scanning identifiers and LPAREN, some rearrangements around who reads chars when

3 files changed, 31 insertions, 20 deletions

diff --git a/minic/parse.c b/minic/parse.c
index 8de56a6..d1464cd 100644
--- a/minic/parse.c
+++ b/minic/parse.c
@@ -1,6 +1,7 @@
 #include "parse.h"
 
 #include "io.h"
+#include "string.h"
 
 void parser_init( Parser *p, Scanner *s )
 {
@@ -20,9 +21,13 @@ void parser_debug( Parser *p, int enable )
 	}
 }
 
-static void print_symbol( scanner_Symbol s )
+static void print_symbol( scanner_Symbol sym )
 {
-	switch( s.sym ) {
+	char s[MAX_TMP_LEN];
+	
+	s[0] = '\0';
+	
+	switch( sym.sym ) {
 		case S_eof:
 			print( "PARSER(EOF)" );
 			break;
@@ -33,8 +38,13 @@ static void print_symbol( scanner_Symbol s )
 			print( "PARSER(DIV)" );
 			break;
 		case S_IDENT:
-			// TODO: print identifier
-			print( "PARSER(IDENT)" );
+			strcat( s, "PARSER(IDENT," );
+			strcat( s, sym.data.s );
+			strcat_c( s, ')' );
+			print( s );
+			break;
+		case S_LPARENT:
+			print( "PARSER(LPARENT)" );		
 			break;
 		default:
 			print( "PARSER(<unknown symbol>)" );
diff --git a/minic/scan.c b/minic/scan.c
index 6e325bc..33803a5 100644
--- a/minic/scan.c
+++ b/minic/scan.c
@@ -65,12 +65,10 @@ static void skip_char( Scanner *s, int seek )
 static void skip_whitespace( Scanner *s )
 {
 	for( ; ; get_char( s ) ) {
-		if( s->peek == ' ' || s->peek == '\t' ) {
+		if( s->peek == ' ' || s->peek == '\t' || s->peek == '\n' ) {
 			continue;
-		} else if( s->peek == '\n' ) {
-			break;
 		} else if( s->peek == '\0' ) {
-			break;
+			return;
 		} else {
 			break;
 		}
@@ -126,20 +124,16 @@ static scanner_Symbol parse_ident( Scanner *s )
 {
 	scanner_Symbol sym;
 	int len = 0;
-	char tok[MAX_IDENT_LEN+1];
+
+	sym.sym = S_IDENT;
+	sym.data.s[0] = '\0';
 	
-	while( len < MAX_IDENT_LEN && isalnum( s->peek ) ) {
-		strcat_c( tok, s->peek );
+	while( len < MAX_IDENT_LEN && ( isalnum( s->peek ) || s->peek == '_' ) ) {
+		strcat_c( sym.data.s, s->peek );
 		len++;
 		get_char( s );
 	}
 
-	/* TODO: we can add the token into the identifier table ourselves or
-	 * pass it to the parser and let the parser do it?
-	 */
-	 
-	sym.sym = S_IDENT;
-	
 	return sym;
 }
 
@@ -149,7 +143,6 @@ scanner_Symbol scanner_scan( Scanner *s )
 	int c;
 
 	for( ;; ) {
-		get_char( s );
 		skip_whitespace( s );	
 		switch( s->peek ) {
 			case '\0':
@@ -171,6 +164,7 @@ scanner_Symbol scanner_scan( Scanner *s )
 					}
 					continue;
 				} else {
+					get_char( s );
 					sym.sym = S_DIV;
 					return sym;
 				}
@@ -181,8 +175,9 @@ scanner_Symbol scanner_scan( Scanner *s )
 					case 'n':
 						switch( peek_char( s, 2 ) ) {
 							case 't':
-								// todo: consume two chars
 								skip_char( s, 2 );
+								get_char( s );
+								get_char( s );
 								sym.sym = S_INT;
 								return sym;
 						}
@@ -194,6 +189,11 @@ scanner_Symbol scanner_scan( Scanner *s )
 			case 'm': case 'a': case 'n':
 				sym = parse_ident( s );
 				return sym;
+			
+			case '(':
+				get_char( s );
+				sym.sym = S_LPARENT;
+				return sym;
 				
 			default:
 				unexpected_char( s, s->peek );
diff --git a/minic/scan.h b/minic/scan.h
index be24e08..1a7cc79 100644
--- a/minic/scan.h
+++ b/minic/scan.h
@@ -7,7 +7,8 @@ typedef enum scanner_Sym {
 	S_eof,
 	S_DIV,
 	S_INT,
-	S_IDENT
+	S_IDENT,
+	S_LPARENT
 } scanner_Sym;
 
 typedef struct scanner_Symbol {