From a java property XML file I want to find each element named entry
(inside the root element properties
). Then put the content of its attribute key
in a std::map<std::string, std::string>
as key and the content of the element (between <entry>
and </entry>
) as value.
So far I am using boost property_tree.
But since I'm unfamiliar with parsing XML documents, I was wondering if there are any pitfalls that I don't see here or whether there's a simpler approach than this.
std::map<std::string, std::string> mMap;
BOOST_FOREACH(boost::property_tree::ptree::value_type &v, pt.get_child("properties"))
{
if (v.first == "entry")
{
std::string sVal = v.second.get_child("<xmlattr>.key").data();
if (sVal.compare("Foo") == 0)
mMap[sVal] = v.second.data();
else if (sVal.compare("Bar") == 0)
mMap[sVal] = v.second.data();
else if(...)
}
}
XML:
<!DOCTYPE properties SYSTEM "http://java.sun.com/dtd/properties.dtd">
<properties>
<entry key="Foo">This is a sentence.</entry>
<entry key="Bar">And another one.</entry>
<entry key=...
</properties>
I'd make it simpler:
Map read_properties(std::string const& fname) {
boost::property_tree::ptree pt;
read_xml(fname, pt);
Map map;
for (auto& p : pt.get_child("properties")) {
auto key = p.second.get<std::string>("<xmlattr>.key");
if (!map.insert({ key, p.second.data() }).second)
throw std::runtime_error("Duplicate key: " + key);
}
return map;
}
If you crave more validation or "XPath" feel, I'd suggest using an XML library.
Other than that, I see not a lot wrong.