protected virtual void PopTag(string tag)
{
var pop = TagStack.Pop();
if (pop != tag)
{
throw new ArgumentException(string.Format("stack top tag {0} is different with {1}.", pop, tag), "tag");
}
}
protected virtual void PopTag()
{
TagStack.Pop();
}
/// <summary>
/// Simple parsing to check if input fragment is well-formed,
/// HTML elements that do not required end tags (i.e. <BR>)
/// will be ignored by this parser.
/// </summary>
/// <param name="text">
/// text being parsed
/// </param>
internal static bool IsWellFormed(String text)
{
int textPos = 0;
TagStack stack = new TagStack();
StringBuilder builder = new StringBuilder();
for (;;)
{
Match match = null;
// 1: scan for text up to the next tag.
if ((match = _textRegex.Match(text, textPos)).Success)
{
textPos = match.Index + match.Length;
}
// we might be done now
if (textPos == text.Length)
{
while (!stack.IsEmpty())
{
if (!IsEndTagOptional(stack.Pop()))
{
return(false);
}
}
return(true);
}
// First check if it's a unclosed tag (i.e. <mobile:Form >)
if ((match = _unclosedTagRegex.Match(text, textPos)).Success)
{
String startTag = match.Groups["tagname"].Value;
stack.Push(startTag);
}
// Check to see if it's a tag
else if ((match = _tagRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an end tag
else if ((match = _endtagRegex.Match(text, textPos)).Success)
{
String endTag = match.Groups["tagname"].Value;
bool matched = false;
while (!stack.IsEmpty())
{
String startTag = stack.Pop();
if (String.Compare(endTag, startTag, true /*ignoreCase*/, CultureInfo.InvariantCulture) != 0)
{
if (IsEndTagOptional(startTag))
{
continue;
}
// no match against start tag that requires an end tag
return(false);
}
// we found a match here.
matched = true;
break;
}
if (!matched && stack.IsEmpty())
{
return(false);
}
}
// Check to see if it's a directive (i.e. <%@ %> block)
else if ((match = _directiveRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's a server side include
// e.g. <!-- #include file="foo.inc" -->
else if ((match = _includeRegex.Match(text, textPos)).Success)
{
// skip it
}
// Check to see if it's a comment (<%-- --%> block
// e.g. <!-- Blah! -->
else if ((match = _commentRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an asp expression block (i.e. <%= %> block)
else if ((match = _aspExprRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's a databinding expression block (i.e. <%# %> block)
// This does not include <%# %> blocks used as values for
// attributes of server tags.
else if ((match = _databindExprRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an asp code block
else if ((match = _aspCodeRegex.Match(text, textPos)).Success)
{
// skip
}
// Did we process the block that started with a '<'?
if (match == null || !match.Success)
{
// Skip the '<'
textPos++;
}
else
{
textPos = match.Index + match.Length;
}
// we might be done now
if (textPos == text.Length)
{
while (!stack.IsEmpty())
{
if (!IsEndTagOptional(stack.Pop()))
{
return(false);
}
}
return(true);
}
}
}
/// <summary>
/// Parses the given document to find all tags between the beginning of the document and the specified
/// end line.
/// </summary>
/// <remarks>
/// The <paramref name="p_pctCompleteTagCallback" /> is called whenever a complete tag has been parsed.
/// A complete tag is a tag whose opening and closing tags have been found (for example, <b>..</b>).
/// The stack that is returned contains all the unclosed tags found, and so represents where in the
/// document heirarchy the line falls.
/// </remarks>
/// <param name="p_docDocument">The document to parse.</param>
/// <param name="p_intEndLine">The line of the document at which to stop parsing.</param>
/// <param name="p_pctCompleteTagCallback">The method to call whenever a complete tag is parsed.</param>
/// <returns>A stack containing all the unclosed tags found.</returns>
/// <exception cref="ArgumentOutOfRangeException">
/// Thrown if <paramref name="p_intEndLine" /> is greater than
/// or equal to the <see cref="IDocument.TotalNumberOfLines" /> of <paramref name="p_docDocument" />.
/// </exception>
public static TagStack ParseTags(IDocument p_docDocument, Int32 p_intEndLine, ParsedTag p_pctCompleteTagCallback,
ParsedTag p_pctUnclosedTagCallback)
{
if (p_intEndLine >= p_docDocument.TotalNumberOfLines)
{
throw new ArgumentOutOfRangeException("p_intEndLine", p_intEndLine,
"The given end line paramater is outside of the range of lines in the given document.");
}
//parse the buffer
var stkTags = new TagStack();
for (var i = 0; i <= p_intEndLine; i++)
{
var strLine = p_docDocument.GetText(p_docDocument.GetLineSegment(i));
var intLineNum = i;
var intLastOpenPos = strLine.LastIndexOf('<');
if (intLastOpenPos < 0)
{
continue;
}
var intLastClosePos = strLine.LastIndexOf('>');
if ((intLastClosePos > -1) && (intLastOpenPos > intLastClosePos))
{
var stbLines = new StringBuilder(strLine);
//there is an open tag on this line - read lines until it is closed.
for (; i <= p_intEndLine; i++)
{
var strNextLine = p_docDocument.GetText(p_docDocument.GetLineSegment(i));
intLastClosePos = strLine.LastIndexOf('>');
stbLines.Append(strNextLine);
if (intLastClosePos < 0)
{
i--;
break;
}
}
strLine = stbLines.ToString();
}
var mclLineTags = rgxTagContents.Matches(strLine);
foreach (Match mtcTag in mclLineTags)
{
var strTag = mtcTag.Groups[1].Value.Trim();
var strTagName = rgxTagName.Match(strTag).Groups[1].Value;
if (strTag.StartsWith("/"))
{
if (stkTags.Contains(strTagName))
{
while (!stkTags.Peek().Equals(strTagName))
{
var tpsTag = stkTags.Pop();
var tlcStart = new TextLocation(tpsTag.Column, tpsTag.LineNumber);
var tlcEnd = new TextLocation(tpsTag.Column + tpsTag.Name.Length, tpsTag.LineNumber);
if (p_pctUnclosedTagCallback != null)
{
p_pctUnclosedTagCallback(p_docDocument, tpsTag.Name, tlcStart, tlcEnd);
}
}
var tpsCompleteTag = stkTags.Pop();
if (p_pctCompleteTagCallback != null)
{
var tlcStart = new TextLocation(tpsCompleteTag.Column, tpsCompleteTag.LineNumber);
var intEndFoldPos = mtcTag.Groups[1].Index;
var tlcEnd = new TextLocation(intEndFoldPos, intLineNum);
p_pctCompleteTagCallback(p_docDocument, strTagName, tlcStart, tlcEnd);
}
}
}
else
{
if (!strTag.EndsWith("/"))
{
stkTags.Push(strTagName, intLineNum, mtcTag.Groups[1].Index);
}
}
}
}
return(stkTags);
}
/// <summary>
/// Simple parsing to check if input fragment is well-formed,
/// HTML elements that do not required end tags (i.e. <BR>)
/// will be ignored by this parser.
/// </summary>
/// <param name="text">
/// text being parsed
/// </param>
internal static bool IsWellFormed(String text)
{
int textPos = 0;
TagStack stack = new TagStack();
for (;;)
{
Match match = null;
// 1: scan for text up to the next tag.
if ((match = _textRegex.Match(text, textPos)).Success)
{
textPos = match.Index + match.Length;
}
// we might be done now
if (textPos == text.Length)
{
while (!stack.IsEmpty())
{
if (!IsEndTagOptional(stack.Pop()))
{
return false;
}
}
return true;
}
// First check if it's a unclosed tag (i.e. <mobile:Form >)
if ((match = _unclosedTagRegex.Match(text, textPos)).Success)
{
String startTag = match.Groups["tagname"].Value;
stack.Push(startTag);
}
// Check to see if it's a tag
else if ((match = _tagRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an end tag
else if ((match = _endtagRegex.Match(text, textPos)).Success)
{
String endTag = match.Groups["tagname"].Value;
bool matched = false;
while (!stack.IsEmpty())
{
String startTag = stack.Pop();
if (String.Compare(endTag, startTag, StringComparison.OrdinalIgnoreCase) != 0)
{
if (IsEndTagOptional(startTag))
{
continue;
}
// no match against start tag that requires an end tag
return false;
}
// we found a match here.
matched = true;
break;
}
if (!matched && stack.IsEmpty())
{
return false;
}
}
// Check to see if it's a directive (i.e. <%@ %> block)
else if ((match = _directiveRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's a server side include
// e.g. <!-- #include file="foo.inc" -->
else if ((match = _includeRegex.Match(text, textPos)).Success)
{
// skip it
}
// Check to see if it's a comment (<%-- --%> block
// e.g. <!-- Blah! -->
else if ((match = _commentRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an asp expression block (i.e. <%= %> block)
else if ((match = _aspExprRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's a databinding expression block (i.e. <%# %> block)
// This does not include <%# %> blocks used as values for
// attributes of server tags.
else if ((match = _databindExprRegex.Match(text, textPos)).Success)
{
// skip
}
// Check to see if it's an asp code block
else if ((match = _aspCodeRegex.Match(text, textPos)).Success)
{
// skip
}
// Did we process the block that started with a '<'?
if (match == null || !match.Success)
{
// Skip the '<'
textPos++;
}
else
{
textPos = match.Index + match.Length;
}
// we might be done now
if (textPos == text.Length)
{
while (!stack.IsEmpty())
{
if (!IsEndTagOptional(stack.Pop()))
{
return false;
}
}
return true;
}
}
}