Normalize white space in attribute value data

3 files changed, 14 insertions, 2 deletions

diff --git a/yxml-states b/yxml-states
index eeede4a..0c3bc27 100644
--- a/yxml-states
+++ b/yxml-states
@@ -149,6 +149,6 @@ elem3      '>' selfclose misc2
 attr0      Name attrname attr0; SP attrnameend attr1; '=' attrnameend attr2
 attr1      SP attr1; '=' attr2
 attr2      SP attr2; '\''|'"' $quote attr3
-attr3      AttValue setdata attr3; '&' refstart attr4; $quote elem2
+attr3      AttValue setattrval attr3; '&' refstart attr4; $quote elem2
 attr4      Ref ref attr4; '\x3b' refend attr3
 
diff --git a/yxml.c b/yxml.c
index 6d3f825..cf3726c 100644
--- a/yxml.c
+++ b/yxml.c
@@ -116,6 +116,12 @@ static inline int yxml_setdata(yxml_t *x, unsigned ch) {
 }
 
 
+static inline int yxml_setattrval(yxml_t *x, unsigned ch) {
+	/* Normalize attribute values according to the XML spec section 3.3.3. */
+	return yxml_setdata(x, ch == 0x9 || ch == 0xa ? 0x20 : ch);
+}
+
+
 /* Go to the misc1 or misc2 state, depending on whether a tag has been opened
  * previously or not. (This is a hack to work around a limitation in the state
  * machine description in yxml-states) */
@@ -339,7 +345,7 @@ yxml_ret_t yxml_parse(yxml_t *x, int _ch) {
 		break;
 	case YXMLS_attr3:
 		if(yxml_isAttValue(ch))
-			return yxml_setdata(x, ch);
+			return yxml_setattrval(x, ch);
 		if(ch == (unsigned char)'&') {
 			x->state = YXMLS_attr4;
 			return yxml_refstart(x, ch);
diff --git a/yxml.c.in b/yxml.c.in
index f08b913..c6c2bd0 100644
--- a/yxml.c.in
+++ b/yxml.c.in
@@ -59,6 +59,12 @@ static inline int yxml_setdata(yxml_t *x, unsigned ch) {
 }
 
 
+static inline int yxml_setattrval(yxml_t *x, unsigned ch) {
+	/* Normalize attribute values according to the XML spec section 3.3.3. */
+	return yxml_setdata(x, ch == 0x9 || ch == 0xa ? 0x20 : ch);
+}
+
+
 /* Go to the misc1 or misc2 state, depending on whether a tag has been opened
  * previously or not. (This is a hack to work around a limitation in the state
  * machine description in yxml-states) */