added @pragma multilinestr on/off

hawk/README.md

@@ -133,7 +133,7 @@ The typical execution begins with the BEGIN block, goes through pattern-action b
 	}
 
 
-### Value
+### Values
 
 - unitialized value
 - integer
@@ -142,10 +142,22 @@ The typical execution begins with the BEGIN block, goes through pattern-action b
 - byte string
 - array
 
+### Pragmas
+
+A program item of the file scope can be placed in any source files.
+A pragma item of the global scope can appear only once thoughout the all source files.
+
+|               | Scope  | Values        | Description                                            |
+|---------------|--------|---------------|--------------------------------------------------------|
+| implicit      | file   | on, off       | allow undeclared variables                             |
+| multilinestr  | file   | on, off       | allow a multiline string literal without continuation  |
+| entry         | global | function name | change the program entry point                         |
+| striprecspc   | global | on, off       |                                                        |
+| striprecspc   | global | on, off       | trim leading and trailing spaces when convering a string to a number |
+
+
 ### Module
 
-
-
 ### Incompatibility with AWK
 
 #### Parameter passing

hawk/bin/main.c

@@ -499,6 +499,7 @@ struct opttab_t
 } opttab[] =
 {
 	{ "implicit",     HAWK_IMPLICIT,       "allow undeclared variables" },
+	{ "multilinestr", HAWK_MULTILINESTR,   "allow raw multiline string and regular expression literals" },
 	{ "nextofile",    HAWK_NEXTOFILE,      "enable nextofile & OFILENAME" },
 	{ "rio",          HAWK_RIO,            "enable builtin I/O including getline & print" },
 	{ "rwpipe",       HAWK_RWPIPE,         "allow a dual-directional pipe" },
@@ -640,6 +641,7 @@ static int process_argv (int argc, hawk_bch_t* argv[], struct arg_t* arg)
 	static hawk_bcli_lng_t lng[] = 
 	{
 		{ ":implicit",         '\0' },
+		{ ":multilinestr",     '\0' },
 		{ ":nextofile",        '\0' },
 		{ ":rio",              '\0' },
 		{ ":rwpipe",           '\0' },

hawk/lib/hawk.c

@@ -379,7 +379,7 @@ void hawk_clear (hawk_t* awk)
 	awk->parse.depth.loop = 0;
 	awk->parse.depth.expr = 0;
 	awk->parse.depth.incl = 0;
-	awk->parse.pragma.trait = (awk->opt.trait & (HAWK_IMPLICIT | HAWK_STRIPRECSPC | HAWK_STRIPSTRSPC)); /* implicit on if you didn't mask it off in awk->opt.trait with hawk_setopt */
+	awk->parse.pragma.trait = (awk->opt.trait & (HAWK_IMPLICIT | HAWK_MULTILINESTR | HAWK_STRIPRECSPC | HAWK_STRIPSTRSPC)); /* implicit on if you didn't mask it off in awk->opt.trait with hawk_setopt */
 	awk->parse.pragma.rtx_stack_limit = 0;
 	awk->parse.pragma.entry[0] = '\0';
 

hawk/lib/hawk.h

@@ -1151,8 +1151,11 @@ enum hawk_trait_t
 	/** allows undeclared variables */
 	HAWK_IMPLICIT = (1 << 0),
 
+	/** allows multiline string literals or regular expression literals */
+	HAWK_MULTILINESTR = (1 << 1),
+
 	/** enables nextofile and NEXTOFILE */
-	HAWK_NEXTOFILE = (1 << 1),
+	HAWK_NEXTOFILE = (1 << 2),
 
 	/** supports \b getline, \b print, \b printf, \b close, \b fflush,
 	 *  piping, and file rediction */

hawk/lib/parse.c

@@ -872,7 +872,7 @@ static int begin_include (hawk_t* awk, int once)
 	arg->pragma_trait = awk->parse.pragma.trait;
 	/* but don't change awk->parse.pragma.trait. it means the included file inherits
 	 * the existing progma values. 
-	awk->parse.pragma.trait = (awk->opt.trait & (HAWK_IMPLICIT | HAWK_STRIPRECSPC | HAWK_STRIPSTRSPC));
+	awk->parse.pragma.trait = (awk->opt.trait & (HAWK_IMPLICIT | HAWK_MULTILINESTR | HAWK_STRIPRECSPC | HAWK_STRIPSTRSPC));
 	*/
 
 	/* i update the current pointer after opening is successful */
@@ -1039,6 +1039,35 @@ static int parse_progunit (hawk_t* awk)
 				goto error_ident_on_off_expected_for_implicit;
 			}
 		}
+		else if (hawk_comp_oochars_oocstr(name.ptr, name.len, HAWK_T("multilinestr")) == 0)
+		{
+			if (get_token(awk) <= -1) return -1;
+			if (!MATCH(awk, TOK_IDENT))
+			{
+			error_ident_on_off_expected_for_multilinestr:
+				hawk_seterrfmt (awk, &awk->ptok.loc, HAWK_EIDENT, HAWK_T("identifier 'on' or 'off' expected for 'multilinestr'"));
+				return -1;
+			}
+
+			name.len = HAWK_OOECS_LEN(awk->tok.name);
+			name.ptr = HAWK_OOECS_PTR(awk->tok.name);
+			if (hawk_comp_oochars_oocstr(name.ptr, name.len, HAWK_T("on")) == 0)
+			{
+				awk->parse.pragma.trait |= HAWK_MULTILINESTR;
+			}
+			else if (hawk_comp_oochars_oocstr(name.ptr, name.len, HAWK_T("off")) == 0)
+			{
+				awk->parse.pragma.trait &= ~HAWK_MULTILINESTR;
+			}
+			else
+			{
+				goto error_ident_on_off_expected_for_multilinestr;
+			}
+		}
+		/* ---------------------------------------------------------------------
+		 * the pragmas up to this point affect the parser 
+		 * the following pragmas affect runtime 
+		 * --------------------------------------------------------------------- */
 		else if (hawk_comp_oochars_oocstr(name.ptr, name.len, HAWK_T("stack_limit")) == 0)
 		{
 			hawk_int_t sl;
@@ -6131,22 +6160,47 @@ static int get_string (
 			}
 		}
 
-		if (escaped == 0 && c == end_char)
+		if (escaped == 99)
 		{
-			/* terminating quote */
+			escaped = 0;
-			/*GET_CHAR_TO (awk, c);*/
+			if (c == '\n') continue; /* backslash \r \n */
-			GET_CHAR (awk);
-			break;
 		}
 
-		if (escaped == 0 && c == esc_char)
+		if (escaped == 0)
 		{
-			escaped = 1;
+			if (c == end_char)
-			continue;
+			{
+				/* terminating quote */
+				/*GET_CHAR_TO (awk, c);*/
+				GET_CHAR (awk);
+				break;
+			}
+			else if (c == esc_char)
+			{
+				escaped = 1;
+				continue;
+			}
+			else if (!(awk->parse.pragma.trait & HAWK_MULTILINESTR) && c == '\n' || c == '\r')
+			{
+				hawk_seterrnum (awk, &awk->tok.loc, HAWK_ESTRNC);
+				return -1;
+			}
 		}
 
 		if (escaped == 1)
 		{
+			if (c == '\n')
+			{
+				/* line continuation - a backslash at the end of line */
+				escaped = 0;
+				continue;
+			}
+			else if (c == '\r')
+			{
+				escaped = 99;
+				continue;
+			}
+
 			if (c == HAWK_T('n')) c = HAWK_T('\n');
 			else if (c == HAWK_T('r')) c = HAWK_T('\r');
 			else if (c == HAWK_T('t')) c = HAWK_T('\t');