///////////////////////////////////////////////////////////////////////////// // Name: xml/xml.h // Purpose: interface of wxXmlNode, wxXmlAttribute, wxXmlDocument // Author: wxWidgets team // Licence: wxWindows licence ///////////////////////////////////////////////////////////////////////////// /// Represents XML node type. enum wxXmlNodeType { // note: values are synchronized with xmlElementType from libxml wxXML_ELEMENT_NODE = 1, wxXML_ATTRIBUTE_NODE = 2, wxXML_TEXT_NODE = 3, wxXML_CDATA_SECTION_NODE = 4, wxXML_ENTITY_REF_NODE = 5, wxXML_ENTITY_NODE = 6, wxXML_PI_NODE = 7, wxXML_COMMENT_NODE = 8, wxXML_DOCUMENT_NODE = 9, wxXML_DOCUMENT_TYPE_NODE = 10, wxXML_DOCUMENT_FRAG_NODE = 11, wxXML_NOTATION_NODE = 12, wxXML_HTML_DOCUMENT_NODE = 13 }; /** @class wxXmlNode Represents a node in an XML document. See wxXmlDocument. Each node is named and depending on the node type it may also hold content or be given attributes. The two most common node types are @c wxXML_ELEMENT_NODE and @c wxXML_TEXT_NODE. @c wxXML_ELEMENT_NODE represents a pair of XML element tags, whilst @c wxXML_TEXT_NODE represents the text value that can belong to the element. A @c wxXML_ELEMENT_NODE has a title, and optionally attributes, but does not have any content. A @c wxXML_TEXT_NODE does not have a title or attributes but should normally have content. For example: in the XML fragment \hi\ there is an element node with the name @c title and a single text node child with the text @c hi as content. A @c wxXML_PI_NODE represents a Processing Instruction (PI) node with the name parameter set as the target and the contents parameter set as the instructions. Note that whilst the PI instructions are often in the form of pseudo-attributes, these do not use the node's attribute member. It is the user's responsibility to code and decode the PI instruction text. The @c wxXML_DOCUMENT_TYPE_NODE is not implemented at this time. Instead, you should get and set the DOCTYPE values using the wxXmlDocument class. If @c wxUSE_UNICODE is 0, all strings are encoded in the encoding given to wxXmlDocument::Load (default is UTF-8). @note Once a wxXmlNode has been added to a wxXmlDocument it becomes owned by the document and this has two implications. Firstly, the wxXmlDocument takes responsibility for deleting the node so the user should not @c delete it; and secondly, a wxXmlNode must always be created on the heap and never on the stack. @library{wxxml} @category{xml} @see wxXmlDocument, wxXmlDoctype, wxXmlAttribute */ class wxXmlNode { public: /** Creates this XML node and inserts it into the XML tree as a child of the specified parent. Once added, the XML tree takes ownership of this object and there is no need to delete it. @param parent The parent node to which append this node instance. If this argument is @NULL this new node will be floating and it can be appended later to another one using the AddChild() or InsertChild() functions. Otherwise the child is added to the XML tree by this constructor and it shouldn't be done again. @param type One of the ::wxXmlNodeType enumeration value. @param name The name of the node. This is the string which appears between angular brackets. @param content The content of the node. Only meaningful when type is @c wxXML_TEXT_NODE or @c wxXML_CDATA_SECTION_NODE. @param attrs If not @NULL, this wxXmlAttribute object and its eventual siblings are attached to the node. @param next If not @NULL, this node and its eventual siblings are attached to the node. @param lineNo Number of line this node was present at in input file or -1. */ wxXmlNode(wxXmlNode* parent, wxXmlNodeType type, const wxString& name, const wxString& content = wxEmptyString, wxXmlAttribute* attrs = NULL, wxXmlNode* next = NULL, int lineNo = -1); /** A simplified version of the first constructor form, assuming a @NULL parent. @param type One of the ::wxXmlNodeType enumeration value. @param name The name of the node. This is the string which appears between angular brackets. @param content The content of the node. Only meaningful when type is @c wxXML_TEXT_NODE or @c wxXML_CDATA_SECTION_NODE. @param lineNo Number of line this node was present at in input file or -1. */ wxXmlNode(wxXmlNodeType type, const wxString& name, const wxString& content = wxEmptyString, int lineNo = -1); /** Copy constructor. Note that this does NOT copy siblings and parent pointer, i.e. GetParent() and GetNext() will return @NULL after using copy ctor and are never unmodified by operator=(). On the other hand, it DOES copy children and attributes. */ wxXmlNode(const wxXmlNode& node); /** The virtual destructor. Deletes attached children and attributes. */ virtual ~wxXmlNode(); /** Appends a attribute with given @a name and @a value to the list of attributes for this node. */ virtual void AddAttribute(const wxString& name, const wxString& value); /** Appends given attribute to the list of attributes for this node. */ virtual void AddAttribute(wxXmlAttribute* attr); /** Adds node @a child as the last child of this node. Once added, the XML tree takes ownership of this object and there is no need to delete it. @note Note that this function works in O(n) time where @e n is the number of existing children. Consequently, adding large number of child nodes using this method can be expensive, because it has O(n^2) time complexity in number of nodes to be added. Use InsertChildAfter() to populate XML tree in linear time. @see InsertChild(), InsertChildAfter() */ virtual void AddChild(wxXmlNode* child); /** Removes the first attributes which has the given @a name from the list of attributes for this node. */ virtual bool DeleteAttribute(const wxString& name); /** Returns true if a attribute named attrName could be found. The value of that attribute is saved in value (which must not be @NULL). */ bool GetAttribute(const wxString& attrName, wxString* value) const; /** Returns the value of the attribute named @a attrName if it does exist. If it does not exist, the @a defaultVal is returned. */ wxString GetAttribute(const wxString& attrName, const wxString& defaultVal = wxEmptyString) const; /** Return a pointer to the first attribute of this node. */ wxXmlAttribute* GetAttributes() const; /** Returns the first child of this node. To get a pointer to the second child of this node (if it does exist), use the GetNext() function on the returned value. */ wxXmlNode* GetChildren() const; /** Returns the content of this node. Can be an empty string. Be aware that for nodes of type @c wxXML_ELEMENT_NODE (the most used node type) the content is an empty string. See GetNodeContent() for more details. */ const wxString& GetContent() const; /** Returns the number of nodes which separate this node from @c grandparent. This function searches only the parents of this node until it finds @a grandparent or the @NULL node (which is the parent of non-linked nodes or the parent of a wxXmlDocument's root element node). */ int GetDepth(wxXmlNode* grandparent = NULL) const; /** Returns a flag indicating whether encoding conversion is necessary when saving. The default is @false. You can improve saving efficiency considerably by setting this value. */ bool GetNoConversion() const; /** Returns line number of the node in the input XML file or @c -1 if it is unknown. */ int GetLineNumber() const; /** Returns the name of this node. Can be an empty string (e.g. for nodes of type @c wxXML_TEXT_NODE or @c wxXML_CDATA_SECTION_NODE). */ const wxString& GetName() const; /** Returns a pointer to the sibling of this node or @NULL if there are no siblings. */ wxXmlNode* GetNext() const; /** Returns the content of the first child node of type @c wxXML_TEXT_NODE or @c wxXML_CDATA_SECTION_NODE. This function is very useful since the XML snippet @c "tagnametagcontent/tagname" is represented by expat with the following tag tree: @code wxXML_ELEMENT_NODE name="tagname", content="" |-- wxXML_TEXT_NODE name="", content="tagcontent" @endcode or eventually: @code wxXML_ELEMENT_NODE name="tagname", content="" |-- wxXML_CDATA_SECTION_NODE name="", content="tagcontent" @endcode An empty string is returned if the node has no children of type @c wxXML_TEXT_NODE or @c wxXML_CDATA_SECTION_NODE, or if the content of the first child of such types is empty. */ wxString GetNodeContent() const; /** Returns a pointer to the parent of this node or @NULL if this node has no parent. */ wxXmlNode* GetParent() const; /** Returns the type of this node. */ wxXmlNodeType GetType() const; /** Returns @true if this node has a attribute named @a attrName. */ bool HasAttribute(const wxString& attrName) const; /** Inserts the @a child node immediately before @a followingNode in the children list. Once inserted, the XML tree takes ownership of the new child and there is no need to delete it. @return @true if @a followingNode has been found and the @a child node has been inserted. @note For historical reasons, @a followingNode may be @NULL. In that case, then @a child is prepended to the list of children and becomes the first child of this node, i.e. it behaves identically to using the first children (as returned by GetChildren()) for @a followingNode). @see AddChild(), InsertChildAfter() */ virtual bool InsertChild(wxXmlNode* child, wxXmlNode* followingNode); /** Inserts the @a child node immediately after @a precedingNode in the children list. Once inserted, the XML tree takes ownership of the new child and there is no need to delete it. @return @true if @a precedingNode has been found and the @a child node has been inserted. @param child The child to insert. @param precedingNode The node to insert @a child after. As a special case, this can be @NULL if this node has no children yet -- in that case, @a child will become this node's only child node. @since 2.8.8 @see InsertChild(), AddChild() */ virtual bool InsertChildAfter(wxXmlNode* child, wxXmlNode* precedingNode); /** Returns @true if the content of this node is a string containing only whitespaces (spaces, tabs, new lines, etc). Note that this function is locale-independent since the parsing of XML documents must always produce the exact same tree regardless of the locale it runs under. */ bool IsWhitespaceOnly() const; /** Removes the given node from the children list. Returns @true if the node was found and removed or @false if the node could not be found. Note that the caller is responsible for deleting the removed node in order to avoid memory leaks. */ virtual bool RemoveChild(wxXmlNode* child); /** Sets as first attribute the given wxXmlAttribute object. The caller is responsible for deleting any previously present attributes attached to this node. */ void SetAttributes(wxXmlAttribute* attr); /** Sets as first child the given node. The caller is responsible for deleting any previously present children node. */ void SetChildren(wxXmlNode* child); /** Sets the content of this node. */ void SetContent(const wxString& con); /** Sets the name of this node. */ void SetName(const wxString& name); /** Sets as sibling the given node. The caller is responsible for deleting any previously present sibling node. */ void SetNext(wxXmlNode* next); /** Sets a flag to indicate whether encoding conversion is necessary when saving. The default is @false. You can improve saving efficiency considerably by setting this value. */ void SetNoConversion(bool noconversion); /** Sets as parent the given node. The caller is responsible for deleting any previously present parent node. */ void SetParent(wxXmlNode* parent); /** Sets the type of this node. */ void SetType(wxXmlNodeType type); /** See the copy constructor for more info. */ wxXmlNode& operator=(const wxXmlNode& node); }; /** @class wxXmlAttribute Represents a node attribute. Example: in \, @c src is an attribute with value @c hello.gif and @c id is an attribute with value @c 3. @library{wxxml} @category{xml} @see wxXmlDocument, wxXmlNode */ class wxXmlAttribute { public: /** Default constructor. */ wxXmlAttribute(); /** Creates the attribute with given @a name and @a value. If @a next is not @NULL, then sets it as sibling of this attribute. */ wxXmlAttribute(const wxString& name, const wxString& value, wxXmlAttribute* next = NULL); /** The virtual destructor. */ virtual ~wxXmlAttribute(); /** Returns the name of this attribute. */ wxString GetName() const; /** Returns the sibling of this attribute or @NULL if there are no siblings. */ wxXmlAttribute* GetNext() const; /** Returns the value of this attribute. */ wxString GetValue() const; /** Sets the name of this attribute. */ void SetName(const wxString& name); /** Sets the sibling of this attribute. */ void SetNext(wxXmlAttribute* next); /** Sets the value of this attribute. */ void SetValue(const wxString& value); }; /** @class wxXmlDoctype Represents a DOCTYPE Declaration. Example DOCTYPE: \. In the above example, "plist" is the name of root element, "-//Apple//DTD PLIST 1.0//EN" (without the quotes) is the public identifier and "http://www.apple.com/DTDs/PropertyList-1.0.dtd" (again, without the quotes) is the system identifier. A valid DOCTYPE exists in one of following forms: 1. A root element name. 2. A root element name and a system identifier. 3. A root element name, a system identifier and a public identifier. 4. A root element name and a public identifier. Although this form is not valid XML it is valid for SMGL. The DOCTYPE may also contain an internal subset of declarations added between square brackets at the end. These have not been implemented at this time. @since 3.1.0 @library{wxxml} @category{xml} @see wxXmlDocument */ class wxXmlDoctype { public: /** Creates and possible initializes the DOCTYPE. @param name The root name. @param sysid The system identifier. @param pubid The public identifier. */ wxXmlDoctype(const wxString& rootName = wxString(), const wxString& systemId = wxString(), const wxString& publicId = wxString()); /** Removes all the DOCTYPE values. */ void Clear(); /** Returns the root name of the document. */ const wxString& GetRootName() const; /** Returns the system id of the document. */ const wxString& GetSystemId() const; /** Returns the public id of the document. */ const wxString& GetPublicId() const; /** Returns the formatted DOCTYPE contents. This consists of all the text shown between the opening "" of a DOCTYPE declaration. If this object is empty or invalid, i.e. IsValid() returns false, this method returns an empty string. */ wxString GetFullString() const; /** Returns true if the contents can produce a valid DOCTYPE string. For an object to be valid, it must have a non-empty root name and a valid system identifier (currently the validity checks of the latter are limited to checking that it doesn't contain both single and double quotes). */ bool IsValid() const; }; //* Special indentation value for wxXmlDocument::Save. #define wxXML_NO_INDENTATION (-1) //* Flags for wxXmlDocument::Load. enum wxXmlDocumentLoadFlag { wxXMLDOC_NONE, wxXMLDOC_KEEP_WHITESPACE_NODES }; /** @class wxXmlDocument This class holds XML data/document as parsed by XML parser in the root node. wxXmlDocument internally uses the expat library which comes with wxWidgets to parse the given stream. A wxXmlDocument is in fact a list of wxXmlNode organised into a structure that reflects the XML tree being represented by the document. @note Ownership is passed to the XML tree as each wxXmlNode is added to it, and this has two implications. Firstly, the wxXmlDocument takes responsibility for deleting the node so the user should not @c delete it; and secondly, a wxXmlNode must always be created on the heap and never on the stack. A simple example of using XML classes is: @code wxXmlDocument doc; if (!doc.Load("myfile.xml")) return false; // Start processing the XML file. if (doc.GetRoot()->GetName() != "myroot-node") return false; // Examine prologue. wxXmlNode *prolog = doc.GetDocumentNode()->GetChildren(); while (prolog) { if (prolog->GetType() == wxXML_PI_NODE && prolog->GetName() == "target") { // Process Process Instruction (PI) contents. wxString pi = prolog->GetContent(); ... } } wxXmlNode *child = doc.GetRoot()->GetChildren(); while (child) { if (child->GetName() == "tag1") { // Process text enclosed by tag1/tag1. wxString content = child->GetNodeContent(); ... // Process attributes of tag1. wxString attrvalue1 = child->GetAttribute("attr1", "default-value"); wxString attrvalue2 = child->GetAttribute("attr2", "default-value"); ... } else if (child->GetName() == "tag2") { // Process tag2 ... } child = child->GetNext(); } @endcode Note that if you want to preserve the original formatting of the loaded file including whitespaces and indentation, you need to turn off whitespace-only textnode removal and automatic indentation. For example: @code wxXmlDocument doc; doc.Load("myfile.xml", "UTF-8", wxXMLDOC_KEEP_WHITESPACE_NODES); // myfile2.xml will be identical to myfile.xml saving it this way: doc.Save("myfile2.xml", wxXML_NO_INDENTATION); @endcode Using default parameters, you will get a reformatted document which in general is different from the original loaded content: @code wxXmlDocument doc; doc.Load("myfile.xml"); doc.Save("myfile2.xml"); // myfile2.xml != myfile.xml @endcode wxXmlDocument can also be used to create documents. The following code gives an example of creating a simple document with two nested element nodes, the second of which has an attribute, and a text node. It also demonstrates how to write the resulting output to a wxString: @code // Create a document and add the root node. wxXmlDocument xmlDoc; wxXmlNode* root = new wxXmlNode(NULL, wxXML_ELEMENT_NODE, "Root"); xmlDoc.SetRoot(root); // Add some XML. wxXmlNode* library = new wxXmlNode (root, wxXML_ELEMENT_NODE, "Library"); library->AddAttribute("type", "CrossPlatformList"); wxXmlNode* name = new wxXmlNode(library, wxXML_ELEMENT_NODE, "Name"); name->AddChild(new wxXmlNode(wxXML_TEXT_NODE, "", "wxWidgets")); // Write the output to a wxString. wxStringOutputStream stream; xmlDoc.Save(stream); @endcode This will produce a document that looks something like the following: @code wxWidgets @endcode If the root name value of the DOCTYPE is set, either by loading a file with a DOCTYPE declaration or by setting it directly with the SetDoctype member, then a DOCTYPE declaration will be added immediately after the XML declaration. @library{wxxml} @category{xml} @see wxXmlNode, wxXmlAttribute, wxXmlDoctype */ class wxXmlDocument : public wxObject { public: /** Default constructor. */ wxXmlDocument(); /** Copy constructor. Deep copies all the XML tree of the given document. */ wxXmlDocument(const wxXmlDocument& doc); /** Loads the given filename using the given encoding. See Load(). */ wxXmlDocument(const wxString& filename, const wxString& encoding = "UTF-8")); /** Loads the XML document from given stream using the given encoding. See Load(). */ wxXmlDocument(wxInputStream& stream, const wxString& encoding = "UTF-8"); /** Virtual destructor. Frees the document root node. */ virtual ~wxXmlDocument(); /** Appends a Process Instruction or Comment node to the document prologue. Calling this function will create a prologue or attach the node to the end of an existing prologue. @since 2.9.2 */ void AppendToProlog(wxXmlNode* node); /** Detaches the document node and returns it. The document node will be set to @NULL and thus IsOk() will return @false after calling this function. Note that the caller is responsible for deleting the returned node in order to avoid memory leaks. @since 2.9.2 */ wxXmlNode* DetachDocumentNode(); /** Detaches the root entity node and returns it. After calling this function, the document node will remain together with any prologue nodes, but IsOk() will return @false since the root entity will be missing. Note that the caller is responsible for deleting the returned node in order to avoid memory leaks. */ wxXmlNode* DetachRoot(); /** Returns encoding of in-memory representation of the document (same as passed to Load() or constructor, defaults to UTF-8). @note this is meaningless in Unicode build where data are stored as @c wchar_t*. */ wxString GetEncoding() const; /** Returns encoding of document (may be empty). @note This is the encoding original file was saved in, @b not the encoding of in-memory representation! */ const wxString& GetFileEncoding() const; /** Returns the DOCTYPE declaration data for the document. @since 3.1.0 */ const wxXmlDoctype& GetDoctype() const; /** Returns the output line ending format used for documents. @since 3.1.1 */ wxTextFileType GetFileType() const; /** Returns the output line ending string used for documents. This string is determined by the last call to SetFileType(). @since 3.1.1 */ wxString GetEOL() const; /** Returns the document node of the document. @since 2.9.2 */ wxXmlNode* GetDocumentNode() const; /** Returns the root element node of the document. */ wxXmlNode* GetRoot() const; /** Returns the version of document. This is the value in the @c \ header of the XML document. If the version attribute was not explicitly given in the header, this function returns an empty string. */ const wxString& GetVersion() const; /** Returns @true if the document has been loaded successfully. */ bool IsOk() const; /** Parses @a filename as an xml document and loads its data. If @a flags does not contain wxXMLDOC_KEEP_WHITESPACE_NODES, then, while loading, all nodes of type @c wxXML_TEXT_NODE (see wxXmlNode) are automatically skipped if they contain whitespaces only. The removal of these nodes makes the load process slightly faster and requires less memory however makes impossible to recreate exactly the loaded text with a Save() call later. Read the initial description of this class for more info. Returns true on success, false otherwise. */ virtual bool Load(const wxString& filename, const wxString& encoding = "UTF-8", int flags = wxXMLDOC_NONE); /** Like Load(const wxString&, const wxString&, int) but takes the data from given input stream. */ virtual bool Load(wxInputStream& stream, const wxString& encoding = "UTF-8", int flags = wxXMLDOC_NONE); /** Saves XML tree creating a file named with given string. If @a indentstep is greater than or equal to zero, then, while saving, an automatic indentation is added with steps composed by indentstep spaces. If @a indentstep is @c wxXML_NO_INDENTATION, then, automatic indentation is turned off. */ virtual bool Save(const wxString& filename, int indentstep = 2) const; /** Saves XML tree in the given output stream. See Save(const wxString&, int) for a description of @a indentstep. */ virtual bool Save(wxOutputStream& stream, int indentstep = 2) const; /** Sets the document node of this document. Deletes any previous document node. Use DetachDocumentNode() and then SetDocumentNode() if you want to replace the document node without deleting the old document tree. @since 2.9.2 */ void SetDocumentNode(wxXmlNode* node); /** Sets the encoding of the document. */ void SetEncoding(const wxString& enc); /** Sets the encoding of the file which will be used to save the document. */ void SetFileEncoding(const wxString& encoding); /** Sets the data which will appear in the DOCTYPE declaration when the document is saved. @since 3.1.0 */ void SetDoctype(const wxXmlDoctype& doctype); /** Sets the output line ending formats when the document is saved. By default Unix file type is used, i.e. a single ASCII LF (10) character is used at the end of lines. @since 3.1.1 */ void SetFileType(wxTextFileType fileType); /** Sets the root element node of this document. Will create the document node if necessary. Any previous root element node is deleted. */ void SetRoot(wxXmlNode* node); /** Sets the version of the XML file which will be used to save the document. */ void SetVersion(const wxString& version); /** Deep copies the given document. */ wxXmlDocument& operator=(const wxXmlDocument& doc); /** Get expat library version information. @since 2.9.2 @see wxVersionInfo */ static wxVersionInfo GetLibraryVersionInfo(); };