Boost Property Tree and Xml parsing Problems

南笙酒味 提交于 2019-12-30 08:36:08

问题


I'm using boost::property_tree. The documentation is very vague and overall unhelpful for the most part. Looking at the source/examples didn't help that much, either.

What I'm wondering is the following:

<VGHL>
    <StringTable>
        <Language>EN</Language>
        <DataPath>..\\Data\\Resources\\Strings\\stringtable.bst</DataPath>
    </StringTable>
</VGHL>

How can I iterate over all the elements at the current level? If I do this:

read_xml(fin, bifPropTree);
VGHL::String tablePath;
BOOST_FOREACH(boost::property_tree::wiptree::value_type &v, 
              bifPropTree.get_child(L"VGHL.StringTable"))
{
    m_StringTable->ParseEntry(v.second, tablePath);
}

In ParseEntry I try this:

VGHL::String langName = stringTree.get<VGHL::String>(L"StringTable.Language");

Results in an exception (not doesn't exist). I've also tried this:

VGHL::String langName = stringTree.get<VGHL::String>(L"Language");

Same problem.

From my understanding when I call ParseEntry I am passing a reference to the tree at that node.

Is there any way to deal with this, when I have multiple entries of StringTable using property tree?


回答1:


ParseEntry receives a reference to each of the children nodes of the current level. So, you cannot ask the values using the node name, because you already have a child node. The node name is stored in v.first.

You can iterate over all the elements at a given level using get_child to select the level and then BOOST_FOREACH to iterate. Each iterator will be a pair representing the name of the node and the node data:

using boost::property_tree::wiptree;

wiptree &iterationLevel = bifPropTree.get_child(L"VGHL.StringTable");
BOOST_FOREACH(wiptree::value_type &v, iterationLevel)
{   
  wstring name = v.first;
  wstring value = v.second.get<wstring>(L"");
  wcout << L"Name: " << name << L", Value: " << value.c_str() << endl;
}

This code would print:

Name: Language, Value: EN

Name: DataPath, Value: ..\\Data\\Resources\\Strings\\stringtable.bst

If you do not want to iterate, you can select the node level and then look for the nodes using their name:

wiptree &iterationLevel = bifPropTree.get_child(L"VGHL.StringTable");
wstring valueLang = iterationLevel.get<wstring>(L"Language");
wstring valuePath = iterationLevel.get<wstring>(L"DataPath");
wcout << valueLang << endl << valuePath << endl;

This code would print:

EN

..\\Data\\Resources\\Strings\\stringtable.bst




回答2:


I've not used the property tree, but probably will as it looks nifty. A few quick observations though:

Should not the template parameter of get be the same as the return type?

VGHL::String langName = stringTree.get(...);

But this is most likely not a problem here, as this would have resulted in compile-time error.

Not sure if L"VGHL.StringTable.Language" argument works?



来源:https://stackoverflow.com/questions/1794418/boost-property-tree-and-xml-parsing-problems

易学教程内所有资源均来自网络或用户发布的内容,如有违反法律规定的内容欢迎反馈
该文章没有解决你所遇到的问题?点击提问,说说你的问题,让更多的人一起探讨吧!