reimplementation (oupsi)

old/lex.c

@@ -0,0 +1,119 @@
+
+#include <err.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+#include "wscm.h"
+
+// helpers
+static int iswhite(int ch) { return ch == ' ' || ch == '\t' || ch == '\n'; }
+
+static int isdelim(int ch) {
+  return ch == '(' || ch == ')' || ch == '\'' || ch == ';';
+}
+
+static inline void appendchar(Lx *lex, char ch) {
+  if (lex->cursor > LEXER_CAP) {
+    fprintf(stderr, "lexer buffer overflow");
+    abort();
+  }
+  lex->buffer[lex->cursor++] = ch;
+}
+
+static int getcws(Lx *lex) {
+  if (feof(lex->input))
+    return EOF;
+  for (;;) {
+    int ch = getc(lex->input);
+    if (iswhite(ch))
+      continue;
+    return ch;
+  }
+}
+
+static int scanword(Lx *lex) {
+  int ch = getc(lex->input);
+  for (;;) {
+    if (ch == EOF) {
+      if (lex->cursor == 0)
+        lex->kind = TOK_EOF;
+      appendchar(lex, 0);
+      return lex->kind;
+    } else if (iswhite(ch) || isdelim(ch)) {
+      ungetc(ch, lex->input);
+      appendchar(lex, 0);
+      return lex->kind;
+    } else {
+      appendchar(lex, ch);
+      ch = getc(lex->input);
+    }
+  }
+}
+
+static int scanstring(Lx *lex) {
+  int ch;
+  for (;;) {
+    ch = getc(lex->input);
+    switch (ch) {
+    case EOF:
+      goto eof;
+    case '\\':
+      ch = getc(lex->input);
+      if (ch == EOF)
+        goto eof;
+      switch (ch) {
+      case 'n':
+        appendchar(lex, '\n');
+        break;
+      case 't':
+        appendchar(lex, '\t');
+        break;
+      case '"':
+        appendchar(lex, '"');
+        break;
+      }
+      break;
+    case '"':
+      appendchar(lex, 0);
+      return (lex->kind = TOK_STRING);
+    default:
+      appendchar(lex, ch);
+    }
+  }
+
+eof:
+  errx(1, "unterminated string literal");
+}
+
+int nexttoken(Lx *lex) {
+  int ch;
+  lex->cursor = 0;
+
+  if (feof(lex->input)) {
+    lex->kind = TOK_EOF;
+    *lex->buffer = 0;
+    return 0;
+  }
+
+  ch = getcws(lex);
+  switch (ch) {
+  case ';':
+    for (; ch != '\n'; ch = getc(lex->input))
+      appendchar(lex, ch);
+    appendchar(lex, 0);
+    return (lex->kind = TOK_COMMENT);
+  case '(':
+  case ')':
+  case '.':
+  case '\'':
+    return (lex->kind = ch);
+  case '"':
+    return scanstring(lex);
+  default:
+    ungetc(ch, lex->input);
+    lex->kind = TOK_WORD;
+    return scanword(lex);
+  }
+
+  return 0;
+}