brackets (for example
+.Ar \&{
+.Ar sym1 ,
+.Ar sym2 ,
+.Ar ...
+.Ar \&} )
+are assigned the same primary ordering only.
 .El
 .Pp
 The backslash character
 .Ar \e
 is used for continuation. In this case, no characters are permitted
-after the backslash character. And as a quotation mark.
-.Pp
-Symbols enclosed in parentheses are assigned the
-same primary ordering but different secondary
-ordering. Symbols enclosed in curly brackets are
-assigned only the same primary ordering
-and different secondary  ordering.
+after the backslash character.
 .Sh EXIT STATUS
 .Ar colldef
 exits with the following values:
diff --git a/usr.bin/colldef/parse.y b/usr.bin/colldef/parse.y
index 55067c5..5823689 100644
--- a/usr.bin/colldef/parse.y
+++ b/usr.bin/colldef/parse.y
@@ -61,7 +61,6 @@ char *out_file = "LC_COLLATE";
 }
 %token SUBSTITUTE WITH ORDER RANGE
 %token <str> STRING
-%token <str> NAME
 %token <str> CHAIN
 %token <str> DEFN
 %token <ch> CHAR
@@ -117,17 +116,6 @@ item : CHAR { __collate_char_pri_table[$1].prim = prim_pri++; }
 	strcpy(__collate_chain_pri_table[chain_index].str, $1);
 	__collate_chain_pri_table[chain_index++].prim = prim_pri++;
 }
-	| NAME {
-	u_int i;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0)
-			goto findi;
-	yyerror("Name <%s> not defined", $1);
-	findi:
-
-	__collate_char_pri_table[i].prim = prim_pri++;
-}
 	| CHAR RANGE CHAR {
 	u_int i;
 
@@ -137,68 +125,6 @@ item : CHAR { __collate_char_pri_table[$1].prim = prim_pri++; }
 	for (i = $1; i <= $3; i++)
 		__collate_char_pri_table[(u_char)i].prim = prim_pri++;
 }
-	| NAME RANGE CHAR {
-	u_int i, c1;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto find1;
-		}
-	yyerror("Name <%s> not defined", $1);
-	find1:
-
-	if ($3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, $3);
-
-	for (i = c1; i <= $3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri++;
-}
-	| CHAR RANGE NAME {
-	u_int i, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto find3;
-		}
-	yyerror("Name <%s> not defined", $3);
-	find3:
-
-	if (c3 <= $1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			$1, c3);
-
-	for (i = $1; i <= c3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri++;
-}
-	| NAME RANGE NAME {
-	u_int i, c1, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto find21;
-		}
-	yyerror("Name <%s> not defined", $1);
-	find21:
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto find23;
-		}
-	yyerror("Name <%s> not defined", $3);
-	find23:
-
-	if (c3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, c3);
-
-	for (i = c1; i <= c3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri++;
-}
 	| '{' prim_order_list '}' {
 	prim_pri++;
 }
@@ -226,79 +152,6 @@ prim_sub_item : CHAR {
 	for (i = $1; i <= $3; i++)
 		__collate_char_pri_table[(u_char)i].prim = prim_pri;
 }
-	| NAME RANGE CHAR {
-	u_int i, c1;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto findpsi1;
-		}
-	yyerror("Name <%s> not defined", $1);
-	findpsi1:
-
-	if ($3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, $3);
-
-	for (i = c1; i <= $3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-}
-	| CHAR RANGE NAME {
-	u_int i, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto findpsi3;
-		}
-	yyerror("Name <%s> not defined", $3);
-	findpsi3:
-
-	if (c3 <= $1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			$1, c3);
-
-	for (i = $1; i <= c3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-}
-	| NAME RANGE NAME {
-	u_int i, c1, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto findpsi21;
-		}
-	yyerror("Name <%s> not defined", $1);
-	findpsi21:
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto findpsi23;
-		}
-	yyerror("Name <%s> not defined", $3);
-	findpsi23:
-
-	if (c3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, c3);
-
-	for (i = c1; i <= c3; i++)
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-}
-	| NAME {
-	u_int i;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0)
-			goto findpsi;
-	yyerror("Name <%s> not defined", $1);
-	findpsi:
-
-	__collate_char_pri_table[i].prim = prim_pri;
-}
 	| CHAIN {
 	if (chain_index >= TABLE_SIZE - 1)
 		yyerror("__collate_chain_pri_table overflow");
@@ -322,86 +175,6 @@ sec_sub_item : CHAR {
 		__collate_char_pri_table[(u_char)i].sec = sec_pri++;
 	}
 }
-	| NAME RANGE CHAR {
-	u_int i, c1;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto findssi1;
-		}
-	yyerror("Name <%s> not defined", $1);
-	findssi1:
-
-	if ($3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, $3);
-
-	for (i = c1; i <= $3; i++) {
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-		__collate_char_pri_table[(u_char)i].sec = sec_pri++;
-	}
-}
-	| CHAR RANGE NAME {
-	u_int i, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto findssi3;
-		}
-	yyerror("Name <%s> not defined", $3);
-	findssi3:
-
-	if (c3 <= $1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			$1, c3);
-
-	for (i = $1; i <= c3; i++) {
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-		__collate_char_pri_table[(u_char)i].sec = sec_pri++;
-	}
-}
-	| NAME RANGE NAME {
-	u_int i, c1, c3;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0) {
-			c1 = i;
-			goto findssi21;
-		}
-	yyerror("Name <%s> not defined", $1);
-	findssi21:
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $3) == 0) {
-			c3 = i;
-			goto findssi23;
-		}
-	yyerror("Name <%s> not defined", $3);
-	findssi23:
-
-	if (c3 <= c1)
-		yyerror("Illegal range 0x%02x -- 0x%02x",
-			c1, c3);
-
-	for (i = c1; i <= c3; i++) {
-		__collate_char_pri_table[(u_char)i].prim = prim_pri;
-		__collate_char_pri_table[(u_char)i].sec = sec_pri++;
-	}
-}
-	| NAME {
-	u_int i;
-
-	for (i = 0; i <= UCHAR_MAX; i++)
-		if (strcmp(charmap_table[i], $1) == 0)
-			goto findssi;
-	yyerror("Name <%s> not defined", $1);
-	findssi:
-
-	__collate_char_pri_table[i].prim = prim_pri;
-	__collate_char_pri_table[i].sec = sec_pri++;
-}
 	| CHAIN {
 	if (chain_index >= TABLE_SIZE - 1)
 		yyerror("__collate_chain_pri_table overflow");
diff --git a/usr.bin/colldef/scan.l b/usr.bin/colldef/scan.l
index 6a7d0cf..d8c1116 100644
--- a/usr.bin/colldef/scan.l
+++ b/usr.bin/colldef/scan.l
@@ -41,6 +41,7 @@ int line_no = 1, save_no;
 u_char buf[STR_LEN], *ptr;
 FILE *map_fp;
 extern char map_name[];
+extern u_char charmap_table[UCHAR_MAX + 1][STR_LEN];
 YY_BUFFER_STATE main_buf, map_buf;
 #ifdef FLEX_DEBUG
 YYSTYPE yylval;
@@ -137,13 +138,21 @@ YYSTYPE yylval;
 	*ptr++ = '"';
 }
 <name>\>		{
+	u_int i;
+
 	if (ptr == buf)
-		errx(EX_UNAVAILABLE, "name expected near line %u",
+		errx(EX_UNAVAILABLE, "non-empty name expected near line %u",
 		     line_no);
 	*ptr = '\0';
-	strcpy(yylval.str, buf);
+	for (i = 0; i <= UCHAR_MAX; i++)
+		if (strcmp(charmap_table[i], buf) == 0)
+			goto findit;
+	errx(EX_UNAVAILABLE, "name <%s> not 'charmap'-defined near line %u",
+		buf, line_no);
+	findit:
+	yylval.ch = i;
 	BEGIN(INITIAL);
-	return NAME;
+	return CHAR;
 }
 <string>\"		{
 	*ptr = '\0';