parsing ident

4 files changed, 105 insertions, 10 deletions

diff --git a/minic/const.h b/minic/const.h
index 3c47234..e22075e 100644
--- a/minic/const.h
+++ b/minic/const.h
@@ -1,6 +1,7 @@
 #pragma once
 
 enum {
+	MAX_IDENT_LEN = 15,
 	MAX_TMP_LEN = 50
 };
-	
\ No newline at end of file
+	
diff --git a/minic/parse.c b/minic/parse.c
index 0658908..8de56a6 100644
--- a/minic/parse.c
+++ b/minic/parse.c
@@ -20,6 +20,27 @@ void parser_debug( Parser *p, int enable )
 	}
 }
 
+static void print_symbol( scanner_Symbol s )
+{
+	switch( s.sym ) {
+		case S_eof:
+			print( "PARSER(EOF)" );
+			break;
+		case S_INT:
+			print( "PARSER(INT)" );
+			break;
+		case S_DIV:
+			print( "PARSER(DIV)" );
+			break;
+		case S_IDENT:
+			// TODO: print identifier
+			print( "PARSER(IDENT)" );
+			break;
+		default:
+			print( "PARSER(<unknown symbol>)" );
+	}
+}
+
 void parser_parse( Parser *p )
 {
 	scanner_Symbol s;
@@ -27,5 +48,9 @@ void parser_parse( Parser *p )
 	s = scanner_scan( p->s );
 	
 	while( s.sym != S_eof ) {
+		if( p->debug ) {
+			print_symbol( s );
+		}
+		s = scanner_scan( p->s );
 	}
 }
diff --git a/minic/scan.c b/minic/scan.c
index 1e0f358..6e325bc 100644
--- a/minic/scan.c
+++ b/minic/scan.c
@@ -4,6 +4,7 @@
 #include "const.h"
 #include "stdlib.h"
 #include "string.h"
+#include "ctype.h"
 
 void scanner_init( Scanner *s, char *src )
 {
@@ -41,14 +42,26 @@ static char get_char( Scanner *s )
 
 static char peek_char( Scanner *s, int seek )
 {
-	char c;
 	char *pos = s->pos;
+	char c = *pos;
 
-	for( c = *pos; c != '\0' && seek > 0; pos++, seek-- );
+	while( c != '\0' && seek > 0 ) {
+		c = *pos;
+		pos++;
+		seek--;
+	}
 
 	return c;
 }
 
+static void skip_char( Scanner *s, int seek )
+{
+	while( seek > 0 ) {
+		get_char( s );
+		seek--;
+	}
+}
+
 static void skip_whitespace( Scanner *s )
 {
 	for( ; ; get_char( s ) ) {
@@ -83,6 +96,18 @@ static void error( Scanner *s, char *msg )
 	halt( );
 }
 
+static void unexpected_char( Scanner *s, int c )
+{
+	char msg[MAX_TMP_LEN];
+	msg[0] = '\0';
+
+	strcat( msg, "unexpected character '" );
+	strcat_c( msg, c );
+	strcat( msg, "'" );
+
+	error( s, msg );
+}
+
 static void skip_comment( Scanner *s )
 {
 	get_char( s );
@@ -97,6 +122,27 @@ static void skip_comment( Scanner *s )
 	get_char( s );
 }
 
+static scanner_Symbol parse_ident( Scanner *s )
+{
+	scanner_Symbol sym;
+	int len = 0;
+	char tok[MAX_IDENT_LEN+1];
+	
+	while( len < MAX_IDENT_LEN && isalnum( s->peek ) ) {
+		strcat_c( tok, 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;
+}
+
 scanner_Symbol scanner_scan( Scanner *s )
 {
 	scanner_Symbol sym;
@@ -124,11 +170,33 @@ scanner_Symbol scanner_scan( Scanner *s )
 						print( "SCANNER(COMMENT_END)" );
 					}
 					continue;
+				} else {
+					sym.sym = S_DIV;
+					return sym;
+				}
+				break;
+
+			case 'i':
+				switch( peek_char( s, 1 ) ) {
+					case 'n':
+						switch( peek_char( s, 2 ) ) {
+							case 't':
+								// todo: consume two chars
+								skip_char( s, 2 );
+								sym.sym = S_INT;
+								return sym;
+						}
+						goto id;
 				}
 				break;
-			
+
+			id:
+			case 'm': case 'a': case 'n':
+				sym = parse_ident( s );
+				return sym;
+				
 			default:
-				error( s, "unexpected symbol" );
+				unexpected_char( s, s->peek );
 				sym.sym = S_eof;
 				return sym;
 		}
diff --git a/minic/scan.h b/minic/scan.h
index 3e6942a..be24e08 100644
--- a/minic/scan.h
+++ b/minic/scan.h
@@ -1,14 +1,15 @@
 #pragma once
 
+#include "const.h"
+
 typedef enum scanner_Sym {
 	S_undef,
-	S_eof
+	S_eof,
+	S_DIV,
+	S_INT,
+	S_IDENT
 } scanner_Sym;
 
-enum {
-	MAX_IDENT_LEN = 10
-};
-
 typedef struct scanner_Symbol {
 	scanner_Sym sym;
 	union {