Add more tokens

os.c

@@ -7,13 +7,7 @@
 #include "s8slice.h"
 
 #include <stdio.h>
-
-enum {
-	OS_READ  = 1<<0,
-	OS_WRITE = 1<<1,
-	OS_ALLFLAGS = (1<<2)-1
-};
-typedef unsigned int os_open_flags;
+#include "os.h"
 
 S8Slice os_open_file(S8Slice path, os_open_flags flags)
 {

tokenizer.c

@@ -8,6 +8,13 @@
 #include "os.h"
 #include "s8slice.h"
 
+#if 0
+#define log_trace(...) fprintf(stderr, __VA_ARGS__)
+#else
+#define log_trace(...)
+#endif
+
+
 typedef struct Parser {
     S8Slice file;
     int64_t cursor; 
@@ -17,18 +24,30 @@ typedef struct Token {
     enum {
         TOKEN_UNDEFINED,
         TOKEN_INT,
-		TOKEN_IDENTIFIER,
+        TOKEN_ID,
         TOKEN_STRING,
         TOKEN_SYNTAX_ERROR,
+        TOKEN_EOS,
         TOKEN_EOF,
     } kind;
     union {
         S8Slice identifier;
         S8Slice string;
         int     integer;
+        char    undefined;
     };
 } Token;
 
+const char* token_kind_str[] = {
+    [TOKEN_UNDEFINED]    = "TOKEN_UNDEFINED",
+    [TOKEN_INT]          = "TOKEN_INT",
+    [TOKEN_ID]           = "TOKEN_ID",
+    [TOKEN_STRING]       = "TOKEN_STRING",
+    [TOKEN_SYNTAX_ERROR] = "TOKEN_SYNTAX_ERROR",
+    [TOKEN_EOS]          = "TOKEN_EOS",
+    [TOKEN_EOF]          = "TOKEN_EOF",
+};
+
 Parser parser_attach(S8Slice file)
 {
     return (Parser){.file = file, .cursor = 0};
@@ -39,6 +58,7 @@ static int parser_peek(Parser* p)
     const int ch = p->cursor >= p->file.len
         ? EOF
         : (int)p->file.data[p->cursor];
+    log_trace("peeking:\t%c\n", isprint(ch) ? ch : '?');
     return ch;
 }
 
@@ -48,20 +68,31 @@ static int parser_getch(Parser* p)
         ? EOF
         : (int)p->file.data[p->cursor];
     p->cursor += 1;
-	if (ch == '\n' && parser_peek(p) == '\r') {
+    log_trace(stderr, "getch:\t%c\n", isprint(ch) ? ch : '?');
+    if (ch == '\n' && p->file.data[p->cursor] == '\r') {
         p->cursor += 1;
     }
     return ch;
 }
 
+static void parser_skip_char(Parser* p)
+{
+    const int ch = parser_getch(p);
+    log_trace(stderr, "skipping:\t%c\n", isprint(ch) ? ch : '?');
+}
+
 static void parser_ungetch(Parser* p)
 {
+    log_trace(stderr, "ungetch:\n");
     if (p->cursor > 0) {
         p->cursor -= 1;
     }
+    if (p->file.data[p->cursor] == '\r') {
+        p->cursor -= 1;
+    }
 }
 
-static void parser_discard(Parser* p, int (*f) (int))
+static inline void parser_discard(Parser* p, int (*f) (int))
 {
     int ch;
     while (ch = parser_getch(p), f(ch))
@@ -90,21 +121,26 @@ static Token read_integer(Parser* p)
     return t;
 }
 
+int is_identifier_tail(int ch)
+{
+    return isalnum(ch) || ch == '_';
+}
+
 static Token read_identifier(Parser* p)
 {
     int ch = EOF;
 
-	/* the first char should be guaranteed to be isalpha */
+    /* the should be checked by the caller */
     assert(isalpha(parser_peek(p)));
 
     int64_t begin = p->cursor;
-	while (ch = parser_getch(p), isalnum(ch))
+    while (ch = parser_getch(p), is_identifier_tail(ch))
         /* NOOP */;
     parser_ungetch(p);
     int64_t end = p->cursor;
 
     Token t = {
-		.kind = TOKEN_IDENTIFIER,
+        .kind = TOKEN_ID,
         .identifier = s8slice(&p->file, begin, end)
     };
 
@@ -144,17 +180,34 @@ static Token read_token(Parser* p)
 
     if (ch == EOF) {
         t.kind = TOKEN_EOF;
-	} else if (isalpha(ch)) {
+    }
+    else if (ch == '\n') {
+        t.kind = TOKEN_EOS;
+        parser_skip_char(p);
+        parser_discard(p, isspace);
+    }
+    else if (isalpha(ch)) {
         t = read_identifier(p);
-	} else if (isdigit(ch)) {
+    }
+    else if (isdigit(ch)) {
         t = read_integer(p);
-	} else if (ch == '"') {
+    }
+    else if (ch == '"') {
         t = read_string(p);
+    } else {
+        t.kind = TOKEN_UNDEFINED;
+        t.undefined = ch;
+        parser_skip_char(p);
     }
 
     return t;
 }
 
+int isspace_except_newline(int ch)
+{
+    return isspace(ch) && (ch != '\n');
+}
+
 int main(int argc, char** argv)
 {
     if (argc != 2) {
@@ -173,18 +226,27 @@ int main(int argc, char** argv)
     Parser p = parser_attach(f);
     Token t = {0};
     while (true) {
-		parser_discard(&p, isspace);
+        parser_discard(&p, isspace_except_newline);
         Token t = read_token(&p);
-		if (t.kind == TOKEN_IDENTIFIER) {
+        printf("%s\t", token_kind_str[t.kind]);
+        if (t.kind == TOKEN_ID) {
             S8Slice s = t.identifier;
-			printf("{.len = %lld, .data = %.*s}\n", s.len, (int)s.len, s.data);
-		} else if (t.kind == TOKEN_INT) {
+            printf("<%.*s>\n", (int)s.len, s.data);
+        }
+        else if (t.kind == TOKEN_INT) {
             printf("%d\n", t.integer);
-		} else if (t.kind == TOKEN_STRING) {
+        }
+        else if (t.kind == TOKEN_STRING) {
             S8Slice s = t.identifier;
-			printf("{.len = %lld, .data = %.*s}\n", s.len, (int)s.len, s.data);
-		} else if (t.kind == TOKEN_EOF) {
+            printf("\"%.*s\"\n", (int)s.len, s.data);
+        }
+        else if (t.kind == TOKEN_EOF) {
             break;
+        } else if (t.kind == TOKEN_UNDEFINED) {
+            printf("'%c'\n", t.undefined);
+        }
+        else {
+            printf("\n");
         }
     }