\begin{verbatim}
wxXmlDocument doc;
-if (!doc.Load(wxT("myfile.xml"))
+if (!doc.Load(wxT("myfile.xml")))
return false;
// start processing the XML file
}
\end{verbatim}
+{\bf Note:} 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:
+
+\begin{verbatim}
+wxXmlDocument doc;
+doc.Load(wxT("myfile.xml"), wxT("UTF-8"), wxXMLDOC_KEEP_WHITESPACE_NODES);
+doc.Save(wxT("myfile2.xml"), wxXML_NO_INDENTATION); // myfile2.xml will be indentic to myfile.xml
+\end{verbatim}
+
+Using default parameters, you will get a reformatted document which in general is different from
+the original loaded content:
+
+\begin{verbatim}
+wxXmlDocument doc;
+doc.Load(wxT("myfile.xml"));
+doc.Save(wxT("myfile2.xml")); // myfile2.xml != myfile.xml
+\end{verbatim}
\wxheading{Derived from}
\func{}{wxXmlDocument}{\void}
-\func{}{wxXmlDocument}{\param{const wxString\& }{filename}, \param{const wxString\& }{encoding = wxT("UTF-8")}}
+\func{}{wxXmlDocument}{\param{const wxString\& }{filename}, \param{const wxString\& }{encoding = wxT("UTF-8")}, \param{int }{flags = wxXMLDOC_NONE}}
Loads the given {\it filename} using the given encoding. See \helpref{Load()}{wxxmldocumentload}.
-\func{}{wxXmlDocument}{\param{wxInputStream\& }{stream}, \param{const wxString\& }{encoding = wxT("UTF-8")}}
+\func{}{wxXmlDocument}{\param{wxInputStream\& }{stream}, \param{const wxString\& }{encoding = wxT("UTF-8")}, \param{int }{flags = wxXMLDOC_NONE}}
Loads the XML document from given stream using the given encoding. See \helpref{Load()}{wxxmldocumentload}.
\membersection{wxXmlDocument::Load}\label{wxxmldocumentload}
-\func{bool}{Load}{\param{const wxString\& }{filename}, \param{const wxString\& }{encoding = wxT("UTF-8")}}
+\func{bool}{Load}{\param{const wxString\& }{filename}, \param{const wxString\& }{encoding = wxT("UTF-8")}, \param{int }{flags = wxXMLDOC_NONE}}
-Parses {\it filename} as an xml document and loads data. Returns \true on success, \false otherwise.
+Parses {\it filename} as an xml document and loads its data.
-\func{bool}{Load}{\param{wxInputStream\& }{stream}, \param{const wxString\& }{encoding = wxT("UTF-8")}}
+If {\tt flags} does not contain {\tt wxXMLDOC_KEEP_WHITESPACE_NODES}, then, while loading, all nodes of
+type {\tt wxXML_TEXT_NODE} (see \helpref{wxXmlNode}{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 \helpref{Save}{wxxmldocumentsave} call later.
+Read the initial description of this class for more info.
+
+Returns \true on success, \false otherwise.
+
+\func{bool}{Load}{\param{wxInputStream\& }{stream}, \param{const wxString\& }{encoding = wxT("UTF-8")}, \param{int }{flags = wxXMLDOC_NONE}}
Like above but takes the data from given input stream.
\membersection{wxXmlDocument::Save}\label{wxxmldocumentsave}
-\constfunc{bool}{Save}{\param{const wxString\& }{filename}}
+\constfunc{bool}{Save}{\param{const wxString\& }{filename}, \param{int }{indentstep = 1}}
Saves XML tree creating a file named with given string.
-\constfunc{bool}{Save}{\param{wxOutputStream\& }{stream}}
+If {\tt indentstep} is greater than or equal to zero, then, while saving, an automatic indentation
+is added with steps composed by {\tt indentstep} spaces.
+If {\tt indentstep} is {\tt wxXML_NO_INDENTATION}, then, automatic indentation is turned off.
+
+\constfunc{bool}{Save}{\param{wxOutputStream\& }{stream}, \param{int }{indentstep = 1}}
-Saves XML tree in the given output stream.
+Saves XML tree in the given output stream. See other overload for a description of {\tt indentstep}.
\membersection{wxXmlDocument::SetEncoding}\label{wxxmldocumentsetencoding}
\membersection{wxXmlNode::GetChildren}\label{wxxmlnodegetchildren}
-\constfunc{wxXmlNode*}{GetChildren}{\param{void}{}}
+\constfunc{wxXmlNode*}{GetChildren}{\void}
Returns the first child of this node.
To get a pointer to the second child of this node (if it does exist), use the
content is an empty string. See \helpref{GetNodeContent()}{wxxmlnodegetnodecontent} for more details.
+\membersection{wxXmlNode::GetDepth}\label{wxxmlnodegetdepth}
+
+\constfunc{int}{GetDepth}{\param{wxXmlNode* }{grandparent = NULL}}
+
+Returns the number of nodes which separe this node from {\tt grandparent}.
+
+This function searches only the parents of this node until it finds {\tt grandparent}
+or the \NULL node (which is the parent of non-linked nodes or the parent of a
+\helpref{wxXmlDocument}{wxxmldocument}'s root node).
+
+
\membersection{wxXmlNode::GetNodeContent}\label{wxxmlnodegetnodecontent}
\constfunc{wxString}{GetNodeContent}{\void}
becomes the first child of this node.
Returns \true if {\it before\_node} has been found and the {\it child} node has been inserted.
+\membersection{wxXmlNode::IsWhitespaceOnly}\label{wxxmlnodecontainsiswhitespaceonly}
+
+\constfunc{bool}{IsWhitespaceOnly}{\void}
+
+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.
+
\membersection{wxXmlNode::RemoveChild}\label{wxxmlnoderemovechild}
\func{bool}{RemoveChild}{\param{wxXmlNode* }{child}}
wxString GetName() const { return m_name; }
wxString GetContent() const { return m_content; }
+ bool IsWhitespaceOnly() const;
+ int GetDepth(wxXmlNode *grandparent = NULL) const;
+
// Gets node content from wxXML_ENTITY_NODE
// The problem is, <tag>content<tag> is represented as
// wxXML_ENTITY_NODE name="tag", content=""
+// special indentation value for wxXmlDocument::Save
+#define wxXML_NO_INDENTATION (-1)
-
+// flags for wxXmlDocument::Load
+enum wxXmlDocumentLoadFlag
+{
+ wxXMLDOC_NONE = 0,
+ wxXMLDOC_KEEP_WHITESPACE_NODES = 1
+};
// This class holds XML data/document as parsed by XML parser.
// Parses .xml file and loads data. Returns TRUE on success, FALSE
// otherwise.
virtual bool Load(const wxString& filename,
- const wxString& encoding = wxT("UTF-8"));
+ const wxString& encoding = wxT("UTF-8"), int flags = wxXMLDOC_NONE);
virtual bool Load(wxInputStream& stream,
- const wxString& encoding = wxT("UTF-8"));
+ const wxString& encoding = wxT("UTF-8"), int flags = wxXMLDOC_NONE);
// Saves document as .xml file.
- virtual bool Save(const wxString& filename) const;
- virtual bool Save(wxOutputStream& stream) const;
+ virtual bool Save(const wxString& filename, int indentstep = 1) const;
+ virtual bool Save(wxOutputStream& stream, int indentstep = 1) const;
bool IsOk() const { return m_root != NULL; }
IMPLEMENT_CLASS(wxXmlDocument, wxObject)
+// a private utility used by wxXML
+static bool wxIsWhiteOnly(const wxChar *buf);
+
//-----------------------------------------------------------------------------
// wxXmlNode
return wxEmptyString;
}
+int wxXmlNode::GetDepth(wxXmlNode *grandparent) const
+{
+ const wxXmlNode *n = this;
+ int ret = -1;
+
+ do
+ {
+ ret++;
+ n = n->GetParent();
+ if (n == grandparent)
+ return ret;
+
+ } while (n);
+
+ return wxNOT_FOUND;
+}
+
+bool wxXmlNode::IsWhitespaceOnly() const
+{
+ return wxIsWhiteOnly(m_content);
+}
+
//-----------------------------------------------------------------------------
m_root = new wxXmlNode(*doc.m_root);
}
-bool wxXmlDocument::Load(const wxString& filename, const wxString& encoding)
+bool wxXmlDocument::Load(const wxString& filename, const wxString& encoding, int flags)
{
wxFileInputStream stream(filename);
if (!stream.Ok())
return false;
- return Load(stream, encoding);
+ return Load(stream, encoding, flags);
}
-bool wxXmlDocument::Save(const wxString& filename) const
+bool wxXmlDocument::Save(const wxString& filename, int indentstep) const
{
wxFileOutputStream stream(filename);
if (!stream.Ok())
return false;
- return Save(stream);
+ return Save(stream, indentstep);
}
// wxXmlDocument loading routines
//-----------------------------------------------------------------------------
-/*
- FIXME:
- - process all elements, including CDATA
- */
-
// converts Expat-produced string in UTF-8 into wxString using the specified
// conv or keep in UTF-8 if conv is NULL
static wxString CharToString(wxMBConv *conv,
#endif // wxUSE_UNICODE/!wxUSE_UNICODE
}
+// returns true if the given string contains only whitespaces
+bool wxIsWhiteOnly(const wxChar *buf)
+{
+ for (const wxChar *c = buf; *c != wxT('\0'); c++)
+ if (*c != wxT(' ') && *c != wxT('\t') && *c != wxT('\n') && *c != wxT('\r'))
+ return false;
+ return true;
+}
+
+
struct wxXmlParsingContext
{
wxMBConv *conv;
wxString encoding;
wxString version;
bool bLastCdata;
+ bool removeWhiteOnlyNodes;
};
extern "C" {
static void TextHnd(void *userData, const char *s, int len)
{
wxXmlParsingContext *ctx = (wxXmlParsingContext*)userData;
- char *buf = new char[len + 1];
-
- buf[len] = '\0';
- memcpy(buf, s, (size_t)len);
+ wxString str = CharToString(ctx->conv, s, len);
if (ctx->lastAsText)
{
if ( ctx->bLastCdata )
{
ctx->lastAsText->SetContent(ctx->lastAsText->GetContent() +
- CharToString(NULL, buf));
+ CharToString(NULL, s, len));
}
else
{
- ctx->lastAsText->SetContent(ctx->lastAsText->GetContent() +
- CharToString(ctx->conv, buf));
+ ctx->lastAsText->SetContent(ctx->lastAsText->GetContent() + str);
}
}
else
{
- bool whiteOnly = true;
- for (char *c = buf; *c != '\0'; c++)
- if (*c != ' ' && *c != '\t' && *c != '\n' && *c != '\r')
- {
- whiteOnly = false;
- break;
- }
+ bool whiteOnly = false;
+ if (ctx->removeWhiteOnlyNodes)
+ whiteOnly = wxIsWhiteOnly(str);
+
if (!whiteOnly)
{
- ctx->lastAsText = new wxXmlNode(wxXML_TEXT_NODE, wxT("text"),
- CharToString(ctx->conv, buf));
+ ctx->lastAsText = new wxXmlNode(wxXML_TEXT_NODE, wxT("text"), str);
ctx->node->AddChild(ctx->lastAsText);
}
}
-
- delete[] buf;
}
}
}
}
-bool wxXmlDocument::Load(wxInputStream& stream, const wxString& encoding)
+bool wxXmlDocument::Load(wxInputStream& stream, const wxString& encoding, int flags)
{
#if wxUSE_UNICODE
(void)encoding;
if ( encoding != wxT("UTF-8") && encoding != wxT("utf-8") )
ctx.conv = new wxCSConv(encoding);
#endif
+ ctx.removeWhiteOnlyNodes = (flags & wxXMLDOC_KEEP_WHITESPACE_NODES) == 0;
ctx.bLastCdata = false;
XML_SetUserData(parser, (void*)&ctx);
}
static void OutputNode(wxOutputStream& stream, wxXmlNode *node, int indent,
- wxMBConv *convMem, wxMBConv *convFile)
+ wxMBConv *convMem, wxMBConv *convFile, int indentstep)
{
wxXmlNode *n, *prev;
wxXmlProperty *prop;
n = node->GetChildren();
while (n)
{
- if (n && n->GetType() != wxXML_TEXT_NODE)
- OutputIndentation(stream, indent + 1);
- OutputNode(stream, n, indent + 1, convMem, convFile);
+ if (indentstep >= 0 && n && n->GetType() != wxXML_TEXT_NODE)
+ OutputIndentation(stream, indent + indentstep);
+ OutputNode(stream, n, indent + indentstep, convMem, convFile, indentstep);
prev = n;
n = n->GetNext();
}
- if (prev && prev->GetType() != wxXML_TEXT_NODE)
+ if (indentstep >= 0 && prev && prev->GetType() != wxXML_TEXT_NODE)
OutputIndentation(stream, indent);
OutputString(stream, wxT("</"));
OutputString(stream, node->GetName());
}
}
-bool wxXmlDocument::Save(wxOutputStream& stream) const
+bool wxXmlDocument::Save(wxOutputStream& stream, int indentstep) const
{
if ( !IsOk() )
return false;
GetVersion().c_str(), GetFileEncoding().c_str());
OutputString(stream, s);
- OutputNode(stream, GetRoot(), 0, convMem, convFile);
+ OutputNode(stream, GetRoot(), 0, convMem, convFile, indentstep);
OutputString(stream, wxT("\n"));
if ( convFile )