harec

commit 31c7f1381d12aba90f690945d4c4f1cd87782e88
parent f3d7aa01e280d8cdeac6106fd808b9cffd6f57ff
Author: Drew DeVault <sir@cmpwn.com>
Date:   Sun, 22 Nov 2020 09:10:28 -0500

lex: add unlex

Diffstat:
M
include/lex.h
 | 
2
++
M
src/lex.c
 | 
14
++++++++++++++

2 files changed, 16 insertions(+), 0 deletions(-)
diff --git a/include/lex.h b/include/lex.h
@@ -138,11 +138,13 @@ struct lexer {
 	char *buf;
 	size_t bufsz, buflen;
 	uint32_t c[2];
+	struct token un;
 };
 
 void lex_init(struct lexer *lexer, FILE *f);
 void lex_finish(struct lexer *lexer);
 enum lexical_token lex(struct lexer *lexer, struct token *out);
+void unlex(struct lexer *lexer, struct token *in);
 
 void token_finish(struct token *tok);
 const char *token_str(const struct token *tok);
diff --git a/src/lex.c b/src/lex.c
@@ -117,6 +117,7 @@ lex_init(struct lexer *lexer, FILE *f)
 	lexer->in = f;
 	lexer->bufsz = 256;
 	lexer->buf = calloc(1, lexer->bufsz);
+	lexer->un.token = T_ERROR;
 }
 
 void
@@ -711,6 +712,12 @@ lex2(struct lexer *lexer, struct token *out, uint32_t c)
 enum lexical_token
 lex(struct lexer *lexer, struct token *out)
 {
+	if (lexer->un.token != T_ERROR) {
+		*out = lexer->un;
+		lexer->un.token = T_ERROR;
+		return out->token;
+	}
+
 	uint32_t c = wgetc(lexer);
 	if (c == UTF8_INVALID) {
 		out->token = T_EOF;
@@ -836,3 +843,10 @@ token_str(const struct token *tok)
 		return lexical_token_str(tok->token);
 	}
 }
+
+void
+unlex(struct lexer *lexer, struct token *in)
+{
+	assert(lexer->un.token == T_ERROR && "Only one unlex is supported");
+	lexer->un = *in;
+}