ox 2.0.6 → 2.0.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 2b9a20f56f4e7eb7090f1ed2a58eab7f85276585
-  data.tar.gz: 7c2c55961c8286576d47b925d97f1073ab9722cd
+  metadata.gz: cbb2d5fe1b036f9e4802ed6357cbe71472d3fbd0
+  data.tar.gz: 7732319c54eaa32c56a8f35a44032b99bc29a42c
 SHA512:
-  metadata.gz: 97a5d814b0cc4329c0c0014bc5c42276bba065c7bfda21b3f1056d0993530ff19b11485c34a72c75c85f6dea04798568985e81b73a7ba2fc6f91745ad4a3c847
-  data.tar.gz: b21814309b97237dde61f86312881a91d645a7fa873e1d4dbc5c14c1a7d28df17215180296f5288612451df904a47e3171af4d4564e0ec34c45151cf5cc3a432
+  metadata.gz: b20bea5b1377b022d4ad864574cc8ed4c7b613f47e28722eba7ff1a438fd471068952a69a4e762df7a3d06a9f786fd6d38397e8fb3c36da4c75900b72c834b1e
+  data.tar.gz: 916f26854111928b6b402dcfbae0288b4c99b6544b31601ec7860b19a36e4160f53f2839b62598a196d426c9685f6b4a4ba9367b7f03654a0fd81963d86a4a35

data/README.md

@@ -34,13 +34,9 @@ A fast XML parser and Object marshaller as a Ruby gem.
 
 ## <a name="release">Release Notes</a>
 
-### Release 2.0.6
+### Release 2.0.7
 
- - Fixed bug in special character decoding that chopped of text.
-
- - Limit depth on dump to 1000 to avoid core dump on circular references if the user does not specify circular.
-
- - Handles dumping non-string values for attributes correctly by converting the value to a string.
+ - Fixed DOCTYPE parsing to handle nested '>' characters.
 
 ## <a name="description">Description</a>
 

data/ext/ox/parse.c

@@ -294,32 +294,60 @@ read_instruction(PInfo pi) {
     attr_stack_cleanup(&attrs);
 }
 
+static void
+read_delimited(PInfo pi, char end) {
+    char	c;
+
+    if ('"' == end || '\'' == end) {
+	for (c = *pi->s++; end != c; c = *pi->s++) {
+	    if ('\0' == c) {
+		set_error(&pi->err, "invalid format, dectype not terminated", pi->str, pi->s);
+		return;
+	    }
+	}
+    } else {
+	while (1) {
+	    c = *pi->s++;
+	    if (end == c) {
+		return;
+	    }
+	    switch (c) {
+	    case '\0':
+		set_error(&pi->err, "invalid format, dectype not terminated", pi->str, pi->s);
+		return;
+	    case '"':
+		read_delimited(pi, c);
+		break;
+	    case '\'':
+		read_delimited(pi, c);
+		break;
+	    case '[':
+		read_delimited(pi, ']');
+		break;
+	    case '<':
+		read_delimited(pi, '>');
+		break;
+	    default:
+		break;
+	    }
+	}
+    }
+}
+
 /* Entered after the "<!DOCTYPE" sequence plus the first character after
- * that. Ready to read the rest. Returns error code.
+ * that. Ready to read the rest.
  */
 static void
 read_doctype(PInfo pi) {
     char	*docType;
-    int		depth = 1;
-    char	c;
 
     next_non_white(pi);
     docType = pi->s;
-    while (1) {
-	c = *pi->s++;
-	if ('\0' == c) {
-	    set_error(&pi->err, "invalid format, prolog not terminated", pi->str, pi->s);
-	    return;
-	} else if ('<' == c) {
-	    depth++;
-	} else if ('>' == c) {
-	    depth--;
-	    if (0 == depth) {	/* done, at the end */
-		pi->s--;
-		break;
-	    }
-	}
+    read_delimited(pi, '>');
+    if (err_has(&pi->err)) {
+	return;
     }
+    pi->s--;
     *pi->s = '\0';
     pi->s++;
     if (0 != pi->pcb->add_doctype) {

data/ext/ox/sax.c

@@ -483,23 +483,58 @@ read_instruction(SaxDrive dr) {
     return c;
 }
 
+static char
+read_delimited(SaxDrive dr, char end) {
+    char	c;
+
+    if ('"' == end || '\'' == end) {
+	while (end != (c = buf_get(&dr->buf))) {
+	    if ('\0' == c) {
+		ox_sax_drive_error(dr, NO_TERM "doctype not terminated");
+		return c;
+	    }
+	}
+    } else {
+	while (1) {
+	    c = buf_get(&dr->buf);
+	    if (end == c) {
+		return c;
+	    }
+	    switch (c) {
+	    case '\0':
+		ox_sax_drive_error(dr, NO_TERM "doctype not terminated");
+		return c;
+	    case '"':
+		c = read_delimited(dr, c);
+		break;
+	    case '\'':
+		c = read_delimited(dr, c);
+		break;
+	    case '[':
+		c = read_delimited(dr, ']');
+		break;
+	    case '<':
+		c = read_delimited(dr, '>');
+		break;
+	    default:
+		break;
+	    }
+	}
+    }
+    return c;
+}
+
 /* Entered after the "<!DOCTYPE" sequence. Ready to read the rest.
  */
 static char
 read_doctype(SaxDrive dr) {
-    char        c;
     int		line = dr->buf.line;
     int		col = dr->buf.col - 10;
     char	*s;
 
     buf_backup(&dr->buf); /* back up to the start in case the cdata is empty */
     buf_protect(&dr->buf);
-    while ('>' != (c = buf_get(&dr->buf))) {
-        if ('\0' == c) {
-            ox_sax_drive_error(dr, NO_TERM "doctype not terminated");
-            return c;
-        }
-    }
+    read_delimited(dr, '>');
     if (dr->options.smart && 0 == dr->hints) {
 	for (s = dr->buf.str; is_white(*s); s++) { }
 	if (0 == strncasecmp("HTML", s, 4)) {
@@ -1226,6 +1261,8 @@ ox_sax_collapse_special(SaxDrive dr, char *str, int line, int col) {
 		    b = ox_ucs_to_utf8_chars(b, u);
 #endif
 		} else {
+		    b = ox_ucs_to_utf8_chars(b, u);
+		    /*
 		    ox_sax_drive_error(dr, NO_TERM "Invalid encoding, need UTF-8 encoding to parse &#nnnn; character sequences.");
 		    *b++ = '&';
 		    *b++ = '#';
@@ -1233,6 +1270,7 @@ ox_sax_collapse_special(SaxDrive dr, char *str, int line, int col) {
 			*b++ = x;
 		    }
 		    continue;
+		    */
 		}
 		s = end + 1;
 		continue;

data/lib/ox/version.rb

@@ -1,5 +1,5 @@
 
 module Ox
   # Current version of the module. 
-  VERSION = '2.0.6'
+  VERSION = '2.0.7'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ox
 version: !ruby/object:Gem::Version
-  version: 2.0.6
+  version: 2.0.7
 platform: ruby
 authors:
 - Peter Ohler
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-07-23 00:00:00.000000000 Z
+date: 2013-08-04 00:00:00.000000000 Z
 dependencies: []
 description: "A fast XML parser and object serializer that uses only standard C lib.\n
   \           \nOptimized XML (Ox), as the name implies was written to provide speed