/// <summary> /// Parses an AOML string into open, close and content nodes /// </summary> /// <param name="aoml">The aoml string to be parsed</param> public NodeCollection Parse(string aoml) { NodeCollection nodes = new NodeCollection(); // Create regular expression instance if (this.regex == null) { this.regex = new Regex( "[<]\\s*(?<closer>[/]?)\\s*" + "(?<name>[a-z]+)" + "(\\s+((?<attribute>[a-z_\\-]+)\\s*" + "(=\\s*(" + /**/ (Mode == ParserMode.Strict ? "" : "(?<value>[^\\s<>'\"]+)|") + /**/ "[\"](?<value>[^\"]*)[\"]|" + /**/ (Mode == ParserMode.Compatibility ? "['](?<value>[^'>]*)[']?)|" : "['](?<value>[^']*)['])|") + /**/ "(?<value>)" + ")?)\\s*)*" + "\\s*(?<closed>[/]?)[>]", RegexOptions.IgnoreCase | RegexOptions.Compiled | RegexOptions.Multiline); } // Some setup int regexCloser = this.regex.GroupNumberFromName("closer"); int regexName = this.regex.GroupNumberFromName("name"); int regexAttribute = this.regex.GroupNumberFromName("attribute"); int regexValue = this.regex.GroupNumberFromName("value"); int regexClosed = this.regex.GroupNumberFromName("closed"); // Convert to unix line endings aoml = aoml.Replace("\r\n", "\n"); // Parse AOML MatchCollection matches = this.regex.Matches(aoml); int offset = 0; foreach (Match match in matches) { // Extract content if (match.Index > offset) { string content = aoml.Substring(offset, match.Index - offset); AddContent(nodes, content, content); } offset = match.Index + match.Length; // Extract tag string raw = match.Groups[0].Value; string name = match.Groups[regexName].Value; if (string.IsNullOrEmpty(name)) { // No name, let's assume it's just text AddContent(nodes, raw, raw); continue; } bool closer = !string.IsNullOrEmpty(match.Groups[regexCloser].Value); bool closed = !string.IsNullOrEmpty(match.Groups[regexClosed].Value); if (closer) { // Closing tag nodes.Add(new CloseNode(name, raw)); } else { // Opening tag OpenNode node = new OpenNode(name, raw, closed); // Add attributes for (int i = 0; i < match.Groups[regexAttribute].Captures.Count; i++) { node.AddAttribute( match.Groups[regexAttribute].Captures[i].Value, match.Groups[regexValue].Captures[i].Value); } nodes.Add(node); } } // Remainder is content if (offset < aoml.Length) { string content = aoml.Substring(offset); AddContent(nodes, content, content); } return(nodes); }
/// <summary> /// Balances out the stream of nodes to form a valid tree. /// This process will always try to resolve conflicts by introducing extra inline elements over block elements. /// </summary> public void Balance(NodeCollection nodes) { if (nodes == null) { throw new ArgumentNullException(); } // Fill in missing elements List <string> nameStack = new List <string>(); nodes.Reset(); Node node = null; while ((node = nodes.Next()) != null) { switch (node.Type) { case NodeType.Open: OpenNode openNode = (OpenNode)node; if (openNode.Closed) { continue; } nameStack.Insert(0, openNode.Name); break; case NodeType.Close: CloseNode closeNode = (CloseNode)node; if (!nameStack.Contains(closeNode.Name)) { // Found CloseNode without OpenNode // Fix by insert an OpenNode before it nodes.InsertBefore( closeNode, new OpenNode(closeNode.Name, "", false)); } nameStack.Remove(closeNode.Name); break; } } while (nameStack.Count > 0) { // Found OpenNode without CloseNode // Fix by adding a CloseNode at the end of the stream string name = nameStack[0]; nameStack.RemoveAt(0); nodes.Add(new CloseNode(name, "")); } // Resolve tree structure List <OpenNode> nodeStack = new List <OpenNode>(); nodes.Reset(); while ((node = nodes.Next()) != null) { switch (node.Type) { case NodeType.Open: OpenNode on = (OpenNode)node; if (on.Closed) { continue; } nodeStack.Insert(0, on); break; case NodeType.Close: // A bit of setup CloseNode closeNode = (CloseNode)node; bool block = blockElements.Contains(closeNode.Name); // Find matching OpenNode OpenNode openNode = null; foreach (OpenNode o in nodeStack) { if (o.Name == closeNode.Name) { openNode = o; break; } } if (openNode == null) { throw new InvalidOperationException("Unable to find matching OpenNode to CloseNode"); } // Handle incorrect balance int offset = 0; while (nodeStack[offset].Name != closeNode.Name) { OpenNode n = nodeStack[offset]; if (block) { nodes.InsertBefore(node, new CloseNode(n.Name, "")); nodes.InsertAfter(node, n.Clone()); nodeStack.RemoveAt(offset); offset--; } else { nodes.InsertBefore(n, closeNode.Clone()); nodes.InsertAfter(n, openNode.Clone()); } offset++; } // All fixed nodeStack.RemoveAt(offset); break; } } // And we're done! nodes.Reset(); }