From: Vadim Zeitlin Date: Wed, 30 Mar 2011 14:16:19 +0000 (+0000) Subject: Add support for elements preceding the document node in wxXML. X-Git-Url: https://git.saurik.com/wxWidgets.git/commitdiff_plain/cee8636e94d49319b1c108d8b2d826e080e2e366 Add support for elements preceding the document node in wxXML. This is mainly useful for parsing and generating processing instructions but can be used for any kind of elements, e.g. also comments, occurring before the document node in XML documents. Closes #11593. git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@67346 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775 --- diff --git a/docs/changes.txt b/docs/changes.txt index 50377642f3..b1aac62be2 100644 --- a/docs/changes.txt +++ b/docs/changes.txt @@ -454,6 +454,7 @@ All: - Added precision parameter to wxString::From[C]Double(). - Added wxThread::Wait() and Delete() "wait mode" parameter (Catalin Raceanu). - Allow showing milliseconds in wxLog time stamps (tan). +- Added support for processing instructions in wxXmlDocument (Nick Matthews). Unix: diff --git a/include/wx/xml/xml.h b/include/wx/xml/xml.h index 4a40d32506..fbd6faa4ac 100644 --- a/include/wx/xml/xml.h +++ b/include/wx/xml/xml.h @@ -257,7 +257,7 @@ public: const wxString& encoding = wxT("UTF-8")); wxXmlDocument(wxInputStream& stream, const wxString& encoding = wxT("UTF-8")); - virtual ~wxXmlDocument() { wxDELETE(m_root); } + virtual ~wxXmlDocument() { wxDELETE(m_docNode); } wxXmlDocument(const wxXmlDocument& doc); wxXmlDocument& operator=(const wxXmlDocument& doc); @@ -273,10 +273,13 @@ public: virtual bool Save(const wxString& filename, int indentstep = 2) const; virtual bool Save(wxOutputStream& stream, int indentstep = 2) const; - bool IsOk() const { return m_root != NULL; } + bool IsOk() const { return GetRoot() != NULL; } // Returns root node of the document. - wxXmlNode *GetRoot() const { return m_root; } + wxXmlNode *GetRoot() const; + // Returns the document node. + wxXmlNode *GetDocumentNode() const { return m_docNode; } + // Returns version of document (may be empty). const wxString& GetVersion() const { return m_version; } @@ -286,10 +289,13 @@ public: const wxString& GetFileEncoding() const { return m_fileEncoding; } // Write-access methods: - wxXmlNode *DetachRoot() { wxXmlNode *old=m_root; m_root=NULL; return old; } - void SetRoot(wxXmlNode *node) { wxDELETE(m_root); m_root = node; } + wxXmlNode *DetachDocumentNode() { wxXmlNode *old=m_docNode; m_docNode=NULL; return old; } + void SetDocumentNode(wxXmlNode *node) { wxDELETE(m_docNode); m_docNode = node; } + wxXmlNode *DetachRoot(); + void SetRoot(wxXmlNode *node); void SetVersion(const wxString& version) { m_version = version; } void SetFileEncoding(const wxString& encoding) { m_fileEncoding = encoding; } + void AppendToProlog(wxXmlNode *node); #if !wxUSE_UNICODE // Returns encoding of in-memory representation of the document @@ -307,7 +313,7 @@ private: #if !wxUSE_UNICODE wxString m_encoding; #endif - wxXmlNode *m_root; + wxXmlNode *m_docNode; void DoCopy(const wxXmlDocument& doc); diff --git a/interface/wx/xml/xml.h b/interface/wx/xml/xml.h index bdd3344bb8..8605491dd1 100644 --- a/interface/wx/xml/xml.h +++ b/interface/wx/xml/xml.h @@ -40,6 +40,11 @@ enum wxXmlNodeType @c title and irrelevant content and one child of type @c wxXML_TEXT_NODE with @c hi as content. + The @c wxXML_PI_NODE type sets the name to the PI target and the contents to + the instructions. Note that whilst the PI instructions are often in the form + of pseudo-attributes these do not use the nodes attribute system. It is the users + responsibility to code and decode the instruction text. + If @c wxUSE_UNICODE is 0, all strings are encoded in the encoding given to wxXmlDocument::Load (default is UTF-8). @@ -179,7 +184,7 @@ public: 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 node). + nodes or the parent of a wxXmlDocument's root element node). */ int GetDepth(wxXmlNode* grandparent = NULL) const; @@ -215,14 +220,14 @@ public: is represented by expat with the following tag tree: @code - wxXML_ENTITY_NODE name="tagname", content="" + wxXML_ELEMENT_NODE name="tagname", content="" |-- wxXML_TEXT_NODE name="", content="tagcontent" @endcode or eventually: @code - wxXML_ENTITY_NODE name="tagname", content="" + wxXML_ELEMENT_NODE name="tagname", content="" |-- wxXML_CDATA_SECTION_NODE name="", content="tagcontent" @endcode @@ -449,6 +454,20 @@ public: 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 contents + wxString pi = prolog->GetContent(); + + ... + + } + } + wxXmlNode *child = doc.GetRoot()->GetChildren(); while (child) { @@ -533,13 +552,37 @@ public: virtual ~wxXmlDocument(); /** - Detaches the document root node and returns it. + 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 root node will be set to @NULL and thus IsOk() will + 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 reponsible for deleting the returned node in order + to avoid memory leaks. */ wxXmlNode* DetachRoot(); @@ -560,7 +603,14 @@ public: const wxString& GetFileEncoding() const; /** - Returns the root node of the document. + 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; @@ -620,7 +670,18 @@ public: virtual bool Save(wxOutputStream& stream, int indentstep = 2) const; /** - Sets the enconding of the document. + 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); @@ -630,9 +691,10 @@ public: void SetFileEncoding(const wxString& encoding); /** - Sets the root node of this document. Deletes previous root node. - Use DetachRoot() and then SetRoot() if you want to replace the root - node without deleting the old document tree. + 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); diff --git a/src/xml/xml.cpp b/src/xml/xml.cpp index 922332b1d6..0e82fffd01 100644 --- a/src/xml/xml.cpp +++ b/src/xml/xml.cpp @@ -396,7 +396,7 @@ bool wxXmlNode::IsWhitespaceOnly() const //----------------------------------------------------------------------------- wxXmlDocument::wxXmlDocument() - : m_version(wxS("1.0")), m_fileEncoding(wxS("utf-8")), m_root(NULL) + : m_version(wxS("1.0")), m_fileEncoding(wxS("UTF-8")), m_docNode(NULL) { #if !wxUSE_UNICODE m_encoding = wxS("UTF-8"); @@ -404,20 +404,20 @@ wxXmlDocument::wxXmlDocument() } wxXmlDocument::wxXmlDocument(const wxString& filename, const wxString& encoding) - :wxObject(), m_root(NULL) + :wxObject(), m_docNode(NULL) { if ( !Load(filename, encoding) ) { - wxDELETE(m_root); + wxDELETE(m_docNode); } } wxXmlDocument::wxXmlDocument(wxInputStream& stream, const wxString& encoding) - :wxObject(), m_root(NULL) + :wxObject(), m_docNode(NULL) { if ( !Load(stream, encoding) ) { - wxDELETE(m_root); + wxDELETE(m_docNode); } } @@ -429,7 +429,7 @@ wxXmlDocument::wxXmlDocument(const wxXmlDocument& doc) wxXmlDocument& wxXmlDocument::operator=(const wxXmlDocument& doc) { - wxDELETE(m_root); + wxDELETE(m_docNode); DoCopy(doc); return *this; } @@ -442,10 +442,10 @@ void wxXmlDocument::DoCopy(const wxXmlDocument& doc) #endif m_fileEncoding = doc.m_fileEncoding; - if (doc.m_root) - m_root = new wxXmlNode(*doc.m_root); + if (doc.m_docNode) + m_docNode = new wxXmlNode(*doc.m_docNode); else - m_root = NULL; + m_docNode = NULL; } bool wxXmlDocument::Load(const wxString& filename, const wxString& encoding, int flags) @@ -464,7 +464,83 @@ bool wxXmlDocument::Save(const wxString& filename, int indentstep) const return Save(stream, indentstep); } +wxXmlNode *wxXmlDocument::GetRoot() const +{ + wxXmlNode *node = m_docNode; + if (node) + { + node = m_docNode->GetChildren(); + while (node != NULL && node->GetType() != wxXML_ELEMENT_NODE) + node = node->GetNext(); + } + return node; +} + +wxXmlNode *wxXmlDocument::DetachRoot() +{ + wxXmlNode *node = m_docNode; + if (node) + { + node = m_docNode->GetChildren(); + wxXmlNode *prev = NULL; + while (node != NULL && node->GetType() != wxXML_ELEMENT_NODE) + { + prev = node; + node = node->GetNext(); + } + if (node) + { + if (node == m_docNode->GetChildren()) + m_docNode->SetChildren(node->GetNext()); + + if (prev) + prev->SetNext(node->GetNext()); + + node->SetParent(NULL); + node->SetNext(NULL); + } + } + return node; +} +void wxXmlDocument::SetRoot(wxXmlNode *root) +{ + wxXmlNode *node = m_docNode; + if (node) + { + node = m_docNode->GetChildren(); + wxXmlNode *prev = NULL; + while (node != NULL && node->GetType() != wxXML_ELEMENT_NODE) + { + prev = node; + node = node->GetNext(); + } + if (node) + { + root->SetNext( node->GetNext() ); + wxDELETE(node); + } + if (prev) + prev->SetNext(root); + else + m_docNode->SetChildren(root); + } + else + { + m_docNode = new wxXmlNode(wxXML_DOCUMENT_NODE, wxEmptyString); + } + root->SetParent(m_docNode); +} + +void wxXmlDocument::AppendToProlog(wxXmlNode *node) +{ + if (!m_docNode) + m_docNode = new wxXmlNode(wxXML_DOCUMENT_NODE, wxEmptyString); + if (IsOk()) + m_docNode->InsertChild( node, GetRoot() ); + else + m_docNode->AddChild( node ); +} //----------------------------------------------------------------------------- // wxXmlDocument loading routines @@ -509,7 +585,6 @@ struct wxXmlParsingContext { wxXmlParsingContext() : conv(NULL), - root(NULL), node(NULL), lastChild(NULL), lastAsText(NULL), @@ -518,7 +593,6 @@ struct wxXmlParsingContext XML_Parser parser; wxMBConv *conv; - wxXmlNode *root; wxXmlNode *node; // the node being parsed wxXmlNode *lastChild; // the last child of "node" wxXmlNode *lastAsText; // the last _text_ child of "node" @@ -551,16 +625,8 @@ static void StartElementHnd(void *userData, const char *name, const char **atts) a += 2; } - if (ctx->root == NULL) - { - ctx->root = node; - } - else - { - ASSERT_LAST_CHILD_OK(ctx); - ctx->node->InsertChildAfter(node, ctx->lastChild); - } - + ASSERT_LAST_CHILD_OK(ctx); + ctx->node->InsertChildAfter(node, ctx->lastChild); ctx->lastAsText = NULL; ctx->lastChild = NULL; // our new node "node" has no children yet @@ -636,21 +702,29 @@ static void CommentHnd(void *userData, const char *data) { wxXmlParsingContext *ctx = (wxXmlParsingContext*)userData; - if (ctx->node) - { - wxXmlNode *commentnode = - new wxXmlNode(wxXML_COMMENT_NODE, - wxS("comment"), CharToString(ctx->conv, data), - XML_GetCurrentLineNumber(ctx->parser)); - - ASSERT_LAST_CHILD_OK(ctx); - ctx->node->InsertChildAfter(commentnode, ctx->lastChild); - ctx->lastChild = commentnode; - } - //else: ctx->node == NULL happens if there is a comment before - // the root element. We current don't have a way to represent - // these in wxXmlDocument (FIXME). + wxXmlNode *commentnode = + new wxXmlNode(wxXML_COMMENT_NODE, + wxS("comment"), CharToString(ctx->conv, data), + XML_GetCurrentLineNumber(ctx->parser)); + ASSERT_LAST_CHILD_OK(ctx); + ctx->node->InsertChildAfter(commentnode, ctx->lastChild); + ctx->lastChild = commentnode; + ctx->lastAsText = NULL; +} + +static void PIHnd(void *userData, const char *target, const char *data) +{ + wxXmlParsingContext *ctx = (wxXmlParsingContext*)userData; + + wxXmlNode *pinode = + new wxXmlNode(wxXML_PI_NODE, CharToString(ctx->conv, target), + CharToString(ctx->conv, data), + XML_GetCurrentLineNumber(ctx->parser)); + + ASSERT_LAST_CHILD_OK(ctx); + ctx->node->InsertChildAfter(pinode, ctx->lastChild); + ctx->lastChild = pinode; ctx->lastAsText = NULL; } @@ -718,6 +792,7 @@ bool wxXmlDocument::Load(wxInputStream& stream, const wxString& encoding, int fl wxXmlParsingContext ctx; bool done; XML_Parser parser = XML_ParserCreate(NULL); + wxXmlNode *root = new wxXmlNode(wxXML_DOCUMENT_NODE, wxEmptyString); ctx.encoding = wxS("UTF-8"); // default in absence of encoding="" ctx.conv = NULL; @@ -727,12 +802,14 @@ bool wxXmlDocument::Load(wxInputStream& stream, const wxString& encoding, int fl #endif ctx.removeWhiteOnlyNodes = (flags & wxXMLDOC_KEEP_WHITESPACE_NODES) == 0; ctx.parser = parser; + ctx.node = root; XML_SetUserData(parser, (void*)&ctx); XML_SetElementHandler(parser, StartElementHnd, EndElementHnd); XML_SetCharacterDataHandler(parser, TextHnd); XML_SetCdataSectionHandler(parser, StartCdataHnd, EndCdataHnd);; XML_SetCommentHandler(parser, CommentHnd); + XML_SetProcessingInstructionHandler(parser, PIHnd); XML_SetDefaultHandler(parser, DefaultHnd); XML_SetUnknownEncodingHandler(parser, UnknownEncodingHnd, NULL); @@ -759,11 +836,11 @@ bool wxXmlDocument::Load(wxInputStream& stream, const wxString& encoding, int fl SetVersion(ctx.version); if (!ctx.encoding.empty()) SetFileEncoding(ctx.encoding); - SetRoot(ctx.root); + SetDocumentNode(root); } else { - delete ctx.root; + delete root; } XML_ParserFree(parser); @@ -995,6 +1072,14 @@ bool OutputNode(wxOutputStream& stream, OutputString(stream, wxS("-->"), convMem, convFile); break; + case wxXML_PI_NODE: + rc = OutputString(stream, wxT("GetName(), convMem, convFile) && + OutputString(stream, wxT(" "), convMem, convFile) && + OutputString(stream, node->GetContent(), convMem, convFile) && + OutputString(stream, wxT("?>"), convMem, convFile); + break; + default: wxFAIL_MSG("unsupported node type"); rc = false; @@ -1023,17 +1108,24 @@ bool wxXmlDocument::Save(wxOutputStream& stream, int indentstep) const //else: file and in-memory encodings are the same, no conversion needed #endif - return OutputString(stream, - wxString::Format - ( - wxS("\n"), - GetVersion(), GetFileEncoding() - ), - convMem.get(), - convFile.get()) && - OutputNode(stream, GetRoot(), 0, - convMem.get(), convFile.get(), indentstep) && - OutputString(stream, wxS("\n"), convMem.get(), convFile.get()); + wxString dec = wxString::Format( + wxS("\n"), + GetVersion(), GetFileEncoding() + ); + bool rc = OutputString(stream, dec, convMem.get(), convFile.get()); + + wxXmlNode *node = GetDocumentNode(); + if ( node ) + node = node->GetChildren(); + + while( rc && node ) + { + rc = OutputNode(stream, node, 0, convMem.get(), + convFile.get(), indentstep) && + OutputString(stream, wxS("\n"), convMem.get(), convFile.get()); + node = node->GetNext(); + } + return rc; } /*static*/ wxVersionInfo wxXmlDocument::GetLibraryVersionInfo() diff --git a/tests/xml/xmltest.cpp b/tests/xml/xmltest.cpp index f04b10365f..153b9bcc4c 100644 --- a/tests/xml/xmltest.cpp +++ b/tests/xml/xmltest.cpp @@ -77,14 +77,20 @@ private: CPPUNIT_TEST( InsertChildAfter ); CPPUNIT_TEST( LoadSave ); CPPUNIT_TEST( CDATA ); + CPPUNIT_TEST( PI ); CPPUNIT_TEST( Escaping ); + CPPUNIT_TEST( DetachRoot ); + CPPUNIT_TEST( AppendToProlog ); CPPUNIT_TEST_SUITE_END(); void InsertChild(); void InsertChildAfter(); void LoadSave(); void CDATA(); + void PI(); void Escaping(); + void DetachRoot(); + void AppendToProlog(); DECLARE_NO_COPY_CLASS(XmlTestCase) }; @@ -150,6 +156,7 @@ void XmlTestCase::LoadSave() const char *xmlText = "\n" "\n" +" \n" " \n" " \n" " \n" @@ -194,6 +201,30 @@ void XmlTestCase::LoadSave() CPPUNIT_ASSERT_EQUAL( wxString(utf8xmlText), wxString(sos8.GetString().ToUTF8()) ); #endif // wxUSE_UNICODE + + const char *xmlTextProlog = +"\n" +"\n" +"\n" +"\n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +"\n" +"\n" + ; + + wxStringInputStream sisp(xmlTextProlog); + CPPUNIT_ASSERT( doc.Load(sisp, "UTF-8") ); + + wxStringOutputStream sosp; + CPPUNIT_ASSERT( doc.Save(sosp) ); + + CPPUNIT_ASSERT_EQUAL( xmlTextProlog, sosp.GetString() ); } void XmlTestCase::CDATA() @@ -221,6 +252,28 @@ void XmlTestCase::CDATA() CPPUNIT_ASSERT_EQUAL( "Giovanni Mittone", n->GetContent() ); } +void XmlTestCase::PI() +{ + const char *xmlText = + "\n" + "\n" + " \n" + "\n" + ; + + wxStringInputStream sis(xmlText); + wxXmlDocument doc; + CPPUNIT_ASSERT( doc.Load(sis) ); + + wxXmlNode *n = doc.GetRoot(); + CPPUNIT_ASSERT( n ); + + n = n->GetChildren(); + CPPUNIT_ASSERT( n ); + + CPPUNIT_ASSERT_EQUAL( "index=\"no\" follow=\"no\"", n->GetContent() ); +} + void XmlTestCase::Escaping() { // Verify that attribute values are escaped correctly, see @@ -243,3 +296,131 @@ void XmlTestCase::Escaping() CPPUNIT_ASSERT_EQUAL( xmlText, sos.GetString() ); } + +void XmlTestCase::DetachRoot() +{ + const char *xmlTextProlog = +"\n" +"\n" +"\n" +"\n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +"\n" +"\n" + ; + const char *xmlTextHtm = +"\n" +"\n" +" \n" +" Testing wxXml\n" +" \n" +" \n" +"

Some body text

\n" +" \n" +"\n" + ; + wxXmlDocument doc; + + wxStringInputStream sish(xmlTextHtm); + CPPUNIT_ASSERT( doc.Load(sish) ); + + wxXmlNode *root = doc.DetachRoot(); + + wxStringInputStream sisp(xmlTextProlog); + CPPUNIT_ASSERT( doc.Load(sisp) ); + + doc.SetRoot(root); + + wxStringOutputStream sos; + CPPUNIT_ASSERT( doc.Save(sos) ); + + const char *xmlTextResult1 = +"\n" +"\n" +"\n" +"\n" +" \n" +" Testing wxXml\n" +" \n" +" \n" +"

Some body text

\n" +" \n" +"\n" +"\n" + ; + CPPUNIT_ASSERT_EQUAL( xmlTextResult1, sos.GetString() ); + + wxStringInputStream sisp2(xmlTextProlog); + CPPUNIT_ASSERT( doc.Load(sisp2) ); + + root = doc.DetachRoot(); + + wxStringInputStream sish2(xmlTextHtm); + CPPUNIT_ASSERT( doc.Load(sish2) ); + + doc.SetRoot(root); + + wxStringOutputStream sos2; + CPPUNIT_ASSERT( doc.Save(sos2) ); + + const char *xmlTextResult2 = +"\n" +"\n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +" \n" +"\n" + ; + CPPUNIT_ASSERT_EQUAL( xmlTextResult2, sos2.GetString() ); +} + +void XmlTestCase::AppendToProlog() +{ + const char *xmlText = +"\n" +"\n" +"

Some text

\n" +"\n" + ; + wxXmlDocument rootdoc; + wxStringInputStream sis(xmlText); + CPPUNIT_ASSERT( rootdoc.Load(sis) ); + wxXmlNode *root = rootdoc.DetachRoot(); + + wxXmlNode *comment1 = new wxXmlNode(wxXML_COMMENT_NODE, "comment", + " 1st prolog entry "); + wxXmlNode *pi = new wxXmlNode(wxXML_PI_NODE, "xml-stylesheet", + "href=\"style.css\" type=\"text/css\""); + wxXmlNode *comment2 = new wxXmlNode(wxXML_COMMENT_NODE, "comment", + " 3rd prolog entry "); + + wxXmlDocument doc; + doc.AppendToProlog( comment1 ); + doc.AppendToProlog( pi ); + doc.SetRoot( root ); + doc.AppendToProlog( comment2 ); + + wxStringOutputStream sos; + CPPUNIT_ASSERT( doc.Save(sos) ); + + const char *xmlTextResult = +"\n" +"\n" +"\n" +"\n" +"\n" +"

Some text

\n" +"\n" + ; + CPPUNIT_ASSERT_EQUAL( xmlTextResult, sos.GetString() ); +}