[Xml] Add in a strict flag to the parser, on by default. If the flag …

…is off, we can attempt to work around some errors, like tags which are not properly closed (common in HTML docs). This allows us to parse XML docs with unbalanced tags
Whiteknight · Jun 28, 2012 · d02a703 · d02a703
1 parent 4897e48
commit d02a703
Show file tree

Hide file tree

Showing 3 changed files with 30 additions and 9 deletions.
diff --git a/src/unstable/xml/Document.winxed b/src/unstable/xml/Document.winxed
@@ -80,9 +80,9 @@ class Rosella.Xml.Document
     }
 
     // Read the xml from string and parse it into this document
-    function read_from_string(string xml, int do_validate = false)
+    function read_from_string(string xml, int do_validate = false, int strict = true)
     {
-        Rosella.Xml.parse(xml, self, do_validate);
+        Rosella.Xml.parse(xml, self, do_validate, strict);
     }
 
     // Validate this document against its DTD

diff --git a/src/unstable/xml/Parser.winxed b/src/unstable/xml/Parser.winxed
@@ -1,6 +1,6 @@
 namespace Rosella.Xml.Parser
 {
-    function __parse_xml(string xml, var s, var b, int len, var document)
+    function __parse_xml(string xml, var s, var b, int len, var document, int strict)
     {
         var tag_stack = [];
         var current_tag = new Rosella.Xml.Tag.DocumentRoot();
@@ -32,8 +32,29 @@ namespace Rosella.Xml.Parser
                 if (is_close) {
                     if (tag.xmlns != current_tag.xmlns)
                         Rosella.Error.error("Syntax error, mismatched tags. Tag namespaces and/or names do not match at position %d", current_position(len, s, b));
-                    if (tag.name != current_tag.name)
-                        Rosella.Error.error("Syntax error, mismatched tags. Found closing tag '%s' at position %d but current tag is '%s'", tag.name, current_position(len, s, b), current_tag.name);
+                    if (tag.name != current_tag.name) {
+                        if (strict)
+                            Rosella.Error.error("Syntax error, mismatched tags. Found closing tag '%s' at position %d but current tag is '%s'", tag.name, current_position(len, s, b), current_tag.name);
+                        else {
+                            /* If we're not in strict mode, we can try to search
+                               up the tag for the matching open tag and just
+                               deal with it. */
+                            current_tag = null;
+                            while(elements(tag_stack) > 0) {
+                                var dead_tag = tag_stack.pop();
+                                if (dead_tag instanceof Rosella.Xml.Tag.DocumentRoot) {
+                                    current_tag = dead_tag;
+                                    break;
+                                }
+                                if (dead_tag.name == tag.name) {
+                                    dead_tag.set_end(tag);
+                                    current_tag = tag_stack.pop();
+                                    break;
+                                }
+                            }
+                            continue;
+                        }
+                    }
                     current_tag.set_end(tag);
                     current_tag = tag_stack.pop();
                     continue;

diff --git a/src/unstable/xml/Xml.winxed b/src/unstable/xml/Xml.winxed
@@ -17,10 +17,10 @@ namespace Rosella.Xml
     }
 
     // Read XML from a string. Return an Xml.Document object
-    function read_string(string xml)
+    function read_string(string xml, int strict = true)
     {
         var document = new Rosella.Xml.Document();
-        document.read_from_string(xml, false);
+        document.read_from_string(xml, false, strict);
         return document;
     }
 
@@ -45,12 +45,12 @@ namespace Rosella.Xml
 
     // Parse the XML string into the given document. Optionally perform
     // verification on the generated structure
-    function parse(string xml, var document, int do_verify = true)
+    function parse(string xml, var document, int do_verify = true, int strict = true)
     {
         if (document == null)
             document = new Rosella.Xml.Document();
         :(xml, var s, var b, int len) = Rosella.Parse.setup_parse(xml);
-        Rosella.Xml.Parser.__parse_xml(xml, s, b, len, document);
+        Rosella.Xml.Parser.__parse_xml(xml, s, b, len, document, strict);
         if (do_verify)
             document.validate();
         return document;