// // System.Xml.XmlTextReader // // Author: // Jason Diamond (jason@injektilo.org) // Adam Treat (manyoso@yahoo.com) // Atsushi Enomoto (ginga@kit.hi-ho.ne.jp) // // (C) 2001, 2002 Jason Diamond http://injektilo.org/ // // FIXME: // // NameTables aren't being used completely yet. // // Some thought needs to be given to performance. There's too many // strings being allocated. // // If current node is on an Attribute, Prefix might be null, and // in several fields which uses XmlReader, it should be considered. // using System; using System.Collections; using System.Collections.Specialized; using System.IO; using System.Text; using System.Xml.Schema; using Mono.Xml; using Mono.Xml.Native; namespace System.Xml { public class XmlTextReader : XmlReader, IXmlLineInfo { #region Constructors protected XmlTextReader () { } public XmlTextReader (Stream input) : this (new XmlStreamReader (input)) { } public XmlTextReader (string url) : this(url, new NameTable ()) { } public XmlTextReader (TextReader input) : this (input, new NameTable ()) { } protected XmlTextReader (XmlNameTable nt) : this (String.Empty, null, XmlNodeType.None, null) { } public XmlTextReader (Stream input, XmlNameTable nt) : this(new XmlStreamReader (input), nt) { } public XmlTextReader (string url, Stream input) : this (url, new XmlStreamReader (input)) { } public XmlTextReader (string url, TextReader input) : this (url, input, new NameTable ()) { } public XmlTextReader (string url, XmlNameTable nt) : this (url, new XmlStreamReader (url, null, null), nt) { } public XmlTextReader (TextReader input, XmlNameTable nt) : this (String.Empty, input, nt) { } public XmlTextReader (Stream xmlFragment, XmlNodeType fragType, XmlParserContext context) : this (context != null ? context.BaseURI : String.Empty, new XmlStreamReader (xmlFragment), fragType, context) { } public XmlTextReader (string url, Stream input, XmlNameTable nt) : this (url, new XmlStreamReader (input), nt) { } public XmlTextReader (string url, TextReader input, XmlNameTable nt) : this (url, input, XmlNodeType.Document, null) { } public XmlTextReader (string xmlFragment, XmlNodeType fragType, XmlParserContext context) : this (context != null ? context.BaseURI : String.Empty, new StringReader (xmlFragment), fragType, context) { } XmlTextReader (string url, TextReader fragment, XmlNodeType fragType, XmlParserContext context) { InitializeContext (url, context, fragment, fragType); } #endregion #region Properties public override int AttributeCount { get { return attributes.Count; } } public override string BaseURI { get { return parserContext.BaseURI; } } public override int Depth { get { if (NodeType == XmlNodeType.Attribute) return elementDepth + 1; else if (insideAttribute) return elementDepth + 2; // inside attribute value. return elementDepth; } } public Encoding Encoding { get { return parserContext.Encoding; } } public override bool EOF { get { return readState == ReadState.EndOfFile || readState == ReadState.Closed; } } public override bool HasValue { get { if (this.valueBuilderAvailable) return valueBuilder.Length != 0; else return value != String.Empty; } } public override bool IsDefault { get { // XmlTextReader does not expand default attributes. return false; } } public override bool IsEmptyElement { get { return isEmptyElement; } } public override string this [int i] { get { return GetAttribute (i); } } public override string this [string name] { get { return GetAttribute (name); } } public override string this [string localName, string namespaceName] { get { return GetAttribute (localName, namespaceName); } } public int LineNumber { get { return currentInput.LineNumber; } } public int LinePosition { get { return currentInput.LinePosition; } } public override string LocalName { get { return localName; } } public override string Name { get { return name; } } public bool Namespaces { get { return namespaces; } set { if (readState != ReadState.Initial) throw new InvalidOperationException ("Namespaces have to be set before reading."); namespaces = value; } } public override string NamespaceURI { get { return namespaceURI; } } public override XmlNameTable NameTable { get { return parserContext.NameTable; } } public override XmlNodeType NodeType { get { return nodeType; } } [MonoTODO] public bool Normalization { get { return normalization; } set { normalization = value; } } public override string Prefix { get { return prefix; } } public override char QuoteChar { get { // value string holds attribute quotation char. if (NodeType == XmlNodeType.Attribute) return value [0]; else return '"'; } } public override ReadState ReadState { get { return readState; } } public override string Value { get { string v = value; if (valueBuilderAvailable) v = valueBuilder.ToString (); if(NodeType == XmlNodeType.Attribute) return UnescapeAttributeValue(v); else return v; } } public WhitespaceHandling WhitespaceHandling { get { return whitespaceHandling; } set { whitespaceHandling = value; } } public override string XmlLang { get { return parserContext.XmlLang; } } public XmlResolver XmlResolver { set { resolver = value; } } public override XmlSpace XmlSpace { get { return parserContext.XmlSpace; } } #endregion #region Methods public override void Close () { readState = ReadState.Closed; foreach (XmlParserInput input in parserInputStack.ToArray ()) input.Close (); this.currentInput.Close (); } public override string GetAttribute (int i) { if (i > attributes.Count) throw new ArgumentOutOfRangeException ("i is smaller than AttributeCount"); else return UnescapeAttributeValue (attributes [orderedAttributes [i]] as string); } // MS.NET 1.0 documentation says that this method returns String.Empty for // not-exist attribute, but in fact it returns null. // That description is corrected in MS.NET 1.1 documentation. public override string GetAttribute (string name) { return UnescapeAttributeValue (attributes [name] as string); } private int GetIndexOfQualifiedAttribute (string localName, string namespaceURI) { for(int i = 0; i < orderedAttributes.Count; i++) { string thisName = (string) orderedAttributes [i]; int indexOfColon = thisName.IndexOf (':'); if (indexOfColon != -1) { string thisLocalName = thisName.Substring (indexOfColon + 1); if (localName == thisLocalName) { string thisPrefix = thisName.Substring (0, indexOfColon); string thisNamespaceURI = LookupNamespace (thisPrefix); if (namespaceURI == thisNamespaceURI) return i; } } else if (localName == "xmlns" && namespaceURI == "http://www.w3.org/2000/xmlns/" && thisName == "xmlns") return i; } return -1; } internal XmlParserContext GetInternalParserContext () { return parserContext; } public override string GetAttribute (string localName, string namespaceURI) { int idx = this.GetIndexOfQualifiedAttribute (localName, namespaceURI); if (idx < 0) return null; return UnescapeAttributeValue (attributes [orderedAttributes [idx]] as string); } [MonoTODO] public TextReader GetRemainder () { throw new NotImplementedException (); } bool IXmlLineInfo.HasLineInfo () { return true; } public override string LookupNamespace (string prefix) { return parserContext.NamespaceManager.LookupNamespace (prefix); } public override void MoveToAttribute (int i) { MoveToElement (); if (attributes == null || orderedAttributes.Count < i || i < 0) throw new ArgumentOutOfRangeException ("attribute index out of range."); if (orderedAttributesPos == -1) { SaveProperties (); } orderedAttributesPos = i; string name = orderedAttributes [i] as string; string value = attributes [name] as string; SetProperties ( XmlNodeType.Attribute, // nodeType name, // name false, // isEmptyElement value, // value false // clearAttributes ); insideAttribute = true; attributeValuePos = 0; } public override bool MoveToAttribute (string name) { MoveToElement (); bool match = false; if (attributes == null) return false; if (orderedAttributesPos == -1) { SaveProperties (); } for (orderedAttributesPos = 0; orderedAttributesPos < orderedAttributes.Count; orderedAttributesPos++) if (name == orderedAttributes [orderedAttributesPos] as string) { match = true; break; } if (match) { string value = attributes [name] as string; SetProperties ( XmlNodeType.Attribute, // nodeType name, // name false, // isEmptyElement value, // value false // clearAttributes ); insideAttribute = true; attributeValuePos = 0; } return match; } public override bool MoveToAttribute (string localName, string namespaceName) { MoveToElement (); if (attributes == null) return false; int idx = GetIndexOfQualifiedAttribute (localName, namespaceName); if (idx < 0) return false; MoveToAttribute (idx); return true; } public override bool MoveToElement () { if (orderedAttributesPos != -1) { orderedAttributesPos = -1; if (isPropertySaved) RestoreProperties (); insideAttribute = false; return true; } return false; } public override bool MoveToFirstAttribute () { MoveToElement (); return MoveToNextAttribute (); } public override bool MoveToNextAttribute () { if (attributes == null) return false; if (orderedAttributesPos == -1) SaveProperties (); if (++orderedAttributesPos < orderedAttributes.Count) { string name = orderedAttributes [orderedAttributesPos] as string; string value = attributes [name] as string; SetProperties ( XmlNodeType.Attribute, // nodeType name, // name false, // isEmptyElement value, // value false // clearAttributes ); insideAttribute = true; attributeValuePos = 0; return true; } return false; } public override bool Read () { bool more = false; isPropertySaved = false; readState = ReadState.Interactive; insideAttribute = false; // It was moved from end of ReadStartTag (). if (depthUp) ++depth; depthUp = false; more = ReadContent (); if (depth == 0 && !allowMultipleRoot && (IsEmptyElement || NodeType == XmlNodeType.EndElement)) currentState = XmlNodeType.EndElement; if (maybeTextDecl != 0) maybeTextDecl--; if (!more && startNodeType == XmlNodeType.Document && currentState != XmlNodeType.EndElement) throw new XmlException ("Document element did not appear."); return more; } public override bool ReadAttributeValue () { // 'attributeString' holds real string value (without their // quotation characters). // // 'attributeValuePos' holds current position // of 'attributeString' while iterating ReadAttribute(). // It may be: // -1 if ReadAttributeValue() has already finished. // 0 if ReadAttributeValue() ready to start reading. // >0 if ReadAttributeValue() already got 1 or more values // // local 'refPosition' holds the position on the // attributeString which may be used next time. if (attributeValuePos < 0) return false; // If not started, then initialize attributeString when parsing is at start. if (attributeValuePos == 0) { attributeString = value.Substring (1, value.Length - 2); // If it has an empty value, this method still returns true. if (attributeString.Length == 0) { attributeValuePos = -1; SetProperties (XmlNodeType.Text, "", false, "", false); return true; } } // It occurs when attribute dully consists of entity reference. if (attributeValuePos == attributeString.Length) return false; returnEntityReference = false; value = String.Empty; int refPosition; int loop = 0; do { refPosition = attributeString.IndexOf ('&', attributeValuePos); if (refPosition < 0) { // Reached to the end of value string. value += attributeString.Substring (attributeValuePos); attributeValuePos = -1; break; } else if (refPosition == attributeValuePos) { string parsed = ReadAttributeValueReference (); if (parsed != null) value += parsed; else { // Found that an entity reference starts from this point. // reset position to after '&'. attributeValuePos = refPosition; if (value.Length <= 0) { int endNamePos = attributeString.IndexOf (";", attributeValuePos); value = attributeString.Substring (attributeValuePos+1, endNamePos - attributeValuePos - 1); attributeValuePos += value.Length + 2; returnEntityReference = true; } break; } } else { value += attributeString.Substring (attributeValuePos, refPosition - attributeValuePos); attributeValuePos = refPosition; continue; } } while (++loop > 0); if (returnEntityReference) SetProperties (XmlNodeType.EntityReference, value, false, String.Empty, false); else SetProperties (XmlNodeType.Text, "", false, value, false); returnEntityReference = false; return true; } [MonoTODO] public int ReadBase64 (byte [] buffer, int offset, int length) { throw new NotImplementedException (); } [MonoTODO] public int ReadBinHex (byte [] buffer, int offset, int length) { throw new NotImplementedException (); } [MonoTODO] public int ReadChars (char [] buffer, int offset, int length) { throw new NotImplementedException (); } #if NET_1_0 public override string ReadInnerXml () { if (readState != ReadState.Interactive) return String.Empty; switch (NodeType) { case XmlNodeType.Attribute: return value.Substring (1, value.Length - 2); case XmlNodeType.Element: if (IsEmptyElement) return String.Empty; int startDepth = depth; if (innerXmlBuilder == null) innerXmlBuilder = new StringBuilder (); innerXmlBuilder.Length = 0; bool loop = true; do { Read (); if (NodeType ==XmlNodeType.None) throw new XmlException ("unexpected end of xml."); else if (NodeType == XmlNodeType.EndElement && depth == startDepth) { loop = false; Read (); } else innerXmlBuilder.Append (currentTag); } while (loop); string xml = innerXmlBuilder.ToString (); innerXmlBuilder.Length = 0; return xml; case XmlNodeType.None: // MS document is incorrect. Seems not to progress. return String.Empty; default: Read (); return String.Empty; } } public override string ReadOuterXml () { if (readState != ReadState.Interactive) return String.Empty; switch (NodeType) { case XmlNodeType.Attribute: // strictly incompatible with MS... (it holds spaces attribute between name, value and "=" char (very trivial). return String.Format ("{0}={1}{2}{1}", Name, QuoteChar, ReadInnerXml ()); case XmlNodeType.Element: bool isEmpty = IsEmptyElement; string startTag = currentTag.ToString (); string name = Name; if (NodeType == XmlNodeType.Element && !isEmpty) return String.Format ("{0}{1}", startTag, ReadInnerXml (), name); else return currentTag.ToString (); case XmlNodeType.None: // MS document is incorrect. Seems not to progress. return String.Empty; default: Read (); return String.Empty; } } #endif public override string ReadString () { return ReadStringInternal (); } public void ResetState () { Init (); } public override void ResolveEntity () { // XmlTextReader does not resolve entities. throw new InvalidOperationException ("XmlTextReader cannot resolve external entities."); } #endregion #region Internals // Parsed DTD Objects internal DTDObjectModel DTD { get { return parserContext.Dtd; } } internal bool MaybeTextDecl { set { if (value) this.maybeTextDecl = 2; } } #endregion #region Privates private XmlParserContext parserContext; private XmlParserInput currentInput; private Stack parserInputStack; private ReadState readState; private int depth; private int elementDepth; private bool depthUp; private bool popScope; private Stack elementStack; private bool allowMultipleRoot; private bool isStandalone; private XmlNodeType nodeType; private string name; private string prefix; private string localName; private string namespaceURI; private bool isEmptyElement; private string value; private StringBuilder valueBuilder; private bool valueBuilderAvailable = false; private bool insideAttribute; private bool isPropertySaved; private XmlNodeType saveNodeType; private string saveName; private string savePrefix; private string saveLocalName; private string saveNamespaceURI; private bool saveIsEmptyElement; private Hashtable attributes; private ArrayList orderedAttributes; private int orderedAttributesPos = -1; private bool returnEntityReference; private string entityReferenceName; private char [] nameBuffer; private int nameLength; private int nameCapacity; private const int initialNameCapacity = 256; private StringBuilder valueBuffer; // A buffer for ReadContent for ReadOuterXml private StringBuilder currentTag { get { return currentInput.CurrentMarkup; } } private string attributeString; private int attributeValuePos; // This should be only referenced(used) by ReadInnerXml(). Kind of flyweight pattern. private StringBuilder innerXmlBuilder; // Parameter entity placeholder private Hashtable parameterEntities; private int dtdIncludeSect; private XmlNodeType startNodeType; // State machine attribute. // XmlDeclaration: after the first node. // DocumentType: after doctypedecl // Element: inside document element // EndElement: after document element private XmlNodeType currentState; private int maybeTextDecl; // These values are never re-initialized. private XmlResolver resolver = new XmlUrlResolver (); private bool namespaces = true; private WhitespaceHandling whitespaceHandling = WhitespaceHandling.All; private bool normalization = false; private void Init () { readState = ReadState.Initial; currentState = XmlNodeType.None; maybeTextDecl = 0; allowMultipleRoot = false; depth = 0; depthUp = false; popScope = false; parserInputStack = new Stack (); elementStack = new Stack(); nodeType = XmlNodeType.None; name = String.Empty; prefix = String.Empty; localName = string.Empty; isEmptyElement = false; value = String.Empty; attributes = new Hashtable (); attributeString = String.Empty; orderedAttributes = new ArrayList (); orderedAttributesPos = -1; returnEntityReference = false; entityReferenceName = String.Empty; nameBuffer = new char [initialNameCapacity]; nameLength = 0; nameCapacity = initialNameCapacity; valueBuffer = new StringBuilder (8192); parameterEntities = new Hashtable (); } private void InitializeContext (string url, XmlParserContext context, TextReader fragment, XmlNodeType fragType) { startNodeType = fragType; parserContext = context; if (context == null) { XmlNameTable nt = new NameTable (); parserContext = new XmlParserContext (nt, new XmlNamespaceManager (nt), String.Empty, XmlSpace.None); } if (url != null && url != String.Empty) { string path = Path.GetFullPath ("./a"); Uri uri = new Uri (new Uri (path), url); parserContext.BaseURI = uri.ToString (); } Init (); switch (fragType) { case XmlNodeType.Attribute: value = String.Format ("{0}{1}{0}", "'", fragment.ReadToEnd ().Replace ("'", "'")); break; case XmlNodeType.Element: currentState = XmlNodeType.Element; allowMultipleRoot = true; break; case XmlNodeType.Document: break; default: throw new XmlException (String.Format ("NodeType {0} is not allowed to create XmlTextReader.", fragType)); } this.currentInput = new XmlParserInput (fragment, url); StreamReader sr = fragment as StreamReader; } // Use this method rather than setting the properties // directly so that all the necessary properties can // be changed in harmony with each other. Maybe the // fields should be in a seperate class to help enforce // this. private void SetProperties ( XmlNodeType nodeType, string name, bool isEmptyElement, string value, bool clearAttributes) { this.nodeType = nodeType; this.name = name; this.isEmptyElement = isEmptyElement; this.value = value; this.elementDepth = depth; this.valueBuilderAvailable = false; if (clearAttributes) ClearAttributes (); if (namespaces) { int indexOfColon = name.IndexOf (':'); if (indexOfColon == -1) { prefix = String.Empty; // prefix = nodeType == XmlNodeType.Attribute ? null : String.Empty; localName = name; } else { prefix = name.Substring (0, indexOfColon); localName = name.Substring (indexOfColon + 1); } } else { prefix = String.Empty; localName = name; } switch (nodeType) { case XmlNodeType.Attribute: if (prefix == string.Empty) namespaceURI = string.Empty; // if (prefix == string.Empty || prefix == null) namespaceURI = string.Empty; else namespaceURI = LookupNamespace (prefix); if (localName == "xmlns" && prefix == "") namespaceURI = "http://www.w3.org/2000/xmlns/"; break; case XmlNodeType.Element: case XmlNodeType.EndElement: namespaceURI = LookupNamespace (prefix); break; default: namespaceURI = ""; break; } } private void SetProperties ( XmlNodeType nodeType, string name, bool isEmptyElement, StringBuilder value, bool clearAttributes) { SetProperties (nodeType, name, isEmptyElement, (string)null, clearAttributes); this.valueBuilderAvailable = true; this.valueBuilder = value; } private void SaveProperties () { // If already saved, then return. if (isPropertySaved) return; saveNodeType = nodeType; saveName = name; savePrefix = prefix; saveLocalName = localName; saveNamespaceURI = namespaceURI; saveIsEmptyElement = isEmptyElement; // An element's value is always String.Empty. isPropertySaved = true; } private void RestoreProperties () { nodeType = saveNodeType; name = saveName; prefix = savePrefix; localName = saveLocalName; namespaceURI = saveNamespaceURI; isEmptyElement = saveIsEmptyElement; value = String.Empty; isPropertySaved = false; } private void AddAttribute (string name, string value) { if (attributes.ContainsKey (name)) throw new XmlException (this as IXmlLineInfo, String.Format ("Attribute {0} already exists.", name)); attributes.Add (name, value); orderedAttributes.Add (name); } private void ClearAttributes () { if (attributes.Count > 0) { attributes.Clear (); orderedAttributes.Clear (); } orderedAttributesPos = -1; } private int PeekChar () { return currentInput.PeekChar (); } private int ReadChar () { return currentInput.ReadChar (); } // This should really keep track of some state so // that it's not possible to have more than one document // element or text outside of the document element. private bool ReadContent () { currentTag.Length = 0; if (popScope) { parserContext.NamespaceManager.PopScope (); popScope = false; } if (returnEntityReference) SetEntityReferenceProperties (); else { switch (PeekChar ()) { case '<': ReadChar (); ReadTag (); break; case '\r': goto case ' '; case '\n': goto case ' '; case '\t': goto case ' '; case ' ': if (whitespaceHandling == WhitespaceHandling.All || whitespaceHandling == WhitespaceHandling.Significant) ReadWhitespace (); else { SkipWhitespace (); return ReadContent (); } break; case -1: if (depth > 0) throw new XmlException ("unexpected end of file. Current depth is " + depth); readState = ReadState.EndOfFile; SetProperties ( XmlNodeType.None, // nodeType String.Empty, // name false, // isEmptyElement String.Empty, // value true // clearAttributes ); break; default: ReadText (true); break; } } if (NodeType == XmlNodeType.XmlDeclaration && maybeTextDecl == 1) return ReadContent (); return this.ReadState != ReadState.EndOfFile; } private void SetEntityReferenceProperties () { /* if (resolver != null) { if (DTD == null) throw new XmlException (this as IXmlLineInfo, "Entity reference is not allowed without document type declaration."); else if((!DTD.InternalSubsetHasPEReference || isStandalone) && DTD.EntityDecls [entityReferenceName] == null) throw new XmlException (this as IXmlLineInfo, "Required entity declaration for '" + entityReferenceName + "' was not found."); string dummy = DTD.EntityDecls [entityReferenceName].EntityValue; } */ SetProperties ( XmlNodeType.EntityReference, // nodeType entityReferenceName, // name false, // isEmptyElement String.Empty, // value true // clearAttributes ); returnEntityReference = false; entityReferenceName = String.Empty; } // The leading '<' has already been consumed. private void ReadTag () { switch (PeekChar ()) { case '/': ReadChar (); ReadEndTag (); break; case '?': ReadChar (); ReadProcessingInstruction (); break; case '!': ReadChar (); ReadDeclaration (); break; default: ReadStartTag (); break; } } // The leading '<' has already been consumed. private void ReadStartTag () { if (currentState == XmlNodeType.EndElement) throw new XmlException (this as IXmlLineInfo, "Element cannot appear in this state."); currentState = XmlNodeType.Element; parserContext.NamespaceManager.PushScope (); string name = ReadName (); if (currentState == XmlNodeType.EndElement) throw new XmlException (this as IXmlLineInfo,"document has terminated, cannot open new element"); bool isEmptyElement = false; ClearAttributes (); SkipWhitespace (); if (XmlChar.IsFirstNameChar (PeekChar ())) ReadAttributes (false); string baseUri = GetAttribute ("xml:base"); if (baseUri != null) parserContext.BaseURI = baseUri; string xmlLang = GetAttribute ("xml:lang"); if (xmlLang != null) parserContext.XmlLang = xmlLang; string xmlSpaceAttr = GetAttribute ("xml:space"); if (xmlSpaceAttr != null) { if (xmlSpaceAttr == "preserve") parserContext.XmlSpace = XmlSpace.Preserve; else if (xmlSpaceAttr == "default") parserContext.XmlSpace = XmlSpace.Default; else throw new XmlException (this as IXmlLineInfo,String.Format ("Invalid xml:space value: {0}", xmlSpaceAttr)); } if (PeekChar () == '/') { ReadChar (); isEmptyElement = true; popScope = true; } else { depthUp = true; elementStack.Push (name); parserContext.PushScope (); } Expect ('>'); SetProperties ( XmlNodeType.Element, // nodeType name, // name isEmptyElement, // isEmptyElement String.Empty, // value false // clearAttributes ); } // The reader is positioned on the first character // of the element's name. private void ReadEndTag () { if (currentState != XmlNodeType.Element) throw new XmlException (this as IXmlLineInfo, "End tag cannot appear in this state."); string name = ReadName (); if (elementStack.Count == 0) throw new XmlException (this as IXmlLineInfo,"closing element without matching opening element"); string expected = (string)elementStack.Pop(); if (expected != name) throw new XmlException (this as IXmlLineInfo,String.Format ("unmatched closing element: expected {0} but found {1}", expected, name)); parserContext.PopScope (); SkipWhitespace (); Expect ('>'); --depth; SetProperties ( XmlNodeType.EndElement, // nodeType name, // name false, // isEmptyElement String.Empty, // value true // clearAttributes ); popScope = true; } private void AppendNameChar (int ch) { CheckNameCapacity (); nameBuffer [nameLength++] = (char)ch; } private void CheckNameCapacity () { if (nameLength == nameCapacity) { nameCapacity = nameCapacity * 2; char [] oldNameBuffer = nameBuffer; nameBuffer = new char [nameCapacity]; Array.Copy (oldNameBuffer, nameBuffer, nameLength); } } private string CreateNameString () { return parserContext.NameTable.Add (nameBuffer, 0, nameLength); } private void AppendValueChar (int ch) { valueBuffer.Append ((char)ch); } private string CreateValueString () { return valueBuffer.ToString (); } private void ClearValueBuffer () { valueBuffer.Length = 0; } // The reader is positioned on the first character // of the text. private void ReadText (bool notWhitespace) { if (currentState != XmlNodeType.Element) throw new XmlException (this as IXmlLineInfo, "Text node cannot appear in this state."); if (notWhitespace) ClearValueBuffer (); int ch = PeekChar (); int previousCloseBracketLine = 0; int previousCloseBracketColumn = 0; while (ch != '<' && ch != -1) { if (ch == '&') { ReadChar (); if (ReadReference (false)) break; } else { if (XmlConstructs.IsInvalid (ch)) throw new XmlException (this as IXmlLineInfo, "Not allowed character was found."); AppendValueChar (ReadChar ()); if (ch == ']') { if (previousCloseBracketColumn == LinePosition - 1 && previousCloseBracketLine == LineNumber) if (PeekChar () == '>') throw new XmlException (this as IXmlLineInfo, "Inside text content, character sequence ']]>' is not allowed."); previousCloseBracketColumn = LinePosition; previousCloseBracketLine = LineNumber; } } ch = PeekChar (); notWhitespace = true; } if (returnEntityReference && valueBuffer.Length == 0) { SetEntityReferenceProperties (); } else { XmlNodeType nodeType = notWhitespace ? XmlNodeType.Text : XmlNodeType.Whitespace; SetProperties ( nodeType, // nodeType String.Empty, // name false, // isEmptyElement valueBuffer, // value true // clearAttributes ); } } // The leading '&' has already been consumed. // Returns true if the entity reference isn't a simple // character reference or one of the predefined entities. // This allows the ReadText method to break so that the // next call to Read will return the EntityReference node. private bool ReadReference (bool ignoreEntityReferences) { if (PeekChar () == '#') { ReadChar (); ReadCharacterReference (); } else ReadEntityReference (ignoreEntityReferences); return returnEntityReference; } private void ReadCharacterReference () { int value = 0; if (PeekChar () == 'x') { ReadChar (); while (PeekChar () != ';' && PeekChar () != -1) { int ch = ReadChar (); if (ch >= '0' && ch <= '9') value = (value << 4) + ch - '0'; else if (ch >= 'A' && ch <= 'F') value = (value << 4) + ch - 'A' + 10; else if (ch >= 'a' && ch <= 'f') value = (value << 4) + ch - 'a' + 10; else throw new XmlException (this as IXmlLineInfo, String.Format ( "invalid hexadecimal digit: {0} (#x{1:X})", (char)ch, ch)); } } else { while (PeekChar () != ';' && PeekChar () != -1) { int ch = ReadChar (); if (ch >= '0' && ch <= '9') value = value * 10 + ch - '0'; else throw new XmlException (this as IXmlLineInfo, String.Format ( "invalid decimal digit: {0} (#x{1:X})", (char)ch, ch)); } } ReadChar (); // ';' // FIXME: how to handle such chars larger than 0xffff? if (value < 0xffff && !XmlConstructs.IsValid (value)) throw new XmlException (this as IXmlLineInfo, "Referenced character was not allowed in XML."); AppendValueChar (value); } private void ReadEntityReference (bool ignoreEntityReferences) { nameLength = 0; int ch = PeekChar (); while (ch != ';' && ch != -1) { AppendNameChar (ReadChar ()); ch = PeekChar (); } Expect (';'); string name = CreateNameString (); if (!XmlChar.IsName (name)) throw new XmlException (this as IXmlLineInfo, "Invalid entity reference name was found."); char predefined = XmlChar.GetPredefinedEntity (name); if (predefined != 0) AppendValueChar (predefined); else { if (ignoreEntityReferences) { AppendValueChar ('&'); foreach (char ch2 in name) { AppendValueChar (ch2); } AppendValueChar (';'); } else { returnEntityReference = true; entityReferenceName = name; } } } // The reader is positioned on the first character of // the attribute name. private void ReadAttributes (bool allowPIEnd) { int peekChar = -1; bool requireWhitespace = false; do { if (!SkipWhitespace () && requireWhitespace) throw new XmlException ("Unexpected token. Name is required here."); string name = ReadName (); SkipWhitespace (); Expect ('='); SkipWhitespace (); string value = ReadAttribute (false); if (name == "xmlns") parserContext.NamespaceManager.AddNamespace (String.Empty, UnescapeAttributeValue (value)); else if (name.StartsWith ("xmlns:")) parserContext.NamespaceManager.AddNamespace (name.Substring (6), UnescapeAttributeValue (value)); AddAttribute (name, value); if (!SkipWhitespace ()) requireWhitespace = true; peekChar = PeekChar (); if (peekChar == '?' && allowPIEnd) break; } while (peekChar != '/' && peekChar != '>' && peekChar != -1); } // The reader is positioned on the quote character. // *Keeps quote char* to value to get_QuoteChar() correctly. private string ReadAttribute (bool isDefaultValue) { ClearValueBuffer (); int quoteChar = ReadChar (); if (quoteChar != '\'' && quoteChar != '\"') throw new XmlException (this as IXmlLineInfo,"an attribute value was not quoted"); AppendValueChar (quoteChar); while (PeekChar () != quoteChar) { int ch = ReadChar (); switch (ch) { case '<': throw new XmlException (this as IXmlLineInfo,"attribute values cannot contain '<'"); case -1: throw new XmlException (this as IXmlLineInfo,"unexpected end of file in an attribute value"); /* case '&': if (isDefaultValue) { AppendValueChar (ch); break; } AppendValueChar (ch); if (PeekChar () == '#') break; // Check XML 1.0 section 3.1 WFC. string entName = ReadName (); Expect (';'); if (XmlChar.GetPredefinedEntity (entName) == 0) { DTDEntityDeclaration entDecl = DTD == null ? null : DTD.EntityDecls [entName]; if (entDecl == null || entDecl.SystemId != null) throw new XmlException (this as IXmlLineInfo, "Reference to external entities is not allowed in attribute value."); } valueBuffer.Append (entName); AppendValueChar (';'); break; */ default: AppendValueChar (ch); break; } } ReadChar (); // quoteChar AppendValueChar (quoteChar); return CreateValueString (); } // The reader is positioned on the first character // of the target. // // It may be xml declaration or processing instruction. private void ReadProcessingInstruction () { string target = ReadName (); if (target == "xml") { ReadXmlDeclaration (); return; } else if (target.ToLower () == "xml") throw new XmlException (this as IXmlLineInfo, "Not allowed processing instruction name which starts with 'X', 'M', 'L' was found."); if (currentState == XmlNodeType.None) currentState = XmlNodeType.XmlDeclaration; if (!SkipWhitespace ()) if (PeekChar () != '?') throw new XmlException (this as IXmlLineInfo, "Invalid processing instruction name was found."); ClearValueBuffer (); while (PeekChar () != -1) { int ch = ReadChar (); if (ch == '?' && PeekChar () == '>') { ReadChar (); break; } AppendValueChar ((char)ch); } SetProperties ( XmlNodeType.ProcessingInstruction, // nodeType target, // name false, // isEmptyElement valueBuffer, // value true // clearAttributes ); } // The reader is positioned after " 1 && (orderedAttributes [1] as string != "encoding" && orderedAttributes [1] as string != "standalone")) message = "Invalid Xml Declaration markup was found."; else if (orderedAttributes.Count > 2 && orderedAttributes [2] as string != "standalone") message = "Invalid Xml Declaration markup was found."; string sa = attributes ["standalone"] as string; if (sa != null) sa = sa.Substring (1, sa.Length - 2); if (sa != null && sa != "yes" && sa != "no") message = "Only 'yes' or 'no' is allowed for standalone."; this.isStandalone = (sa == "yes"); } else { int currentCheck = 0; if (orderedAttributes [0] as string == "version") { if (version != "1.0") message = "Version 1.0 declaration is required in Text Declaration."; currentCheck = 1; } if (orderedAttributes.Count <= currentCheck || orderedAttributes [currentCheck] as string != "encoding") message = "Invalid Text Declaration markup was found. encoding specification is required."; } if (message != null) throw new XmlException (this as IXmlLineInfo, message); Expect ("?>"); if (maybeTextDecl != 0) if (this ["standalone"] != null) throw new XmlException (this as IXmlLineInfo, "Invalid text declaration."); if (maybeTextDecl == 2) maybeTextDecl = 1; SetProperties ( XmlNodeType.XmlDeclaration, // nodeType "xml", // name false, // isEmptyElement currentInput.CurrentMarkup.ToString (6, currentInput.CurrentMarkup.Length - 6), // value false // clearAttributes ); } // The reader is positioned on the first character after // the leading '') throw new XmlException (this as IXmlLineInfo,"comments cannot contain '--'"); ReadChar (); break; } if (XmlConstructs.IsInvalid (ch)) throw new XmlException (this as IXmlLineInfo, "Not allowed character was found."); AppendValueChar ((char)ch); } SetProperties ( XmlNodeType.Comment, // nodeType String.Empty, // name false, // isEmptyElement valueBuffer, // value true // clearAttributes ); } // The reader is positioned on the first character after // the leading '') { ReadChar (); // '>' break; } else { skip = true; // AppendValueChar (']'); // AppendValueChar (']'); // ch = ReadChar (); } } AppendValueChar ((char)ch); } SetProperties ( XmlNodeType.CDATA, // nodeType String.Empty, // name false, // isEmptyElement valueBuffer, // value true // clearAttributes ); } // The reader is positioned on the first character after // the leading ''); GenerateDTDObjectModel (doctypeName, publicId, systemId, parserContext.InternalSubset, intSubsetStartLine, intSubsetStartColumn); // set properties for node SetProperties ( XmlNodeType.DocumentType, // nodeType doctypeName, // name false, // isEmptyElement parserContext.InternalSubset, // value true // clearAttributes ); } internal DTDObjectModel GenerateDTDObjectModel (string name, string publicId, string systemId, string internalSubset) { return GenerateDTDObjectModel (name, publicId, systemId, internalSubset, 0, 0); } internal DTDObjectModel GenerateDTDObjectModel (string name, string publicId, string systemId, string internalSubset, int intSubsetStartLine, int intSubsetStartColumn) { // now compile DTD parserContext.Dtd = new DTDObjectModel (); // merges both internal and external subsets in the meantime, DTD.BaseURI = BaseURI; DTD.Name = name; DTD.PublicId = publicId; DTD.SystemId = systemId; DTD.InternalSubset = internalSubset; DTD.XmlResolver = resolver; int originalParserDepth = parserInputStack.Count; if (internalSubset != null && internalSubset.Length > 0) { XmlParserInput original = currentInput; currentInput = new XmlParserInput (new StringReader (internalSubset), BaseURI, intSubsetStartLine, intSubsetStartColumn); do { CompileDTDSubset (); if (PeekChar () == -1 && parserInputStack.Count > 0) PopParserInput (); } while (nodeType != XmlNodeType.None || parserInputStack.Count > originalParserDepth); if (dtdIncludeSect != 0) throw new XmlException (this as IXmlLineInfo,"INCLUDE section is not ended correctly."); currentInput = original; } if (systemId != null && systemId != String.Empty && resolver != null) { PushParserInput (systemId); do { this.CompileDTDSubset (); if (PeekChar () == -1 && parserInputStack.Count > 1) PopParserInput (); } while (nodeType != XmlNodeType.None || parserInputStack.Count > originalParserDepth + 1); PopParserInput (); } return DTD; } private void PushParserInput (string url) { Uri baseUri = null; try { baseUri = new Uri (BaseURI); } catch (UriFormatException) { } Uri absUri = resolver.ResolveUri (baseUri, url); string absPath = absUri.ToString (); foreach (XmlParserInput i in parserInputStack.ToArray ()) { if (i.BaseURI == absPath) throw new XmlException (this as IXmlLineInfo, "Nested inclusion is not allowed: " + url); } parserInputStack.Push (currentInput); currentInput = new XmlParserInput (new XmlStreamReader (url, false, resolver, BaseURI), absPath); parserContext.PushScope (); parserContext.BaseURI = absPath; maybeTextDecl = 2; } private void PopParserInput () { currentInput = parserInputStack.Pop () as XmlParserInput; parserContext.PopScope (); } private enum DtdInputState { Free = 1, ElementDecl, AttlistDecl, EntityDecl, NotationDecl, PI, Comment, InsideSingleQuoted, InsideDoubleQuoted, } private class DtdInputStateStack { Stack intern = new Stack (); public DtdInputStateStack () { Push (DtdInputState.Free); } public DtdInputState Peek () { return (DtdInputState) intern.Peek (); } public DtdInputState Pop () { return (DtdInputState) intern.Pop (); } public void Push (DtdInputState val) { intern.Push (val); } } DtdInputStateStack stateStack = new DtdInputStateStack (); DtdInputState State { get { return stateStack.Peek (); } } // Simply read but not generate any result. private void ReadInternalSubset () { bool continueParse = true; while (continueParse) { switch (ReadChar ()) { case ']': switch (State) { case DtdInputState.Free: continueParse = false; break; case DtdInputState.InsideDoubleQuoted: continue; case DtdInputState.InsideSingleQuoted: continue; default: throw new XmlException (this as IXmlLineInfo,"unexpected end of file at DTD."); } break; case -1: throw new XmlException (this as IXmlLineInfo,"unexpected end of file at DTD."); case '<': if (State == DtdInputState.InsideDoubleQuoted || State == DtdInputState.InsideSingleQuoted) continue; // well-formed switch (ReadChar ()) { case '?': stateStack.Push (DtdInputState.PI); break; case '!': switch (ReadChar ()) { case 'E': switch (ReadChar ()) { case 'L': Expect ("EMENT"); stateStack.Push (DtdInputState.ElementDecl); break; case 'N': Expect ("TITY"); stateStack.Push (DtdInputState.EntityDecl); break; default: throw new XmlException (this as IXmlLineInfo,"unexpected token ''."); } break; case '\'': if (State == DtdInputState.InsideSingleQuoted) stateStack.Pop (); else if (State != DtdInputState.InsideDoubleQuoted && State != DtdInputState.Comment) stateStack.Push (DtdInputState.InsideSingleQuoted); break; case '"': if (State == DtdInputState.InsideDoubleQuoted) stateStack.Pop (); else if (State != DtdInputState.InsideSingleQuoted && State != DtdInputState.Comment) stateStack.Push (DtdInputState.InsideDoubleQuoted); break; case '>': switch (State) { case DtdInputState.ElementDecl: goto case DtdInputState.NotationDecl; case DtdInputState.AttlistDecl: goto case DtdInputState.NotationDecl; case DtdInputState.EntityDecl: goto case DtdInputState.NotationDecl; case DtdInputState.NotationDecl: stateStack.Pop (); break; case DtdInputState.InsideDoubleQuoted: continue; case DtdInputState.InsideSingleQuoted: continue; // well-formed case DtdInputState.Comment: continue; default: throw new XmlException (this as IXmlLineInfo,"unexpected token '>'"); } break; case '?': if (State == DtdInputState.PI) { if (ReadChar () == '>') stateStack.Pop (); } break; case '-': if (State == DtdInputState.Comment) { if (PeekChar () == '-') { ReadChar (); Expect ('>'); stateStack.Pop (); } } break; case '%': if (State != DtdInputState.Free && State != DtdInputState.EntityDecl && State != DtdInputState.Comment && State != DtdInputState.InsideDoubleQuoted && State != DtdInputState.InsideSingleQuoted) throw new XmlException (this as IXmlLineInfo,"Parameter Entity Reference cannot appear as a part of markupdecl (see XML spec 2.8)."); break; } } } // Read any one of following: // elementdecl, AttlistDecl, EntityDecl, NotationDecl, // PI, Comment, Parameter Entity, or doctype termination char(']') // // returns a node of some nodeType or null, setting nodeType. // (if None then ']' was found.) private void CompileDTDSubset() { SkipWhitespace (); switch(PeekChar ()) { case -1: nodeType = XmlNodeType.None; break; case '%': // It affects on entity references' well-formedness if (this.parserInputStack.Count == 0) DTD.InternalSubsetHasPEReference = true; ReadChar (); string peName = ReadName (); Expect (';'); currentInput.InsertParameterEntityBuffer (GetPEValue (peName)); int currentLine = currentInput.LineNumber; int currentColumn = currentInput.LinePosition; while (currentInput.HasPEBuffer) CompileDTDSubset (); if (currentInput.LineNumber != currentLine || currentInput.LinePosition != currentColumn) throw new XmlException (this as IXmlLineInfo, "Incorrectly nested parameter entity."); break; case '<': ReadChar (); switch(ReadChar ()) { case '?': // Only read, no store. ReadProcessingInstruction (); break; case '!': CompileDeclaration (); break; default: throw new XmlException (this as IXmlLineInfo,"Syntax Error after '<' character."); } break; case ']': if (dtdIncludeSect == 0) throw new XmlException (this as IXmlLineInfo, "Unbalanced end of INCLUDE/IGNORE section."); // End of inclusion Expect ("]]>"); dtdIncludeSect--; SkipWhitespace (); break; default: throw new XmlException (this as IXmlLineInfo,String.Format ("Syntax Error inside doctypedecl markup : {0}({1})", PeekChar (), (char) PeekChar ())); } } private void CompileDeclaration () { nodeType = XmlNodeType.DocumentType; // Hack!! switch(ReadChar ()) { case '-': Expect ('-'); // Only read, no store. ReadComment (); break; case 'E': switch(ReadChar ()) { case 'N': Expect ("TITY"); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required after ' 0) { switch (skip ? PeekChar () : ReadChar ()) { case -1: throw new XmlException (this as IXmlLineInfo,"Unexpected IGNORE section end."); case '<': if (ReadChar () == '!' && ReadChar () == '[') dtdIgnoreSect++; break; case ']': if (ReadChar () == ']') { if (ReadChar () == '>') dtdIgnoreSect--; else skip = true; } break; } skip = false; } } // The reader is positioned on the head of the name. private DTDElementDeclaration ReadElementDecl () { DTDElementDeclaration decl = new DTDElementDeclaration (DTD); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between ''); return decl; } // read 'children'(BNF) of contentspec private void ReadContentSpec (DTDElementDeclaration decl) { TryExpandPERef (); SkipWhitespace (); switch(PeekChar ()) { case 'E': decl.IsEmpty = true; Expect ("EMPTY"); break; case 'A': decl.IsAny = true; Expect ("ANY"); break; case '(': DTDContentModel model = decl.ContentModel; ReadChar (); SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); if(PeekChar () == '#') { // Mixed Contents. "#PCDATA" must appear first. decl.IsMixedContent = true; model.Occurence = DTDOccurence.ZeroOrMore; model.OrderType = DTDContentOrderType.Or; Expect ("#PCDATA"); SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); while(PeekChar () != ')') { Expect('|'); SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); DTDContentModel elem = new DTDContentModel (DTD, decl.Name); elem.ElementName = ReadName (); model.ChildModels.Add (elem); SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); } Expect (')'); if (model.ChildModels.Count > 0) Expect ('*'); else if (PeekChar () == '*') Expect ('*'); } else { // Non-Mixed Contents model.ChildModels.Add (ReadCP (decl)); SkipWhitespace (); do { // copied from ReadCP() ...;-) TryExpandPERef (); SkipWhitespace (); if(PeekChar ()=='|') { // CPType=Or if (model.OrderType == DTDContentOrderType.Seq) throw new XmlException (this as IXmlLineInfo, "Inconsistent choice markup in sequence cp."); model.OrderType = DTDContentOrderType.Or; ReadChar (); SkipWhitespace (); model.ChildModels.Add (ReadCP (decl)); SkipWhitespace (); } else if(PeekChar () == ',') { // CPType=Seq if (model.OrderType == DTDContentOrderType.Or) throw new XmlException (this as IXmlLineInfo, "Inconsistent sequence markup in choice cp."); model.OrderType = DTDContentOrderType.Seq; ReadChar (); SkipWhitespace (); model.ChildModels.Add (ReadCP (decl)); SkipWhitespace (); } else break; } while(true); Expect (')'); switch(PeekChar ()) { case '?': model.Occurence = DTDOccurence.Optional; ReadChar (); break; case '*': model.Occurence = DTDOccurence.ZeroOrMore; ReadChar (); break; case '+': model.Occurence = DTDOccurence.OneOrMore; ReadChar (); break; } SkipWhitespace (); } SkipWhitespace (); break; } } // Read 'cp' (BNF) of contentdecl (BNF) private DTDContentModel ReadCP (DTDElementDeclaration elem) { DTDContentModel model = null; TryExpandPERef (); SkipWhitespace (); if(PeekChar () == '(') { model = new DTDContentModel (DTD, elem.Name); ReadChar (); SkipWhitespace (); model.ChildModels.Add (ReadCP (elem)); SkipWhitespace (); do { TryExpandPERef (); SkipWhitespace (); if(PeekChar ()=='|') { // CPType=Or if (model.OrderType == DTDContentOrderType.Seq) throw new XmlException (this as IXmlLineInfo, "Inconsistent choice markup in sequence cp."); model.OrderType = DTDContentOrderType.Or; ReadChar (); SkipWhitespace (); model.ChildModels.Add (ReadCP (elem)); SkipWhitespace (); } else if(PeekChar () == ',') { // CPType=Seq if (model.OrderType == DTDContentOrderType.Or) throw new XmlException (this as IXmlLineInfo, "Inconsistent sequence markup in choice cp."); model.OrderType = DTDContentOrderType.Seq; ReadChar (); SkipWhitespace (); model.ChildModels.Add (ReadCP (elem)); SkipWhitespace (); } else break; } while(true); SkipWhitespace (); Expect (')'); } else { TryExpandPERef (); model = new DTDContentModel (DTD, elem.Name); SkipWhitespace (); model.ElementName = ReadName (); } switch(PeekChar ()) { case '?': model.Occurence = DTDOccurence.Optional; ReadChar (); break; case '*': model.Occurence = DTDOccurence.ZeroOrMore; ReadChar (); break; case '+': model.Occurence = DTDOccurence.OneOrMore; ReadChar (); break; } return model; } // The reader is positioned on the first name char. private void ReadParameterEntityDecl () { DTDParameterEntityDeclaration decl = new DTDParameterEntityDeclaration(); decl.BaseURI = BaseURI; decl.Name = ReadName (); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required after name in DTD parameter entity declaration."); if (PeekChar () == 'S' || PeekChar () == 'P') { // throw new NotImplementedException ("External parameter entity reference is not implemented yet."); // read publicId/systemId ReadExternalID (); decl.PublicId = attributes ["PUBLIC"] as string; decl.SystemId = attributes ["SYSTEM"] as string; SkipWhitespace (); decl.Resolve (resolver); } else { TryExpandPERef (); int quoteChar = ReadChar (); int start = currentTag.Length; ClearValueBuffer (); bool loop = true; while (loop) { int c = PeekChar (); switch (c) { case -1: throw new XmlException ("unexpected end of stream in entity value definition."); case '"': ReadChar (); if (quoteChar == '"') loop = false; else AppendValueChar ('"'); break; case '\'': ReadChar (); if (quoteChar == '\'') loop = false; else AppendValueChar ('\''); break; case '&': ReadChar (); if (PeekChar () == '#') { ReadChar (); ReadCharacterReference (); } else AppendValueChar ('&'); break; case '%': ReadChar (); string peName = ReadName (); Expect (';'); valueBuffer.Append (GetPEValue (peName)); break; default: AppendValueChar (ReadChar ()); break; } } decl.LiteralValue = CreateValueString (); // currentTag.ToString (start, currentTag.Length - start - 1); ClearValueBuffer (); } SkipWhitespace (); Expect ('>'); if (parameterEntities [decl.Name] == null) { parameterEntities.Add (decl.Name, decl); } } private string GetPEValue (string peName) { DTDParameterEntityDeclaration peDecl = this.parameterEntities [peName] as DTDParameterEntityDeclaration; if (peDecl != null) return peDecl.Value; // See XML 1.0 section 4.1 for both WFC and VC. if ((DTD.SystemId == null && !DTD.InternalSubsetHasPEReference) || this.isStandalone) throw new XmlException (this as IXmlLineInfo, "Parameter entity " + peName + " not found."); DTD.AddError (new XmlSchemaException ( "Parameter entity " + peName + " not found.", null)); return ""; } private void TryExpandPERef () { if (PeekChar () == '%') { // ReadChar (); // if (!XmlChar.IsNameChar (PeekChar ())) // return; // ExpandPERef (); ImportAsPERef (); } } // reader is positioned on '%' private void ImportAsPERef () { ReadChar (); string peName = ReadName (); Expect (';'); DTDParameterEntityDeclaration peDecl = this.parameterEntities [peName] as DTDParameterEntityDeclaration; if (peDecl == null) { DTD.AddError (new XmlSchemaException ("Parameter entity " + peName + " not found.", null)); return; // do nothing } currentInput.InsertParameterEntityBuffer (" " + peDecl.Value + " "); } // The reader is positioned on the head of the name. private DTDEntityDeclaration ReadEntityDecl () { DTDEntityDeclaration decl = new DTDEntityDeclaration (DTD); decl.IsInternalSubset = (parserInputStack.Count == 0); TryExpandPERef (); SkipWhitespace (); decl.Name = ReadName (); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between name and content in DTD entity declaration."); TryExpandPERef (); SkipWhitespace (); if (PeekChar () == 'S' || PeekChar () == 'P') { // external entity ReadExternalID (); decl.PublicId = attributes ["PUBLIC"] as string; decl.SystemId = attributes ["SYSTEM"] as string; if (SkipWhitespace ()) { if (PeekChar () == 'N') { // NDataDecl Expect ("NDATA"); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required after NDATA."); decl.NotationName = ReadName (); // ndata_name } } } else { // literal entity ReadEntityValueDecl (decl); } SkipWhitespace (); // This expanding is only allowed as a non-validating parser. TryExpandPERef (); SkipWhitespace (); Expect ('>'); return decl; } private void ReadEntityValueDecl (DTDEntityDeclaration decl) { SkipWhitespace (); // quotation char will be finally removed on unescaping int quoteChar = ReadChar (); int start = currentTag.Length; if (quoteChar != '\'' && quoteChar != '"') throw new XmlException ("quotation char was expected."); ClearValueBuffer (); while (PeekChar () != quoteChar) { switch (PeekChar ()) { case '%': ReadChar (); string name = ReadName (); Expect (';'); if (decl.IsInternalSubset) throw new XmlException (this as IXmlLineInfo, "Parameter entity is not allowed in internal subset entity '" + name + "'"); valueBuffer.Append (GetPEValue (name)); break; case -1: throw new XmlException ("unexpected end of stream."); default: AppendValueChar (ReadChar ()); break; } } // string value = Dereference (currentTag.ToString (start, currentTag.Length - start), false); string value = Dereference (CreateValueString (), false); ClearValueBuffer (); Expect (quoteChar); decl.LiteralEntityValue = value; } private DTDAttListDeclaration ReadAttListDecl () { SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); string name = ReadName (); // target element name DTDAttListDeclaration decl = DTD.AttListDecls [name] as DTDAttListDeclaration; if (decl == null) decl = new DTDAttListDeclaration (DTD); decl.Name = name; if (!SkipWhitespace ()) if (PeekChar () != '>') throw new XmlException (this as IXmlLineInfo, "Whitespace is required between name and content in non-empty DTD attlist declaration."); TryExpandPERef (); SkipWhitespace (); while (XmlChar.IsNameChar ((char) PeekChar ())) { DTDAttributeDefinition def = ReadAttributeDefinition (); if (decl [def.Name] == null) decl.Add (def); SkipWhitespace (); TryExpandPERef (); SkipWhitespace (); } SkipWhitespace (); // This expanding is only allowed as a non-validating parser. TryExpandPERef (); SkipWhitespace (); Expect ('>'); return decl; } private DTDAttributeDefinition ReadAttributeDefinition () { DTDAttributeDefinition def = new DTDAttributeDefinition (); // attr_name TryExpandPERef (); SkipWhitespace (); def.Name = ReadName (); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between name and content in DTD attribute definition."); // attr_value TryExpandPERef (); SkipWhitespace (); switch(PeekChar ()) { case 'C': // CDATA Expect ("CDATA"); def.Datatype = XmlSchemaDatatype.FromName ("normalizedString"); break; case 'I': // ID, IDREF, IDREFS Expect ("ID"); if(PeekChar () == 'R') { Expect ("REF"); if(PeekChar () == 'S') { // IDREFS ReadChar (); def.Datatype = XmlSchemaDatatype.FromName ("IDREFS"); } else // IDREF def.Datatype = XmlSchemaDatatype.FromName ("IDREF"); } else // ID def.Datatype = XmlSchemaDatatype.FromName ("ID"); break; case 'E': // ENTITY, ENTITIES Expect ("ENTIT"); switch(ReadChar ()) { case 'Y': // ENTITY def.Datatype = XmlSchemaDatatype.FromName ("ENTITY"); break; case 'I': // ENTITIES Expect ("ES"); def.Datatype = XmlSchemaDatatype.FromName ("ENTITIES"); break; } break; case 'N': // NMTOKEN, NMTOKENS, NOTATION ReadChar (); switch(PeekChar ()) { case 'M': Expect ("MTOKEN"); if(PeekChar ()=='S') { // NMTOKENS ReadChar (); def.Datatype = XmlSchemaDatatype.FromName ("NMTOKENS"); } else // NMTOKEN def.Datatype = XmlSchemaDatatype.FromName ("NMTOKEN"); break; case 'O': Expect ("OTATION"); def.Datatype = XmlSchemaDatatype.FromName ("NOTATION"); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between name and content in DTD attribute definition."); Expect ('('); SkipWhitespace (); def.EnumeratedNotations.Add (ReadName ()); // notation name SkipWhitespace (); while(PeekChar () == '|') { ReadChar (); SkipWhitespace (); def.EnumeratedNotations.Add (ReadName ()); // notation name SkipWhitespace (); } Expect (')'); break; default: throw new XmlException ("attribute declaration syntax error."); } break; default: // Enumerated Values def.Datatype = XmlSchemaDatatype.FromName ("NMTOKEN"); TryExpandPERef (); SkipWhitespace (); Expect ('('); SkipWhitespace (); def.EnumeratedAttributeDeclaration.Add ( def.Datatype.Normalize (ReadNmToken ())); // enum value SkipWhitespace (); while(PeekChar () == '|') { ReadChar (); SkipWhitespace (); def.EnumeratedAttributeDeclaration.Add ( def.Datatype.Normalize (ReadNmToken ())); // enum value SkipWhitespace (); } Expect (')'); break; } TryExpandPERef (); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between type and occurence in DTD attribute definition."); // def_value if(PeekChar () == '#') { ReadChar (); switch(PeekChar ()) { case 'R': Expect ("REQUIRED"); def.OccurenceType = DTDAttributeOccurenceType.Required; break; case 'I': Expect ("IMPLIED"); def.OccurenceType = DTDAttributeOccurenceType.Optional; break; case 'F': Expect ("FIXED"); def.OccurenceType = DTDAttributeOccurenceType.Fixed; if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between FIXED and actual value in DTD attribute definition."); def.UnresolvedDefaultValue = ReadAttribute (true); break; } } else { // one of the enumerated value TryExpandPERef (); SkipWhitespace (); def.UnresolvedDefaultValue = ReadAttribute (true); } return def; } private DTDNotationDeclaration ReadNotationDecl() { DTDNotationDeclaration decl = new DTDNotationDeclaration (); TryExpandPERef (); SkipWhitespace (); decl.Name = ReadName (); // notation name if (namespaces) { // copy from SetProperties ;-) int indexOfColon = decl.Name.IndexOf (':'); if (indexOfColon == -1) { decl.Prefix = String.Empty; decl.LocalName = decl.Name; } else { decl.Prefix = decl.Name.Substring (0, indexOfColon); decl.LocalName = decl.Name.Substring (indexOfColon + 1); } } else { decl.Prefix = String.Empty; decl.LocalName = decl.Name; } SkipWhitespace (); if(PeekChar () == 'P') { decl.PublicId = ReadPubidLiteral (); bool wsSkipped = SkipWhitespace (); if (PeekChar () == '\'' || PeekChar () == '"') { if (!wsSkipped) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between public id and system id."); decl.SystemId = ReadSystemLiteral (false); SkipWhitespace (); } } else if(PeekChar () == 'S') { decl.SystemId = ReadSystemLiteral (true); SkipWhitespace (); } if(decl.PublicId == null && decl.SystemId == null) throw new XmlException ("public or system declaration required for \"NOTATION\" declaration."); // This expanding is only allowed as a non-validating parser. TryExpandPERef (); SkipWhitespace (); Expect ('>'); return decl; } private void ReadExternalID() { switch(PeekChar ()) { case 'S': attributes ["PUBLIC"] = null; attributes ["SYSTEM"] = ReadSystemLiteral (true); break; case 'P': attributes ["PUBLIC"] = ReadPubidLiteral (); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required between PUBLIC id and SYSTEM id."); attributes ["SYSTEM"] = ReadSystemLiteral (false); break; } } // The reader is positioned on the first 'S' of "SYSTEM". private string ReadSystemLiteral (bool expectSYSTEM) { if(expectSYSTEM) { Expect ("SYSTEM"); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required after 'SYSTEM'."); } else SkipWhitespace (); int quoteChar = ReadChar (); // apos or quot int startPos = currentTag.Length; int c = 0; while(c != quoteChar) { c = ReadChar (); if(c < 0) throw new XmlException (this as IXmlLineInfo,"Unexpected end of stream in ExternalID."); } return currentTag.ToString (startPos, currentTag.Length - 1 - startPos); } private string ReadPubidLiteral() { Expect ("PUBLIC"); if (!SkipWhitespace ()) throw new XmlException (this as IXmlLineInfo, "Whitespace is required after 'PUBLIC'."); int quoteChar = ReadChar (); int startPos = currentTag.Length; int c = 0; while(c != quoteChar) { c = ReadChar (); if(c < 0) throw new XmlException (this as IXmlLineInfo,"Unexpected end of stream in ExternalID."); if(c != quoteChar && !XmlChar.IsPubidChar (c)) throw new XmlException (this as IXmlLineInfo,"character '" + (char)c + "' not allowed for PUBLIC ID"); } return currentTag.ToString (startPos, currentTag.Length - 1 - startPos); } // The reader is positioned on the first character // of the name. internal string ReadName () { return ReadNameOrNmToken(false); } // The reader is positioned on the first character // of the name. private string ReadNmToken () { return ReadNameOrNmToken(true); } private string ReadNameOrNmToken(bool isNameToken) { int ch = PeekChar (); if(isNameToken) { if (!XmlChar.IsNameChar ((char) ch)) throw new XmlException (this as IXmlLineInfo,String.Format ("a nmtoken did not start with a legal character {0} ({1})", ch, (char)ch)); } else { if (!XmlChar.IsFirstNameChar (ch)) throw new XmlException (this as IXmlLineInfo,String.Format ("a name did not start with a legal character {0} ({1})", ch, (char)ch)); } nameLength = 0; AppendNameChar (ReadChar ()); while (XmlChar.IsNameChar (PeekChar ())) { AppendNameChar (ReadChar ()); } return CreateNameString (); } // Read the next character and compare it against the // specified character. private void Expect (int expected) { int ch = ReadChar (); if (ch != expected) { throw new XmlException (this as IXmlLineInfo, String.Format ( "expected '{0}' ({1:X}) but found '{2}' ({3:X})", (char)expected, expected, (char)ch, ch)); } } private void Expect (string expected) { int len = expected.Length; for(int i=0; i< len; i++) Expect (expected[i]); } // Does not consume the first non-whitespace character. private bool SkipWhitespace () { //FIXME: Should not skip if whitespaceHandling == WhiteSpaceHandling.None bool skipped = XmlChar.IsWhitespace (PeekChar ()); while (XmlChar.IsWhitespace (PeekChar ())) ReadChar (); return skipped; } private void ReadWhitespace () { if (currentState == XmlNodeType.None) currentState = XmlNodeType.XmlDeclaration; ClearValueBuffer (); int ch = PeekChar (); do { AppendValueChar (ReadChar ()); } while ((ch = PeekChar ()) != -1 && XmlChar.IsWhitespace (ch)); if (currentState == XmlNodeType.Element && ch != -1 && ch != '<') ReadText (false); else SetProperties (XmlNodeType.Whitespace, String.Empty, false, valueBuffer, true); return; // (PeekChar () != -1); } // read entity reference from attribute string and if parsable then return the value. private string ReadAttributeValueReference () { int endEntityPosition = attributeString.IndexOf(';', attributeValuePos); if (endEntityPosition < 0) throw new XmlException ("Insufficient markup of entity reference"); string entityName = attributeString.Substring (attributeValuePos + 1, endEntityPosition - attributeValuePos - 1); attributeValuePos = endEntityPosition + 1; if(entityName [0] == '#') { char c; // character entity if(entityName [1] == 'x') { // hexadecimal c = (char) int.Parse ("0" + entityName.Substring (2), System.Globalization.NumberStyles.HexNumber); } else { // decimal c = (char) int.Parse (entityName.Substring (1)); } return c.ToString(); } else { switch(entityName) { case "lt": return "<"; case "gt": return ">"; case "amp": return "&"; case "quot": return "\""; case "apos": return "'"; default: return null; } } } private string UnescapeAttributeValue (string unresolved) { if(unresolved == null) return null; // trim start/end edge of quotation character. return Dereference (unresolved.Substring (1, unresolved.Length - 2), true); } private string Dereference (string unresolved, bool expandPredefined) { StringBuilder resolved = new StringBuilder(); int pos = 0; int next = unresolved.IndexOf ('&'); if(next < 0) return unresolved; while(next >= 0) { if(pos < next) resolved.Append (unresolved.Substring (pos, next - pos));// - 1); int endPos = unresolved.IndexOf (';', next+1); string entityName = unresolved.Substring (next + 1, endPos - next - 1); if(entityName [0] == '#') { char c; // character entity if(entityName [1] == 'x') { // hexadecimal c = (char) int.Parse ("0" + entityName.Substring (2), System.Globalization.NumberStyles.HexNumber); } else { // decimal c = (char) int.Parse (entityName.Substring (1)); } resolved.Append (c); } else { char predefined = XmlChar.GetPredefinedEntity (entityName); if (expandPredefined && predefined != 0) resolved.Append (predefined); else // With respect to "Value", MS document is helpless // and the implemention returns inconsistent value // (e.g. XML: "&ent; &ent;" ---> Value: "&ent; &ent;".) resolved.Append ("&" + entityName + ";"); } pos = endPos + 1; if(pos > unresolved.Length) break; next = unresolved.IndexOf('&', pos); } resolved.Append (unresolved.Substring(pos)); return resolved.ToString(); } #endregion } }