core: Add lexer + parser groundwork

Signed-off-by: Ian Moffett <ian@mirocom.org>

1 files changed, 84 insertions, 0 deletions

diff --git a/core/lexer.c b/core/lexer.c
new file mode 100644
index 0000000..d6e5e66
--- /dev/null
+++ b/core/lexer.c
@@ -0,0 +1,84 @@
+#include <errno.h>
+#include <stdbool.h>
+#include "cescal/lexer.h"
+#include "cescal/log.h"
+
+/*
+ * Returns true if the given character is a whitespace
+ *
+ * @c: Character to check
+ */
+static inline bool
+lexer_is_ws(char c)
+{
+    switch (c) {
+    case '\t':
+    case '\n':
+    case ' ':
+    case '\f':
+    case '\r':
+        return true;
+    }
+
+    return false;
+}
+
+/*
+ * Consume a single character from the input source file and
+ * optionally skip whitespace
+ *
+ * @state: Compiler state
+ * @skip_ws: If true skip whitespace
+ */
+static char
+lexer_consume_single(struct cescal_state *state, bool skip_ws)
+{
+    char c;
+
+    if (state == NULL) {
+        return '\0';
+    }
+
+    while ((c = readbuf_read(&state->rb, state->in_fd)) != '\0') {
+        if (lexer_is_ws(c)) {
+            continue;
+        }
+
+        return c;
+    }
+
+    return '\0';
+}
+
+int
+lexer_nom(struct cescal_state *state, struct token *res)
+{
+    char c;
+
+    if (state == NULL || res == NULL) {
+        errno = EINVAL;
+        return -1;
+    }
+
+    if ((c = lexer_consume_single(state, true)) == '\0') {
+        return -1;
+    }
+
+    switch (c) {
+    case '(':
+        res->type = TT_LPAREN;
+        res->c = c;
+        return 0;
+    case ')':
+        res->type = TT_RPAREN;
+        res->c = c;
+        return 0;
+    case ',':
+        res->type = TT_COMMA;
+        res->c = c;
+        return 0;
+    }
+
+    cc_error("got unknown token '%c'\n", c);
+    return -1;
+}