API: Change 'data' field from a single char to a string

Allowing multiple bytes to be returned in a single YXML_DATA token. This
is (unfortunately) necessary for a few special cases:
- &#N; for N > 127,
- <? ? ?>
- <![CDATA[ ] ]]>

I'll fix those separately in the next commits. This data string is now
also re-used as a temporary buffer for entity/char references, removing
the private 'ref' field.

4 files changed, 32 insertions, 21 deletions

diff --git a/test/test.c b/test/test.c
index ca272c9..c0a3c6c 100644
--- a/test/test.c
+++ b/test/test.c
@@ -87,7 +87,7 @@ static void y_printres(yxml_t *x, yxml_ret_t r) {
 			y_printtoken(x, "data ");
 			indata = 1;
 		}
-		y_printchar(x->data);
+		y_printstring(x->data);
 		break;
 	case YXML_PISTART:
 		y_printtoken(x, "pistart ");
diff --git a/yxml.c b/yxml.c
index c300fbd..213b7d1 100644
--- a/yxml.c
+++ b/yxml.c
@@ -110,14 +110,20 @@ typedef enum {
 #define yxml_isRef(c) (yxml_isNum(c) || yxml_isAlpha(c) || c == '#')
 
 
-/* Set the x->data value to ch and tell the application we have some data.
- * This can't be done with simple assignment because char may be unsigned, and
+/* Set the given char value to ch (0<=ch<=255).
+ * This can't be done with simple assignment because char may be signed, and
  * unsigned-to-signed overflow is implementation defined in C. This function
  * /looks/ inefficient, but gcc compiles it down to a single movb instruction
  * on x86, even with -O0. */
-static inline int yxml_setdata(yxml_t *x, unsigned ch) {
+static inline void yxml_setchar(char *dest, unsigned ch) {
 	unsigned char _ch = ch;
-	memcpy(&x->data, &_ch, 1);
+	memcpy(dest, &_ch, 1);
+}
+
+
+static inline int yxml_setdata(yxml_t *x, unsigned ch) {
+	yxml_setchar(x->data, ch);
+	x->data[1] = 0;
 	return YXML_DATA;
 }
 
@@ -209,23 +215,23 @@ static inline int yxml_pivalend (yxml_t *x, unsigned ch) { yxml_popstack(x); x->
 
 
 static inline int yxml_refstart(yxml_t *x, unsigned ch) {
-	memset(x->ref, 0, sizeof(x->ref));
+	memset(x->data, 0, sizeof(x->data));
 	x->reflen = 0;
 	return YXML_OK;
 }
 
 
 static int yxml_ref(yxml_t *x, unsigned ch) {
-	if(x->reflen >= sizeof(x->ref)-1)
+	if(x->reflen >= sizeof(x->data)-1)
 		return YXML_EREF;
-	x->ref[x->reflen] = ch;
+	yxml_setchar(x->data+x->reflen, ch);
 	x->reflen++;
 	return YXML_OK;
 }
 
 
 static int yxml_refend(yxml_t *x, unsigned ch) {
-	unsigned char *r = x->ref;
+	unsigned char *r = (unsigned char *)x->data;
 	ch = 0;
 	if(*r == '#') {
 		if(r[1] == 'x')
diff --git a/yxml.c.in b/yxml.c.in
index 6a32a08..c6c1a4a 100644
--- a/yxml.c.in
+++ b/yxml.c.in
@@ -46,14 +46,20 @@ typedef enum {
 #define yxml_isRef(c) (yxml_isNum(c) || yxml_isAlpha(c) || c == '#')
 
 
-/* Set the x->data value to ch and tell the application we have some data.
- * This can't be done with simple assignment because char may be unsigned, and
+/* Set the given char value to ch (0<=ch<=255).
+ * This can't be done with simple assignment because char may be signed, and
  * unsigned-to-signed overflow is implementation defined in C. This function
  * /looks/ inefficient, but gcc compiles it down to a single movb instruction
  * on x86, even with -O0. */
-static inline int yxml_setdata(yxml_t *x, unsigned ch) {
+static inline void yxml_setchar(char *dest, unsigned ch) {
 	unsigned char _ch = ch;
-	memcpy(&x->data, &_ch, 1);
+	memcpy(dest, &_ch, 1);
+}
+
+
+static inline int yxml_setdata(yxml_t *x, unsigned ch) {
+	yxml_setchar(x->data, ch);
+	x->data[1] = 0;
 	return YXML_DATA;
 }
 
@@ -145,23 +151,23 @@ static inline int yxml_pivalend (yxml_t *x, unsigned ch) { yxml_popstack(x); x->
 
 
 static inline int yxml_refstart(yxml_t *x, unsigned ch) {
-	memset(x->ref, 0, sizeof(x->ref));
+	memset(x->data, 0, sizeof(x->data));
 	x->reflen = 0;
 	return YXML_OK;
 }
 
 
 static int yxml_ref(yxml_t *x, unsigned ch) {
-	if(x->reflen >= sizeof(x->ref)-1)
+	if(x->reflen >= sizeof(x->data)-1)
 		return YXML_EREF;
-	x->ref[x->reflen] = ch;
+	yxml_setchar(x->data+x->reflen, ch);
 	x->reflen++;
 	return YXML_OK;
 }
 
 
 static int yxml_refend(yxml_t *x, unsigned ch) {
-	unsigned char *r = x->ref;
+	unsigned char *r = (unsigned char *)x->data;
 	ch = 0;
 	if(*r == '#') {
 		if(r[1] == 'x')
diff --git a/yxml.h b/yxml.h
index b8c7190..f3a014e 100644
--- a/yxml.h
+++ b/yxml.h
@@ -74,13 +74,13 @@ typedef struct {
 	 * including the YXML_ELEMCLOSE for the corresponding element. */
 	char *elem;
 
-	/* The last read character of an attribute value, element data, or
+	/* The last read character(s) of an attribute value, element data, or
 	 * processing instruction. Changed after YXML_DATA and only valid until the
-	 * next yxml_parse() call.
+	 * next yxml_parse() call. Currently, only the first byte is ever set.
 	 * Note: For processing instructions, the last '?' is considered part of
 	 *   the data, unless the PI was empty (e.g. "<?SomePI?>"), in which case
 	 *   no DATA token is emitted at all. */
-	char data;
+	char data[8];
 
 	/* Name of the current attribute. Changed after YXML_ATTRSTART, valid up to
 	 * and including the next YXML_ATTREND. */
@@ -103,7 +103,6 @@ typedef struct {
 	int state;
 	unsigned char *stack; /* Stack of element names + attribute/PI name, separated by \0. Also starts with a \0. */
 	size_t stacksize, stacklen;
-	unsigned char ref[8];
 	unsigned reflen;
 	unsigned quote;
 	int nextstate; /* Used for '@' state remembering and for the "string" consuming state */