/**
 * DomParser.js
 *
 * Copyright, Moxiecode Systems AB
 * Released under LGPL License.
 *
 * License: http://www.tinymce.com/license
 * Contributing: http://www.tinymce.com/contributing
 */

(function(tinymce) {
	var Node = tinymce.html.Node;

	/**
	 * This class parses HTML code into a DOM like structure of nodes it will remove redundant whitespace and make
	 * sure that the node tree is valid according to the specified schema. So for example: <p>a<p>b</p>c</p> will become <p>a</p><p>b</p><p>c</p>
	 *
	 * @example
	 * var parser = new tinymce.html.DomParser({validate: true}, schema);
	 * var rootNode = parser.parse('<h1>content</h1>');
	 *
	 * @class tinymce.html.DomParser
	 * @version 3.4
	 */

	/**
	 * Constructs a new DomParser instance.
	 *
	 * @constructor
	 * @method DomParser
	 * @param {Object} settings Name/value collection of settings. comment, cdata, text, start and end are callbacks.
	 * @param {tinymce.html.Schema} schema HTML Schema class to use when parsing.
	 */
	tinymce.html.DomParser = function(settings, schema) {
		var self = this, nodeFilters = {}, attributeFilters = [], matchedNodes = {}, matchedAttributes = {};

		settings = settings || {};
		settings.validate = "validate" in settings ? settings.validate : true;
		settings.root_name = settings.root_name || 'body';
		self.schema = schema = schema || new tinymce.html.Schema();

		function fixInvalidChildren(nodes) {
			var ni, node, parent, parents, newParent, currentNode, tempNode, childNode, i,
				childClone, nonEmptyElements, nonSplitableElements, textBlockElements, sibling, nextNode;

			nonSplitableElements = tinymce.makeMap('tr,td,th,tbody,thead,tfoot,table');
			nonEmptyElements = schema.getNonEmptyElements();
			textBlockElements = schema.getTextBlockElements();

			for (ni = 0; ni < nodes.length; ni++) {
				node = nodes[ni];

				// Already removed or fixed
				if (!node.parent || node.fixed)
					continue;

				// If the invalid element is a text block and the text block is within a parent LI element
				// Then unwrap the first text block and convert other sibling text blocks to LI elements similar to Word/Open Office
				if (textBlockElements[node.name] && node.parent.name == 'li') {
					// Move sibling text blocks after LI element
					sibling = node.next;
					while (sibling) {
						if (textBlockElements[sibling.name]) {
							sibling.name = 'li';
							sibling.fixed = true;
							node.parent.insert(sibling, node.parent);
						} else {
							break;
						}

						sibling = sibling.next;
					}

					// Unwrap current text block
					node.unwrap(node);
					continue;
				}

				// Get list of all parent nodes until we find a valid parent to stick the child into
				parents = [node];
				for (parent = node.parent; parent && !schema.isValidChild(parent.name, node.name) && !nonSplitableElements[parent.name]; parent = parent.parent)
					parents.push(parent);

				// Found a suitable parent
				if (parent && parents.length > 1) {
					// Reverse the array since it makes looping easier
					parents.reverse();

					// Clone the related parent and insert that after the moved node
					newParent = currentNode = self.filterNode(parents[0].clone());

					// Start cloning and moving children on the left side of the target node
					for (i = 0; i < parents.length - 1; i++) {
						if (schema.isValidChild(currentNode.name, parents[i].name)) {
							tempNode = self.filterNode(parents[i].clone());
							currentNode.append(tempNode);
						} else
							tempNode = currentNode;

						for (childNode = parents[i].firstChild; childNode && childNode != parents[i + 1]; ) {
							nextNode = childNode.next;
							tempNode.append(childNode);
							childNode = nextNode;
						}

						currentNode = tempNode;
					}

					if (!newParent.isEmpty(nonEmptyElements)) {
						parent.insert(newParent, parents[0], true);
						parent.insert(node, newParent);
					} else {
						parent.insert(node, parents[0], true);
					}

					// Check if the element is empty by looking through it's contents and special treatment for <p><br /></p>
					parent = parents[0];
					if (parent.isEmpty(nonEmptyElements) || parent.firstChild === parent.lastChild && parent.firstChild.name === 'br') {
						parent.empty().remove();
					}
				} else if (node.parent) {
					// If it's an LI try to find a UL/OL for it or wrap it
					if (node.name === 'li') {
						sibling = node.prev;
						if (sibling && (sibling.name === 'ul' || sibling.name === 'ul')) {
							sibling.append(node);
							continue;
						}

						sibling = node.next;
						if (sibling && (sibling.name === 'ul' || sibling.name === 'ul')) {
							sibling.insert(node, sibling.firstChild, true);
							continue;
						}

						node.wrap(self.filterNode(new Node('ul', 1)));
						continue;
					}

					// Try wrapping the element in a DIV
					if (schema.isValidChild(node.parent.name, 'div') && schema.isValidChild('div', node.name)) {
						node.wrap(self.filterNode(new Node('div', 1)));
					} else {
						// We failed wrapping it, then remove or unwrap it
						if (node.name === 'style' || node.name === 'script')
							node.empty().remove();
						else
							node.unwrap();
					}
				}
			}
		};

		/**
		 * Runs the specified node though the element and attributes filters.
		 *
		 * @param {tinymce.html.Node} Node the node to run filters on.
		 * @return {tinymce.html.Node} The passed in node.
		 */
		self.filterNode = function(node) {
			var i, name, list;

			// Run element filters
			if (name in nodeFilters) {
				list = matchedNodes[name];

				if (list)
					list.push(node);
				else
					matchedNodes[name] = [node];
			}

			// Run attribute filters
			i = attributeFilters.length;
			while (i--) {
				name = attributeFilters[i].name;

				if (name in node.attributes.map) {
					list = matchedAttributes[name];

					if (list)
						list.push(node);
					else
						matchedAttributes[name] = [node];
				}
			}

			return node;
		};

		/**
		 * Adds a node filter function to the parser, the parser will collect the specified nodes by name
		 * and then execute the callback ones it has finished parsing the document.
		 *
		 * @example
		 * parser.addNodeFilter('p,h1', function(nodes, name) {
		 *		for (var i = 0; i < nodes.length; i++) {
		 *			console.log(nodes[i].name);
		 *		}
		 * });
		 * @method addNodeFilter
		 * @method {String} name Comma separated list of nodes to collect.
		 * @param {function} callback Callback function to execute once it has collected nodes.
		 */
		self.addNodeFilter = function(name, callback) {
			tinymce.each(tinymce.explode(name), function(name) {
				var list = nodeFilters[name];

				if (!list)
					nodeFilters[name] = list = [];

				list.push(callback);
			});
		};

		/**
		 * Adds a attribute filter function to the parser, the parser will collect nodes that has the specified attributes
		 * and then execute the callback ones it has finished parsing the document.
		 *
		 * @example
		 * parser.addAttributeFilter('src,href', function(nodes, name) {
		 *		for (var i = 0; i < nodes.length; i++) {
		 *			console.log(nodes[i].name);
		 *		}
		 * });
		 * @method addAttributeFilter
		 * @method {String} name Comma separated list of nodes to collect.
		 * @param {function} callback Callback function to execute once it has collected nodes.
		 */
		self.addAttributeFilter = function(name, callback) {
			tinymce.each(tinymce.explode(name), function(name) {
				var i;

				for (i = 0; i < attributeFilters.length; i++) {
					if (attributeFilters[i].name === name) {
						attributeFilters[i].callbacks.push(callback);
						return;
					}
				}

				attributeFilters.push({name: name, callbacks: [callback]});
			});
		};

		/**
		 * Parses the specified HTML string into a DOM like node tree and returns the result.
		 *
		 * @example
		 * var rootNode = new DomParser({...}).parse('<b>text</b>');
		 * @method parse
		 * @param {String} html Html string to sax parse.
		 * @param {Object} args Optional args object that gets passed to all filter functions.
		 * @return {tinymce.html.Node} Root node containing the tree.
		 */
		self.parse = function(html, args) {
			var parser, rootNode, node, nodes, i, l, fi, fl, list, name, validate,
				blockElements, startWhiteSpaceRegExp, invalidChildren = [], isInWhiteSpacePreservedElement,
				endWhiteSpaceRegExp, allWhiteSpaceRegExp, isAllWhiteSpaceRegExp, whiteSpaceElements, children, nonEmptyElements, rootBlockName;

			args = args || {};
			matchedNodes = {};
			matchedAttributes = {};
			blockElements = tinymce.extend(tinymce.makeMap('script,style,head,html,body,title,meta,param'), schema.getBlockElements());
			nonEmptyElements = schema.getNonEmptyElements();
			children = schema.children;
			validate = settings.validate;
			rootBlockName = "forced_root_block" in args ? args.forced_root_block : settings.forced_root_block;

			whiteSpaceElements = schema.getWhiteSpaceElements();
			startWhiteSpaceRegExp = /^[ \t\r\n]+/;
			endWhiteSpaceRegExp = /[ \t\r\n]+$/;
			allWhiteSpaceRegExp = /[ \t\r\n]+/g;
			isAllWhiteSpaceRegExp = /^[ \t\r\n]+$/;

			function addRootBlocks() {
				var node = rootNode.firstChild, next, rootBlockNode;

				while (node) {
					next = node.next;

					if (node.type == 3 || (node.type == 1 && node.name !== 'p' && !blockElements[node.name] && !node.attr('data-mce-type'))) {
						if (!rootBlockNode) {
							// Create a new root block element
							rootBlockNode = createNode(rootBlockName, 1);
							rootNode.insert(rootBlockNode, node);
							rootBlockNode.append(node);
						} else
							rootBlockNode.append(node);
					} else {
						rootBlockNode = null;
					}

					node = next;
				};
			};

			function createNode(name, type) {
				var node = new Node(name, type), list;

				if (name in nodeFilters) {
					list = matchedNodes[name];

					if (list)
						list.push(node);
					else
						matchedNodes[name] = [node];
				}

				return node;
			};

			function removeWhitespaceBefore(node) {
				var textNode, textVal, sibling;

				for (textNode = node.prev; textNode && textNode.type === 3; ) {
					textVal = textNode.value.replace(endWhiteSpaceRegExp, '');

					if (textVal.length > 0) {
						textNode.value = textVal;
						textNode = textNode.prev;
					} else {
						sibling = textNode.prev;
						textNode.remove();
						textNode = sibling;
					}
				}
			};

			function cloneAndExcludeBlocks(input) {
				var name, output = {};

				for (name in input) {
					if (name !== 'li' && name != 'p') {
						output[name] = input[name];
					}
				}

				return output;
			};

			parser = new tinymce.html.SaxParser({
				validate : validate,

				// Exclude P and LI from DOM parsing since it's treated better by the DOM parser
				self_closing_elements: cloneAndExcludeBlocks(schema.getSelfClosingElements()),

				cdata: function(text) {
					node.append(createNode('#cdata', 4)).value = text;
				},

				text: function(text, raw) {
					var textNode;

					// Trim all redundant whitespace on non white space elements
					if (!isInWhiteSpacePreservedElement) {
						text = text.replace(allWhiteSpaceRegExp, ' ');

						if (node.lastChild && blockElements[node.lastChild.name])
							text = text.replace(startWhiteSpaceRegExp, '');
					}

					// Do we need to create the node
					if (text.length !== 0) {
						textNode = createNode('#text', 3);
						textNode.raw = !!raw;
						node.append(textNode).value = text;
					}
				},

				comment: function(text) {
					node.append(createNode('#comment', 8)).value = text;
				},

				pi: function(name, text) {
					node.append(createNode(name, 7)).value = text;
					removeWhitespaceBefore(node);
				},

				doctype: function(text) {
					var newNode;
		
					newNode = node.append(createNode('#doctype', 10));
					newNode.value = text;
					removeWhitespaceBefore(node);
				},

				start: function(name, attrs, empty) {
					var newNode, attrFiltersLen, elementRule, textNode, attrName, text, sibling, parent;

					elementRule = validate ? schema.getElementRule(name) : {};
					if (elementRule) {
						newNode = createNode(elementRule.outputName || name, 1);
						newNode.attributes = attrs;
						newNode.shortEnded = empty;

						node.append(newNode);

						// Check if node is valid child of the parent node is the child is
						// unknown we don't collect it since it's probably a custom element
						parent = children[node.name];
						if (parent && children[newNode.name] && !parent[newNode.name])
							invalidChildren.push(newNode);

						attrFiltersLen = attributeFilters.length;
						while (attrFiltersLen--) {
							attrName = attributeFilters[attrFiltersLen].name;

							if (attrName in attrs.map) {
								list = matchedAttributes[attrName];

								if (list)
									list.push(newNode);
								else
									matchedAttributes[attrName] = [newNode];
							}
						}

						// Trim whitespace before block
						if (blockElements[name])
							removeWhitespaceBefore(newNode);

						// Change current node if the element wasn't empty i.e not <br /> or <img />
						if (!empty)
							node = newNode;

						// Check if we are inside a whitespace preserved element
						if (!isInWhiteSpacePreservedElement && whiteSpaceElements[name]) {
							isInWhiteSpacePreservedElement = true;
						}
					}
				},

				end: function(name) {
					var textNode, elementRule, text, sibling, tempNode;

					elementRule = validate ? schema.getElementRule(name) : {};
					if (elementRule) {
						if (blockElements[name]) {
							if (!isInWhiteSpacePreservedElement) {
								// Trim whitespace of the first node in a block
								textNode = node.firstChild;
								if (textNode && textNode.type === 3) {
									text = textNode.value.replace(startWhiteSpaceRegExp, '');

									// Any characters left after trim or should we remove it
									if (text.length > 0) {
										textNode.value = text;
										textNode = textNode.next;
									} else {
										sibling = textNode.next;
										textNode.remove();
										textNode = sibling;
									}

									// Remove any pure whitespace siblings
									while (textNode && textNode.type === 3) {
										text = textNode.value;
										sibling = textNode.next;

										if (text.length === 0 || isAllWhiteSpaceRegExp.test(text)) {
											textNode.remove();
											textNode = sibling;
										}

										textNode = sibling;
									}
								}

								// Trim whitespace of the last node in a block
								textNode = node.lastChild;
								if (textNode && textNode.type === 3) {
									text = textNode.value.replace(endWhiteSpaceRegExp, '');

									// Any characters left after trim or should we remove it
									if (text.length > 0) {
										textNode.value = text;
										textNode = textNode.prev;
									} else {
										sibling = textNode.prev;
										textNode.remove();
										textNode = sibling;
									}

									// Remove any pure whitespace siblings
									while (textNode && textNode.type === 3) {
										text = textNode.value;
										sibling = textNode.prev;

										if (text.length === 0 || isAllWhiteSpaceRegExp.test(text)) {
											textNode.remove();
											textNode = sibling;
										}

										textNode = sibling;
									}
								}
							}

							// Trim start white space
							// Removed due to: #5424
							/*textNode = node.prev;
							if (textNode && textNode.type === 3) {
								text = textNode.value.replace(startWhiteSpaceRegExp, '');

								if (text.length > 0)
									textNode.value = text;
								else
									textNode.remove();
							}*/
						}

						// Check if we exited a whitespace preserved element
						if (isInWhiteSpacePreservedElement && whiteSpaceElements[name]) {
							isInWhiteSpacePreservedElement = false;
						}

						// Handle empty nodes
						if (elementRule.removeEmpty || elementRule.paddEmpty) {
							if (node.isEmpty(nonEmptyElements)) {
								if (elementRule.paddEmpty)
									node.empty().append(new Node('#text', '3')).value = '\u00a0';
								else {
									// Leave nodes that have a name like <a name="name">
									if (!node.attributes.map.name && !node.attributes.map.id) {
										tempNode = node.parent;
										node.empty().remove();
										node = tempNode;
										return;
									}
								}
							}
						}

						node = node.parent;
					}
				}
			}, schema);

			rootNode = node = new Node(args.context || settings.root_name, 11);

			parser.parse(html);

			// Fix invalid children or report invalid children in a contextual parsing
			if (validate && invalidChildren.length) {
				if (!args.context)
					fixInvalidChildren(invalidChildren);
				else
					args.invalid = true;
			}

			// Wrap nodes in the root into block elements if the root is body
			if (rootBlockName && rootNode.name == 'body')
				addRootBlocks();

			// Run filters only when the contents is valid
			if (!args.invalid) {
				// Run node filters
				for (name in matchedNodes) {
					list = nodeFilters[name];
					nodes = matchedNodes[name];

					// Remove already removed children
					fi = nodes.length;
					while (fi--) {
						if (!nodes[fi].parent)
							nodes.splice(fi, 1);
					}

					for (i = 0, l = list.length; i < l; i++)
						list[i](nodes, name, args);
				}

				// Run attribute filters
				for (i = 0, l = attributeFilters.length; i < l; i++) {
					list = attributeFilters[i];

					if (list.name in matchedAttributes) {
						nodes = matchedAttributes[list.name];

						// Remove already removed children
						fi = nodes.length;
						while (fi--) {
							if (!nodes[fi].parent)
								nodes.splice(fi, 1);
						}

						for (fi = 0, fl = list.callbacks.length; fi < fl; fi++)
							list.callbacks[fi](nodes, list.name, args);
					}
				}
			}

			return rootNode;
		};

		// Remove <br> at end of block elements Gecko and WebKit injects BR elements to
		// make it possible to place the caret inside empty blocks. This logic tries to remove
		// these elements and keep br elements that where intended to be there intact
		if (settings.remove_trailing_brs) {
			self.addNodeFilter('br', function(nodes, name) {
				var i, l = nodes.length, node, blockElements = tinymce.extend({}, schema.getBlockElements()),
					nonEmptyElements = schema.getNonEmptyElements(), parent, lastParent, prev, prevName;

				// Remove brs from body element as well
				blockElements.body = 1;

				// Must loop forwards since it will otherwise remove all brs in <p>a<br><br><br></p>
				for (i = 0; i < l; i++) {
					node = nodes[i];
					parent = node.parent;

					if (blockElements[node.parent.name] && node === parent.lastChild) {
						// Loop all nodes to the left of the current node and check for other BR elements
						// excluding bookmarks since they are invisible
						prev = node.prev;
						while (prev) {
							prevName = prev.name;

							// Ignore bookmarks
							if (prevName !== "span" || prev.attr('data-mce-type') !== 'bookmark') {
								// Found a non BR element
								if (prevName !== "br")
									break;
	
								// Found another br it's a <br><br> structure then don't remove anything
								if (prevName === 'br') {
									node = null;
									break;
								}
							}

							prev = prev.prev;
						}

						if (node) {
							node.remove();

							// Is the parent to be considered empty after we removed the BR
							if (parent.isEmpty(nonEmptyElements)) {
								elementRule = schema.getElementRule(parent.name);

								// Remove or padd the element depending on schema rule
								if (elementRule) {
									if (elementRule.removeEmpty)
										parent.remove();
									else if (elementRule.paddEmpty)
										parent.empty().append(new tinymce.html.Node('#text', 3)).value = '\u00a0';
								}
							}
						}
					} else {
						// Replaces BR elements inside inline elements like <p><b><i><br></i></b></p> so they become <p><b><i>&nbsp;</i></b></p> 
						lastParent = node;
						while (parent.firstChild === lastParent && parent.lastChild === lastParent) {
							lastParent = parent;

							if (blockElements[parent.name]) {
								break;
							}

							parent = parent.parent;
						}

						if (lastParent === parent) {
							textNode = new tinymce.html.Node('#text', 3);
							textNode.value = '\u00a0';
							node.replace(textNode);
						}
					}
				}
			});
		}

		// Force anchor names closed, unless the setting "allow_html_in_named_anchor" is explicitly included.
		if (!settings.allow_html_in_named_anchor) {
			self.addAttributeFilter('id,name', function(nodes, name) {
				var i = nodes.length, sibling, prevSibling, parent, node;

				while (i--) {
					node = nodes[i];
					if (node.name === 'a' && node.firstChild && !node.attr('href')) {
						parent = node.parent;

						// Move children after current node
						sibling = node.lastChild;
						do {
							prevSibling = sibling.prev;
							parent.insert(sibling, node);
							sibling = prevSibling;
						} while (sibling);
					}
				}
			});
		}
	}
})(tinymce);