/// <summary>
/// Parses the text of an HTML tag and updates the <see cref="Name"/> and
/// <see cref="AttributesPart"/> properties. <see cref="ITextParser"/>
/// <paramref name="tp"/> should be pointed at the '<' character that
/// starts an HTML tag.
/// </summary>
public bool Parse(ITextParser tp)
{
if (tp.Peek() != '<')
{
return(false);
}
tp.MoveAhead();
if (tp.Peek() == '/')
{
tp.MoveAhead();
IsEndTag = true;
}
if (tp.EndOfText)
{
return(false);
}
if (!ParseName(tp))
{
return(false);
}
if (!ParseAttributes(tp))
{
return(false);
}
return(true);
}
/// <summary>
/// Gets a named entity starting at the currrent
/// character ('&') and proceeding until ';'.
/// Returns null if the substring does not match
/// the expected entity sequence.
/// </summary>
private static string GetNamedEntity(ITextParser tp)
{
const string kValidCharacters = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ";
var startPos = tp.Position + 1;
var offset = 1;
char c;
while (true)
{
c = tp.Peek(offset);
if (kValidCharacters.IndexOf(c) == -1)
{
break;
}
offset++;
}
if (c != ';')
{
return(null);
}
var length = offset - 1;
if (length < 1)
{
return(null);
}
return(tp.Substring(startPos, length));
}
/// <summary>
/// Returns true if the substring starting at
/// the current position and continuing to
/// the next ';' contains a decimal or hex
/// entity sequence. On entry, tp should be
/// pointing at '&'.
/// </summary>
private static bool EntityIsHexOrDecimal(ITextParser tp)
{
// Handle hex and numeric
if (tp.Peek(1) == '#')
{
var c = tp.Peek(2);
if (c == 'x' || c == 'X')
{
return(EntityContainsValidCharacters(tp, 3, hex));
}
else
{
return(EntityContainsValidCharacters(tp, 2, digits));
}
}
return(false);
}
/// <summary>
/// Parses the element name from the tag using the given
/// <see cref="ITextParser"/> <paramref name="tp"/>. The
/// <see cref="ITextParser"/> position should be set to
/// the first character of the tag following the "<",
/// or following the "</" for end tags.
/// Returns true if a syntactically valid name is found.
/// </summary>
private bool ParseName(ITextParser tp)
{
const string nameCharacters = "abcdefghijklmnopqrstuvwxyz"
+ "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789:";
var startPos = tp.Position;
var offset = 0;
char c;
while ((c = tp.Peek(offset)) != TextParser.NullChar)
{
if (nameCharacters.IndexOf(c) == -1)
{
break;
}
offset++;
}
// Did we get any valid characters?
if (offset < 1)
{
return(false);
}
// Does the tag name end properly?
if (c != '>' && c != '/' && !Char.IsWhiteSpace(c))
{
return(false);
}
if (tp.Peek(offset - 1) == ':')
{
return(false);
}
// Our minimal validation has passed...
tp.MoveAhead(offset);
var length = tp.Position - startPos;
Name = tp.Substring(startPos, length).ToLower();
// Force void elements to be self-closing tags.
IsSelfClosingTag = IsVoidElement(Name);
return(true);
}
string ExtractBlock(
ITextParser parser,
char openChar,
char closeChar)
{
// Track delimiter depth
var depth = 1;
// Extract characters between delimiters
parser.MoveAhead();
var start = parser.Position;
while (!parser.EndOfText)
{
if (parser.Peek() == openChar)
{
// Increase block depth
depth++;
}
else if (parser.Peek() == closeChar)
{
// Decrease block depth
depth--;
// Test for end of block
if (depth == 0)
{
break;
}
}
else if (parser.Peek() == '"')
{
// Don't count delimiters within quoted text
ExtractQuote(parser);
}
// Move to next character
parser.MoveAhead();
}
return(parser.Extract(start, parser.Position));
}
string ExtractQuote(ITextParser parser)
{
// Extract contents of quote
parser.MoveAhead();
var start = parser.Position;
while (!parser.EndOfText && parser.Peek() != '"')
{
parser.MoveAhead();
}
return(parser.Extract(start, parser.Position));
}
/// <summary>
/// Appends the entity substring that begins at
/// the current location, which is assumed to
/// be '&', to the output.
/// </summary>
private static void AppendEntityToOutput(ITextParser tp, StringBuilder sb)
{
while (!tp.EndOfText)
{
var c = tp.Peek();
sb.Append(c);
tp.MoveAhead();
if (c == ';')
{
break;
}
}
}
/// <summary>
/// Returns true if the substring starting at offset
/// and continuing to the next ';' contains characters
/// from "validChars" only.
/// </summary>
private static bool EntityContainsValidCharacters(ITextParser tp, int offset, string validChars)
{
char c;
while ((c = tp.Peek(offset++)) != TextParser.NullChar)
{
if (validChars.IndexOf(c) == -1)
{
return(false);
}
if (c == ';')
{
return(true);
}
}
return(false);
}
private void Converter()
{
while (!tp.EndOfText)
{
var c = tp.Peek();
switch (c)
{
case '<':
var tag = GetTag();
if (tag != null)
{
HandleTag(tag);
}
break;
case '>':
xml.Append(">");
tp.MoveAhead();
break;
case '&':
EntityConverter.Convert(tp, xml);
break;
default:
xml.Append(c);
tp.MoveAhead();
break;
}
}
CloseOpenElements(openElements.Count);
}
/// <summary>
/// Parses attributes from the tag using the given
/// <see cref="ITextParser"/> <paramref name="tp"/>. The
/// <see cref="ITextParser"/> position should be set to
/// the first character of the tag following the element name.
/// </summary>
private bool ParseAttributes(ITextParser tp)
{
const char kDoubleQuote = '"';
const char kSingleQuote = '\'';
const string kDoubleQuoteEntity = "";
if (tp.Peek() == '>')
{
tp.MoveAhead();
return(true);
}
var sb = new StringBuilder();
// Copy current input character
void Copy()
{
sb.Append(tp.Peek());
tp.MoveAhead();
}
// Copy input characters until fence character or end of tag
void CopyTo(char fence)
{
while (!tp.EndOfText)
{
var c = tp.Peek();
if (c == fence || c == '>')
{
break;
}
if (c != kDoubleQuote)
{
sb.Append(c);
}
else
{
sb.Append(kDoubleQuoteEntity);
}
tp.MoveAhead();
}
}
// Copy attributes
var startPos = tp.Position;
while (!tp.EndOfText)
{
var c = tp.Peek();
if (c == '>' || c == '<')
{
break;
}
switch (c)
{
case '=':
Copy();
c = tp.Peek();
if (c == kDoubleQuote)
{
// Copy double-quoted value
Copy();
CopyTo(kDoubleQuote);
sb.Append(kDoubleQuote);
tp.MoveAhead();
}
else if (c == kSingleQuote)
{
// Copy single-quoted value, but with double-quotes
sb.Append(kDoubleQuote);
tp.MoveAhead();
CopyTo(kSingleQuote);
sb.Append(kDoubleQuote);
tp.MoveAhead();
}
else
{
// Copy unqouted value adding double-quotes
sb.Append(kDoubleQuote);
CopyTo(' ');
sb.Append(kDoubleQuote);
}
break;
default:
Copy();
break;
}
}
if (tp.Peek() != '>')
{
return(false);
}
if (tp.CharAt(tp.Position - 1) == '/')
{
IsSelfClosingTag = true;
sb.Length = sb.Length - 1;
}
AttributesPart = sb.ToString();
if (AttributesPart.IndexOf('&') != -1)
{
AttributesPart = ResolveEntities(AttributesPart);
}
tp.MoveAhead();
return(true);
}
