On 5/27/21 10:56 AM, Dmitry Timoshkov wrote:
Nikolay Sivov nsivov@codeweavers.com wrote:
On 5/20/21 9:50 AM, Dmitry Timoshkov wrote:
+static char *xmldoc_encoding(IXMLDOMDocument3 *doc) +{
- HRESULT hr;
- IXMLDOMNode *node;
- char *encoding = NULL;
- hr = IXMLDOMDocument3_get_firstChild(doc, &node);
- if (hr == S_OK)
- {
DOMNodeType type;
hr = IXMLDOMNode_get_nodeType(node, &type);
if (hr == S_OK && type == NODE_PROCESSING_INSTRUCTION)
{
IXMLDOMProcessingInstruction *pi;
IXMLDOMNode *item;
IXMLDOMNamedNodeMap *node_map;
hr = IXMLDOMNode_QueryInterface(node, &IID_IXMLDOMProcessingInstruction, (void **)&pi);
if (hr == S_OK)
{
hr = IXMLDOMNode_get_attributes(node, &node_map);
if (hr == S_OK)
{
static const WCHAR encodingW[] = {'e','n','c','o','d','i','n','g',0};
BSTR bstr;
bstr = SysAllocString(encodingW);
hr = IXMLDOMNamedNodeMap_getNamedItem(node_map, bstr, &item);
SysFreeString(bstr);
if (hr == S_OK)
{
VARIANT var;
hr = IXMLDOMNode_get_nodeValue(item, &var);
if (hr == S_OK)
{
if (V_VT(&var) == VT_BSTR)
encoding = (char *)xmlchar_from_wchar(V_BSTR(&var));
VariantClear(&var);
}
}
IXMLDOMNamedNodeMap_Release(node_map);
}
IXMLDOMProcessingInstruction_Release(pi);
}
}
IXMLDOMNode_Release(node);
- }
- return encoding;
+}
Why all this complexity? Isn't it the same as reading current "encoding" value from artificially linked PI node?
Do you mean checking the first document child node type for XML_PI_NODE and directly calling xmlGetProp("encoding")? Or something else?
xmlGetProp() won't get you anything for PI node, as I understand. You could read it from content like in parse_xml_decl() from first patch, but only for attribute that matters.