/// <summary> Go through <global-tag-attributes> section of the policy file.</summary>
/// <param name="globalAttributeListNode">Top level of <global-tag-attributes>
/// </param>
/// <returns> A HashMap of global Attributes that need validation for every tag.
/// </returns>
/// <throws> PolicyException </throws>
private Hashtable parseGlobalAttributes(XmlNode globalAttributeListNode)
{
XmlNodeList globalAttributeNodes = globalAttributeListNode.SelectNodes("attribute");
Hashtable globalAttributes = new Hashtable();
string _name = "";
//string _value = "";
foreach (XmlNode node in globalAttributeNodes)
{
if (node.NodeType == XmlNodeType.Element)
{
_name = node.Attributes[0].Value;
//_value = node.Attributes[1].Value;
Attribute toAdd = getCommonAttributeByName(_name);
if (toAdd != null)
{
globalAttributes.Add(_name, toAdd);
}
else
{
throw new PolicyException("Global attribute '" + _name + "' was not defined in <common-attributes>");
}
//if (!globalAttributes.ContainsKey(_name))
// globalAttributes.Add(_name, new AntiSamyPattern(_name, _value));
}
}
return(globalAttributes);
}
/// <summary> Private method for parsing the <tag-rules> from the XML file.</summary>
/// <param name="root">The root element for <tag-rules>
/// </param>
/// <returns> A List<Tag> containing the rules.
/// </returns>
/// <throws> PolicyException </throws>
private Hashtable parseTagRules(XmlNode tagAttributeListNode)
{
Hashtable tags = new Hashtable();
XmlNodeList tagList = tagAttributeListNode.SelectNodes("tag");
foreach (XmlNode tagNode in tagList)
{
if (tagNode.NodeType == XmlNodeType.Element)
{
String name = (tagNode.Attributes["name"] == null ? null : tagNode.Attributes["name"].Value);
String action = (tagNode.Attributes["action"] == null ? null : tagNode.Attributes["action"].Value);
Tag tag = new Tag(name);
if (tagNames == null)
tagNames = new ArrayList();
tagNames.Add(name);
tag.Action = action;
//XmlNodeList attributeList = tagNode.SelectNodes("attribute");
XmlNodeList attributeList = tagNode.SelectNodes("attribute");
foreach (XmlNode attributeNode in attributeList)
{
if (!attributeNode.HasChildNodes)
{
Attribute attribute = getCommonAttributeByName(attributeNode.Attributes["name"].Value);
if (attribute != null)
{
String onInvalid = (attributeNode.Attributes["onInvalid"] == null ? null : attributeNode.Attributes["onInvalid"].Value);
String description = (attributeNode.Attributes["description"] == null ? null : attributeNode.Attributes["description"].Value);
if (onInvalid != null && onInvalid.Length > 0)
attribute.OnInvalid = onInvalid;
if (description != null && description.Length > 0)
attribute.Description = description;
tag.addAttribute((org.owasp.validator.html.model.Attribute)attribute.Clone());
}
else
{
//TODO: make this work with .NET
//throw new PolicyException("Attribute '"+XMLUtil.getAttributeValue(attributeNode,"name")+"' was referenced as a common attribute in definition of '"+tag.getName()+"', but does not exist in <common-attributes>");
}
}
else
{
/* Custom attribute for this tag */
Attribute attribute = new Attribute(attributeNode.Attributes["name"].Value);
attribute.OnInvalid = (attributeNode.Attributes["onInvalid"] != null ? attributeNode.Attributes["onInvalid"].Value : null);
attribute.Description = (attributeNode.Attributes["description"] != null ? attributeNode.Attributes["description"].Value : null);
XmlNode regExpListNode = attributeNode.SelectNodes("regexp-list")[0];
if (regExpListNode != null)
{
XmlNodeList regExpList = regExpListNode.SelectNodes("regexp");
foreach (XmlNode regExpNode in regExpList)
{
string regExpName = (regExpNode.Attributes["name"] == null ? null : regExpNode.Attributes["name"].Value);
string value = (regExpNode.Attributes["value"] == null ? null : regExpNode.Attributes["value"].Value);
if (regExpName != null && regExpName.Length > 0)
{
//AntiSamyPattern pattern = getRegularExpression(regExpName);
string pattern = getRegularExpression(regExpName);
if (pattern != null)
attribute.addAllowedRegExp(pattern);
//attribute.addAllowedRegExp(pattern.Pattern);
else
{
throw new PolicyException("Regular expression '" + regExpName + "' was referenced as a common regexp in definition of '" + tag.Name + "', but does not exist in <common-regexp>");
}
}
else if (value != null && value.Length > 0)
{
//TODO: see if I need to reimplement pattern.compile
attribute.addAllowedRegExp(REGEXP_BEGIN + value + REGEXP_END);
}
}
}
XmlNode literalListNode = attributeNode.SelectNodes("literal-list")[0];
if (literalListNode != null)
{
XmlNodeList literalNodes = literalListNode.SelectNodes("literal");
foreach (XmlNode literalNode in literalNodes)
{
string value = (literalNode.Attributes["value"] == null ? null : literalNode.Attributes["value"].Value);
if (value != null && value.Length > 0)
{
attribute.addAllowedValue(value);
}
else if (literalNode.Value != null)
{
attribute.addAllowedValue(literalNode.Value);
}
}
}
tag.addAttribute(attribute);
}
}
tags.Add(name, tag);
}
}
return tags;
}
private void recursiveValidateTag(HtmlNode node)
{
int maxinputsize = int.Parse(policy.getDirective("maxInputSize"));
num++;
HtmlNode parentNode = node.ParentNode;
HtmlNode tmp = null;
string tagName = node.Name;
//check this out
//might not be robust enough
if (tagName.ToLower().Equals("#text")) // || tagName.ToLower().Equals("#comment"))
{
return;
}
Tag tag = policy.getTagByName(tagName.ToLower());
if (tag == null || "filter".Equals(tag.Action))
{
StringBuilder errBuff = new StringBuilder();
if (tagName == null || tagName.Trim().Equals(""))
{
errBuff.Append("An unprocessable ");
}
else
{
errBuff.Append("The <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName.ToLower()) + "</b> ");
}
errBuff.Append("tag has been filtered for security reasons. The contents of the tag will ");
errBuff.Append("remain in place.");
errorMessages.Add(errBuff.ToString());
for (int i = 0; i < node.ChildNodes.Count; i++)
{
tmp = node.ChildNodes[i];
recursiveValidateTag(tmp);
if (tmp.ParentNode == null)
{
i--;
}
}
promoteChildren(node);
return;
}
else if ("validate".Equals(tag.Action))
{
if ("style".Equals(tagName.ToLower()) && policy.getTagByName("style") != null)
{
CssScanner styleScanner = new CssScanner(policy);
try
{
CleanResults cr = styleScanner.scanStyleSheet(node.FirstChild.InnerHtml, maxinputsize);
foreach (string msg in cr.getErrorMessages())
{
errorMessages.Add(msg.ToString());
}
/*
* If IE gets an empty style tag, i.e. <style/>
* it will break all CSS on the page. I wish I
* was kidding. So, if after validation no CSS
* properties are left, we would normally be left
* with an empty style tag and break all CSS. To
* prevent that, we have this check.
*/
if (cr.getCleanHTML() == null || cr.getCleanHTML().Equals(""))
{
//node.getFirstChild().setNodeValue("/* */");
node.FirstChild.InnerHtml = "/* */";
}
else
{
//node.getFirstChild().setNodeValue(cr.getCleanHTML());
node.FirstChild.InnerHtml = cr.getCleanHTML();
}
}
// catch (DomException e)
// {
// addError(ErrorMessageUtil.ERROR_CSS_TAG_MALFORMED, new Object[] { HTMLEntityEncoder.htmlEntityEncode(node.getFirstChild().getNodeValue()) });
// parentNode.removeChild(node);
// }
catch (ScanException e)
{
Console.WriteLine("Scan Exception: " + e.Message);
//addError(ErrorMessageUtil.ERROR_CSS_TAG_MALFORMED, new Object[] { HTMLEntityEncoder.htmlEntityEncode(node.getFirstChild().getNodeValue()) });
parentNode.RemoveChild(node);
}
}
HtmlAttribute attribute = null;
for (int currentAttributeIndex = 0; currentAttributeIndex < node.Attributes.Count; currentAttributeIndex++)
{
attribute = node.Attributes[currentAttributeIndex];
string name = attribute.Name;
string _value = attribute.Value;
Attribute attr = tag.getAttributeByName(name);
if (attr == null)
{
attr = policy.getGlobalAttributeByName(name);
}
bool isAttributeValid = false;
if ("style".Equals(name.ToLower()) && attr != null)
{
CssScanner styleScanner = new CssScanner(policy);
try
{
CleanResults cr = styleScanner.scanInlineStyle(_value, tagName, maxinputsize);
//attribute.setNodeValue(cr.getCleanHTML());
attribute.Value = cr.getCleanHTML();
ArrayList cssScanErrorMessages = cr.getErrorMessages();
foreach (string msg in cr.getErrorMessages())
{
errorMessages.Add(msg.ToString());
}
}
/*
* catch (DOMException e)
* {
*
* addError(ErrorMessageUtil.ERROR_CSS_ATTRIBUTE_MALFORMED, new Object[] { tagName, HTMLEntityEncoder.htmlEntityEncode(node.getNodeValue()) });
*
* ele.removeAttribute(name);
* currentAttributeIndex--;
*
* }
*/
catch (ScanException ex)
{
Console.WriteLine(ex.Message);
//addError(ErrorMessageUtil.ERROR_CSS_ATTRIBUTE_MALFORMED, new Object[] { tagName, HTMLEntityEncoder.htmlEntityEncode(node.getNodeValue()) });
//ele.removeAttribute(name);
currentAttributeIndex--;
}
}
else
{
if (attr != null)
{
//try to find out how robust this is - do I need to do this in a loop?
_value = HtmlEntity.DeEntitize(_value);
foreach (string allowedValue in attr.AllowedValues)
{
if (isAttributeValid)
{
break;
}
if (allowedValue != null && allowedValue.ToLower().Equals(_value.ToLower()))
{
isAttributeValid = true;
}
}
foreach (string ptn in attr.AllowedRegExp)
{
if (isAttributeValid)
{
break;
}
string pattern = "^" + ptn + "$";
Match m = Regex.Match(_value, pattern);
if (m.Success)
{
isAttributeValid = true;
}
}
if (!isAttributeValid)
{
string onInvalidAction = attr.OnInvalid;
StringBuilder errBuff = new StringBuilder();
errBuff.Append("The <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag contained an attribute that we couldn't process. ");
errBuff.Append("The <b>" + HTMLEntityEncoder.htmlEntityEncode(name) + "</b> attribute had a value of <u>" + HTMLEntityEncoder.htmlEntityEncode(_value) + "</u>. ");
errBuff.Append("This value could not be accepted for security reasons. We have chosen to ");
//Console.WriteLine(policy);
if ("removeTag".Equals(onInvalidAction))
{
parentNode.RemoveChild(node);
errBuff.Append("remove the <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag and its contents in order to process this input. ");
}
else if ("filterTag".Equals(onInvalidAction))
{
for (int i = 0; i < node.ChildNodes.Count; i++)
{
tmp = node.ChildNodes[i];
recursiveValidateTag(tmp);
if (tmp.ParentNode == null)
{
i--;
}
}
promoteChildren(node);
errBuff.Append("filter the <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag and leave its contents in place so that we could process this input.");
}
else
{
node.Attributes.Remove(attr.Name);
currentAttributeIndex--;
errBuff.Append("remove the <b>" + HTMLEntityEncoder.htmlEntityEncode(name) + "</b> attribute from the tag and leave everything else in place so that we could process this input.");
}
errorMessages.Add(errBuff.ToString());
if ("removeTag".Equals(onInvalidAction) || "filterTag".Equals(onInvalidAction))
{
return; // can't process any more if we remove/filter the tag
}
}
}
else
{
StringBuilder errBuff = new StringBuilder();
errBuff.Append("The <b>" + HTMLEntityEncoder.htmlEntityEncode(name));
errBuff.Append("</b> attribute of the <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag has been removed for security reasons. ");
errBuff.Append("This removal should not affect the display of the HTML submitted.");
errorMessages.Add(errBuff.ToString());
node.Attributes.Remove(name);
currentAttributeIndex--;
} // end if attribute is or is not found in policy file
} // end if style.equals("name")
} // end while loop through attributes
for (int i = 0; i < node.ChildNodes.Count; i++)
{
tmp = node.ChildNodes[i];
recursiveValidateTag(tmp);
if (tmp.ParentNode == null)
{
i--;
}
}
}
else if ("truncate".Equals(tag.Action))
{
Console.WriteLine("truncate");
HtmlAttributeCollection nnmap = node.Attributes;
while (nnmap.Count > 0)
{
StringBuilder errBuff = new StringBuilder();
errBuff.Append("The <b>" + HTMLEntityEncoder.htmlEntityEncode(nnmap[0].Name));
errBuff.Append("</b> attribute of the <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag has been removed for security reasons. ");
errBuff.Append("This removal should not affect the display of the HTML submitted.");
node.Attributes.Remove(nnmap[0].Name);
errorMessages.Add(errBuff.ToString());
}
HtmlNodeCollection cList = node.ChildNodes;
int i = 0;
int j = 0;
int length = cList.Count;
while (i < length)
{
HtmlNode nodeToRemove = cList[j];
if (nodeToRemove.NodeType != HtmlNodeType.Text && nodeToRemove.NodeType != HtmlNodeType.Comment)
{
node.RemoveChild(nodeToRemove);
}
else
{
j++;
}
i++;
}
}
else
{
errorMessages.Add("The <b>" + HTMLEntityEncoder.htmlEntityEncode(tagName) + "</b> tag has been removed for security reasons.");
parentNode.RemoveChild(node);
}
}
/// <summary> Private method for parsing the <tag-rules> from the XML file.</summary>
/// <param name="root">The root element for <tag-rules>
/// </param>
/// <returns> A List<Tag> containing the rules.
/// </returns>
/// <throws> PolicyException </throws>
private Hashtable parseTagRules(XmlNode tagAttributeListNode)
{
Hashtable tags = new Hashtable();
XmlNodeList tagList = tagAttributeListNode.SelectNodes("tag");
foreach (XmlNode tagNode in tagList)
{
if (tagNode.NodeType == XmlNodeType.Element)
{
String name = (tagNode.Attributes["name"] == null ? null : tagNode.Attributes["name"].Value);
String action = (tagNode.Attributes["action"] == null ? null : tagNode.Attributes["action"].Value);
Tag tag = new Tag(name);
if (tagNames == null)
{
tagNames = new ArrayList();
}
tagNames.Add(name);
tag.Action = action;
//XmlNodeList attributeList = tagNode.SelectNodes("attribute");
XmlNodeList attributeList = tagNode.SelectNodes("attribute");
foreach (XmlNode attributeNode in attributeList)
{
if (!attributeNode.HasChildNodes)
{
Attribute attribute = getCommonAttributeByName(attributeNode.Attributes["name"].Value);
if (attribute != null)
{
String onInvalid = (attributeNode.Attributes["onInvalid"] == null ? null : attributeNode.Attributes["onInvalid"].Value);
String description = (attributeNode.Attributes["description"] == null ? null : attributeNode.Attributes["description"].Value);
if (onInvalid != null && onInvalid.Length > 0)
{
attribute.OnInvalid = onInvalid;
}
if (description != null && description.Length > 0)
{
attribute.Description = description;
}
tag.addAttribute((org.owasp.validator.html.model.Attribute)attribute.Clone());
}
else
{
//TODO: make this work with .NET
//throw new PolicyException("Attribute '"+XMLUtil.getAttributeValue(attributeNode,"name")+"' was referenced as a common attribute in definition of '"+tag.getName()+"', but does not exist in <common-attributes>");
}
}
else
{
/* Custom attribute for this tag */
Attribute attribute = new Attribute(attributeNode.Attributes["name"].Value);
attribute.OnInvalid = (attributeNode.Attributes["onInvalid"] != null ? attributeNode.Attributes["onInvalid"].Value : null);
attribute.Description = (attributeNode.Attributes["description"] != null ? attributeNode.Attributes["description"].Value : null);
XmlNode regExpListNode = attributeNode.SelectNodes("regexp-list")[0];
if (regExpListNode != null)
{
XmlNodeList regExpList = regExpListNode.SelectNodes("regexp");
foreach (XmlNode regExpNode in regExpList)
{
string regExpName = (regExpNode.Attributes["name"] == null ? null : regExpNode.Attributes["name"].Value);
string value = (regExpNode.Attributes["value"] == null ? null : regExpNode.Attributes["value"].Value);
if (regExpName != null && regExpName.Length > 0)
{
//AntiSamyPattern pattern = getRegularExpression(regExpName);
string pattern = getRegularExpression(regExpName);
if (pattern != null)
{
attribute.addAllowedRegExp(pattern);
}
//attribute.addAllowedRegExp(pattern.Pattern);
else
{
throw new PolicyException("Regular expression '" + regExpName + "' was referenced as a common regexp in definition of '" + tag.Name + "', but does not exist in <common-regexp>");
}
}
else if (value != null && value.Length > 0)
{
//TODO: see if I need to reimplement pattern.compile
attribute.addAllowedRegExp(REGEXP_BEGIN + value + REGEXP_END);
}
}
}
XmlNode literalListNode = attributeNode.SelectNodes("literal-list")[0];
if (literalListNode != null)
{
XmlNodeList literalNodes = literalListNode.SelectNodes("literal");
foreach (XmlNode literalNode in literalNodes)
{
string value = (literalNode.Attributes["value"] == null ? null : literalNode.Attributes["value"].Value);
if (value != null && value.Length > 0)
{
attribute.addAllowedValue(value);
}
else if (literalNode.Value != null)
{
attribute.addAllowedValue(literalNode.Value);
}
}
}
tag.addAttribute(attribute);
}
}
tags.Add(name, tag);
}
}
return(tags);
}