/*
This is a major clean up to strip out all the extra stuff you get
when you save as web page from Word 2000. It doesn't yet know what
to do with VML tags, but these will appear as errors unless you
declare them as new tags, such as o:p which needs to be declared
as inline.
*/
public virtual void CleanWord2000(Lexer lexer, Node node)
{
/* used to a list from a sequence of bulletted p's */
Node list = null;
while (node != null)
{
/* discard Word's style verbiage */
if (node.Tag == _tt.TagStyle || node.Tag == _tt.TagMeta || node.Type == Node.COMMENT_TAG)
{
node = Node.DiscardElement(node);
continue;
}
/* strip out all span tags Word scatters so liberally! */
if (node.Tag == _tt.TagSpan)
{
node = StripSpan(lexer, node);
continue;
}
/* get rid of Word's xmlns attributes */
if (node.Tag == _tt.TagHtml)
{
/* check that it's a Word 2000 document */
if (node.GetAttrByName("xmlns:o") == null)
{
return;
}
}
if (node.Tag == _tt.TagLink)
{
AttVal attr = node.GetAttrByName("rel");
if (attr != null && attr.Val != null && attr.Val.Equals("File-List"))
{
node = Node.DiscardElement(node);
continue;
}
}
/* discard empty paragraphs */
if (node.Content == null && node.Tag == _tt.TagP)
{
node = Node.DiscardElement(node);
continue;
}
if (node.Tag == _tt.TagP)
{
AttVal attr = node.GetAttrByName("class");
/* map sequence of <p class="MsoListBullet"> to <ul>...</ul> */
if (attr != null && attr.Val != null && attr.Val.Equals("MsoListBullet"))
{
Node.CoerceNode(lexer, node, _tt.TagLi);
if (list == null || list.Tag != _tt.TagUl)
{
list = lexer.InferredTag("ul");
Node.InsertNodeBeforeElement(node, list);
}
PurgeAttributes(node);
if (node.Content != null)
{
CleanWord2000(lexer, node.Content);
}
/* remove node and append to contents of list */
Node.RemoveNode(node);
Node.InsertNodeAtEnd(list, node);
node = list.Next;
}
else if (attr != null && attr.Val != null && attr.Val.Equals("Code"))
{
/* map sequence of <p class="Code"> to <pre>...</pre> */
Node br = lexer.NewLineNode();
NormalizeSpaces(node);
if (list == null || list.Tag != _tt.TagPre)
{
list = lexer.InferredTag("pre");
Node.InsertNodeBeforeElement(node, list);
}
/* remove node and append to contents of list */
Node.RemoveNode(node);
Node.InsertNodeAtEnd(list, node);
StripSpan(lexer, node);
Node.InsertNodeAtEnd(list, br);
node = list.Next;
}
else
{
list = null;
}
}
else
{
list = null;
}
/* strip out style and class attributes */
if (node.Type == Node.START_TAG || node.Type == Node.START_END_TAG)
{
PurgeAttributes(node);
}
if (node.Content != null)
{
CleanWord2000(lexer, node.Content);
}
node = node.Next;
}
}