Fix a bug on name parsing at the end of current input buffer

For https://bugzilla.gnome.org/show_bug.cgi?id=754946

When hitting the end of the current input buffer while parsing
a name we could end up loosing the beginning of the name, which
led to various issues.

5 files changed, 41 insertions, 9 deletions

diff --git a/parser.c b/parser.c
index 0edd53bc..fd29a395 100644
--- a/parser.c
+++ b/parser.c
@@ -3491,7 +3491,14 @@ xmlParseNCNameComplex(xmlParserCtxtPtr ctxt) {
 	c = CUR_CHAR(l);
 	if (c == 0) {
 	    count = 0;
+	    /*
+	     * when shrinking to extend the buffer we really need to preserve
+	     * the part of the name we already parsed. Hence rolling back
+	     * by current lenght.
+	     */
+	    ctxt->input->cur -= l;
 	    GROW;
+	    ctxt->input->cur += l;
             if (ctxt->instate == XML_PARSER_EOF)
                 return(NULL);
 	    end = ctxt->input->cur;
@@ -3523,7 +3530,7 @@ xmlParseNCNameComplex(xmlParserCtxtPtr ctxt) {
 
 static const xmlChar *
 xmlParseNCName(xmlParserCtxtPtr ctxt) {
-    const xmlChar *in;
+    const xmlChar *in, *e;
     const xmlChar *ret;
     int count = 0;
 
@@ -3535,16 +3542,19 @@ xmlParseNCName(xmlParserCtxtPtr ctxt) {
      * Accelerator for simple ASCII names
      */
     in = ctxt->input->cur;
-    if (((*in >= 0x61) && (*in <= 0x7A)) ||
-	((*in >= 0x41) && (*in <= 0x5A)) ||
-	(*in == '_')) {
+    e = ctxt->input->end;
+    if ((((*in >= 0x61) && (*in <= 0x7A)) ||
+	 ((*in >= 0x41) && (*in <= 0x5A)) ||
+	 (*in == '_')) && (in < e)) {
 	in++;
-	while (((*in >= 0x61) && (*in <= 0x7A)) ||
-	       ((*in >= 0x41) && (*in <= 0x5A)) ||
-	       ((*in >= 0x30) && (*in <= 0x39)) ||
-	       (*in == '_') || (*in == '-') ||
-	       (*in == '.'))
+	while ((((*in >= 0x61) && (*in <= 0x7A)) ||
+	        ((*in >= 0x41) && (*in <= 0x5A)) ||
+	        ((*in >= 0x30) && (*in <= 0x39)) ||
+	        (*in == '_') || (*in == '-') ||
+	        (*in == '.')) && (in < e))
 	    in++;
+	if (in >= e)
+	    goto complex;
 	if ((*in > 0) && (*in < 0x80)) {
 	    count = in - ctxt->input->cur;
             if ((count > XML_MAX_NAME_LENGTH) &&
@@ -3562,6 +3572,7 @@ xmlParseNCName(xmlParserCtxtPtr ctxt) {
 	    return(ret);
 	}
     }
+complex:
     return(xmlParseNCNameComplex(ctxt));
 }
 
diff --git a/result/errors/754946.xml b/result/errors/754946.xml
new file mode 100644
index 00000000..e69de29b
--- /dev/null
+++ b/result/errors/754946.xml
diff --git a/result/errors/754946.xml.err b/result/errors/754946.xml.err
new file mode 100644
index 00000000..423dff56
--- /dev/null
+++ b/result/errors/754946.xml.err
@@ -0,0 +1,16 @@
+Entity: line 1: parser error : internal error: xmlParseInternalSubset: error detected in Markup declaration
+
+ %SYSTEM; 
+         ^
+Entity: line 1: 
+A<lbbbbbbbbbbbbbbbbbbb_
+^
+Entity: line 1: parser error : DOCTYPE improperly terminated
+ %SYSTEM; 
+         ^
+Entity: line 1: 
+A<lbbbbbbbbbbbbbbbbbbb_
+^
+./test/errors/754946.xml:1: parser error : Extra content at the end of the document
+<!DOCTYPEA[<!ENTITY %
+  ^
diff --git a/result/errors/754946.xml.str b/result/errors/754946.xml.str
new file mode 100644
index 00000000..3b748cc7
--- /dev/null
+++ b/result/errors/754946.xml.str
@@ -0,0 +1,4 @@
+./test/errors/754946.xml:1: parser error : Extra content at the end of the document
+<!DOCTYPEA[<!ENTITY %
+          ^
+./test/errors/754946.xml : failed to parse
diff --git a/test/errors/754946.xml b/test/errors/754946.xml
new file mode 100644
index 00000000..6b5f9b06
--- /dev/null
+++ b/test/errors/754946.xml
@@ -0,0 +1 @@
+<!DOCTYPEA[<!ENTITY %

SYSTEM "A<lbbbbbbbbbbbbbbbbbbb_"
>%SYSTEM;<![
\ No newline at end of file