config: implement string tokenizing

1 files changed, 123 insertions, 2 deletions

diff --git a/common.c b/common.c
index 67781dc..7bfe9fe 100644
--- a/common.c
+++ b/common.c
@@ -1009,6 +1009,120 @@ config_tokenizer_error (struct config_tokenizer *self,
 	str_free (&description);
 }
 
+static bool
+config_tokenizer_hexa_escape (struct config_tokenizer *self, struct str *output)
+{
+	int i;
+	unsigned char code = 0;
+
+	for (i = 0; self->len && i < 2; i++)
+	{
+		unsigned char c = tolower_ascii (*self->p);
+		if (c >= '0' && c <= '9')
+			code = (code << 4) | (c - '0');
+		else if (c >= 'a' && c <= 'f')
+			code = (code << 4) | (c - 'a' + 10);
+		else
+			break;
+
+		config_tokenizer_advance (self);
+	}
+
+	if (!i)
+		return false;
+
+	str_append_c (output, code);
+	return true;
+}
+
+static bool
+config_tokenizer_octal_escape
+	(struct config_tokenizer *self, struct str *output)
+{
+	int i;
+	unsigned char code = 0;
+
+	for (i = 0; self->len && i < 3; i++)
+	{
+		unsigned char c = *self->p;
+		if (c >= '0' && c <= '7')
+			code = (code << 3) | (c - '0');
+		else
+			break;
+
+		config_tokenizer_advance (self);
+	}
+
+	if (!i)
+		return false;
+
+	str_append_c (output, code);
+	return true;
+}
+
+static bool
+config_tokenizer_escape_sequence
+	(struct config_tokenizer *self, struct str *output, struct error **e)
+{
+	if (!self->len)
+	{
+		config_tokenizer_error (self, e, "premature end of escape sequence");
+		return false;
+	}
+
+	unsigned char c;
+	switch ((c = *self->p))
+	{
+	case '"':              break;
+	case '\\':             break;
+	case 'a':   c = '\a';  break;
+	case 'b':   c = '\b';  break;
+	case 'f':   c = '\f';  break;
+	case 'n':   c = '\n';  break;
+	case 'r':   c = '\r';  break;
+	case 't':   c = '\t';  break;
+	case 'v':   c = '\v';  break;
+
+	case 'x':
+	case 'X':
+		config_tokenizer_advance (self);
+		if (config_tokenizer_hexa_escape (self, output))
+			return true;
+
+		config_tokenizer_error (self, e, "invalid hexadecimal escape");
+		return false;
+
+	default:
+		if (config_tokenizer_octal_escape (self, output))
+			return true;
+
+		config_tokenizer_error (self, e, "unknown escape sequence");
+		return false;
+	}
+
+	str_append_c (output, c);
+	config_tokenizer_advance (self);
+	return true;
+}
+
+static bool
+config_tokenizer_string
+	(struct config_tokenizer *self, struct str *output, struct error **e)
+{
+	unsigned char c;
+	while (self->len)
+	{
+		if ((c = config_tokenizer_advance (self)) == '"')
+			return true;
+		if (c != '\\')
+			str_append_c (output, c);
+		else if (!config_tokenizer_escape_sequence (self, output, e))
+			return false;
+	}
+	config_tokenizer_error (self, e, "premature end of string");
+	return false;
+}
+
 static enum config_token
 config_tokenizer_next (struct config_tokenizer *self, struct error **e)
 {
@@ -1033,8 +1147,15 @@ config_tokenizer_next (struct config_tokenizer *self, struct error **e)
 		return CONFIG_T_ABORT;
 
 	case '"':
-		// TODO: string, validate as UTF-8
-		break;
+		config_tokenizer_advance (self);
+		str_reset (&self->string);
+		if (!config_tokenizer_string (self, &self->string, e))
+			return CONFIG_T_ABORT;
+		if (!utf8_validate (self->string.str, self->string.len))
+		{
+			config_tokenizer_error (self, e, "not a valid UTF-8 string");
+			return CONFIG_T_ABORT;
+		}
 	}
 
 	bool is_word = false;