Use XQilla and Xerces-c from the system instead of pugixml.

I don't think this commit works or even compiles, I have too many
changes and I have to start committing from somewhere. At the same
time I don't want to make a "lots of changes here and there" kind
of commit.

.gitignore

@@ -1,2 +1,3 @@
 build/
 tags
+compile_commands.json

CMakeLists.txt

@@ -5,8 +5,8 @@ project(duckscraper VERSION 0.2.1 LANGUAGES CXX)
 option(BUILD_SHARED_TIDY "Wheter you want to build tidy-html5 as a shared library" OFF)
 
 include(GetGitRevisionDescription)
-find_package(PugiXML REQUIRED)
 find_package(Boost 1.32.0 COMPONENTS program_options)
+find_package(XQilla 2.3.3 REQUIRED)
 
 set(CMAKE_CXX_STANDARD 17)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)
@@ -59,6 +59,7 @@ target_link_libraries(${PROJECT_NAME}
 	PRIVATE curlcpp
 	PRIVATE ${Boost_LIBRARIES}
 	PRIVATE mstch
+	PRIVATE XQilla::XQilla
 )
 
 target_compile_definitions(${PROJECT_NAME}

cmake/Modules/FindXQilla.cmake

@@ -0,0 +1,28 @@
+# Find the XQilla library
+# originally taken from
+# https://github.com/rug-compling/alpinocorpus/blob/master/cmake/FindXQilla.cmake
+
+find_path(XQILLA_INCLUDE_DIR NAMES xqilla/xqilla-simple.hpp)
+find_library(XQILLA_LIBRARY NAMES xqilla)
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(
+	XQILLA
+	DEFAULT_MSG
+	XQILLA_INCLUDE_DIR
+	XQILLA_LIBRARY
+)
+set(XQILLA_LIBRARIES ${XQILLA_LIBRARY})
+mark_as_advanced(XQILLA_INCLUDE_DIR XQILLA_LIBRARY)
+
+if (XQILLA_FOUND)
+	find_package(XercesC REQUIRED)
+
+	if (NOT TARGET XQilla::XQilla)
+		add_library(XQilla::XQilla UNKNOWN IMPORTED)
+		set_target_properties(XQilla::XQilla PROPERTIES
+			INTERFACE_INCLUDE_DIRECTORIES "${XQILLA_INCLUDE_DIR}"
+			IMPORTED_LOCATION "${XQILLA_LIBRARY}"
+			INTERFACE_LINK_LIBRARIES XercesC::XercesC
+		)
+	endif()
+endif()

src/safe_stack_object.hpp

@@ -0,0 +1,104 @@
+/*
+	Copyright 2016, 2017 Michele "King_DuckZ" Santullo
+
+	This file is part of MyCurry.
+
+	MyCurry is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	MyCurry is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with MyCurry.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#pragma once
+
+#include "kakoune/safe_ptr.hh"
+#include <utility>
+
+namespace curry {
+	template <typename T>
+	class SafeStackObject {
+	public:
+		typedef Kakoune::SafePtr<T> safe_ptr;
+
+		SafeStackObject();
+		SafeStackObject (SafeStackObject&& parOther);
+		SafeStackObject (const SafeStackObject& parOther) = delete;
+		template <typename... Args> explicit SafeStackObject (Args&&... parArgs);
+		~SafeStackObject() noexcept = default;
+
+		SafeStackObject& operator= (SafeStackObject&& parOther) = delete;
+		SafeStackObject& operator= (const SafeStackObject& parOther) = delete;
+
+		operator Kakoune::SafePtr<T>&();
+		template <typename U>
+		operator Kakoune::SafePtr<U>();
+		T& operator*();
+		safe_ptr& operator->();
+
+	private:
+		T m_obj;
+		safe_ptr m_obj_ptr;
+	};
+
+	template <typename T>
+	SafeStackObject<T>::SafeStackObject() :
+		m_obj(),
+		m_obj_ptr(&m_obj)
+	{
+	}
+
+	template <typename T>
+	SafeStackObject<T>::SafeStackObject (SafeStackObject&& parOther) :
+		m_obj(std::move(parOther.m_obj)),
+		m_obj_ptr(&m_obj)
+	{
+	}
+
+	template <typename T>
+	template <typename... Args>
+	SafeStackObject<T>::SafeStackObject (Args&&... parArgs) :
+		m_obj(std::forward<Args>(parArgs)...),
+		m_obj_ptr(&m_obj)
+	{
+	}
+
+	//template <typename T>
+	//SafeStackObject& SafeStackObject<T>::operator= (SafeStackObject&& parOther) {
+	//	m_obj = std::move(parOther.m_obj);
+	//	m_obj_ptr = std::move(parOther.m_obj_ptr);
+	//	m_ob
+	//}
+
+	//template <typename T>
+	//SafeStackObject& SafeStackObject<T>::operator= (const SafeStackObject& parOther) {
+	//}
+
+	template <typename T>
+	SafeStackObject<T>::operator Kakoune::SafePtr<T>&() {
+		return m_obj_ptr;
+	}
+
+	template <typename T>
+	template <typename U>
+	SafeStackObject<T>::operator Kakoune::SafePtr<U>() {
+		return Kakoune::SafePtr<U>(&m_obj);
+	}
+
+	template <typename T>
+	T& SafeStackObject<T>::operator*() {
+		return *m_obj_ptr;
+	}
+
+	template <typename T>
+	auto SafeStackObject<T>::operator->() -> safe_ptr& {
+		return m_obj_ptr;
+	}
+} //namespace curry

src/xpath.cpp

@@ -17,7 +17,9 @@
  */
 
 #include "xpath.hpp"
-#include <pugixml.hpp>
+#include <xercesc/framework/MemBufInputSource.hpp>
+#include <xercesc/util/XMLString.hpp>
+#include <xqilla/exceptions/XMLParseException.hpp>
 #include <sstream>
 #include <stdexcept>
 #include <algorithm>
@@ -46,30 +48,30 @@ namespace duck {
 	} //unnamed namespace
 
 	XPathBatchResults xpath_query (const std::string& parXML, const std::vector<std::string>& parQueries) {
-		pugi::xml_document doc;
+		XQilla& xqilla = m_xqilla;
-		std::istringstream iss(parXML);
+		XercesConfiguration xconfig;
-		pugi::xml_parse_result result(doc.load(iss));
+		AutoDelete<DynamicContext> context(xqilla.createContext(XQilla::XQUERY_UPDATE, &xconfig));
-		if (not result) {
+		xercesc::MemBufInputSource input_buf(reinterpret_cast<const XMLByte*>(parXML.c_str()), parXML.size(), "n/a", false);
-			auto line_col = line_col_from_offset(result.offset, parXML);
+		Node::Ptr ptr;
-			throw ParseError(line_col.first, line_col.second, result.description());
+		try {
+			ptr = context->parseDocument(input_buf);
 		}
+		catch (const XMLParseException& err) {
+			throw ParseError(err.getXQueryLine(), err.getXQueryColumn(), xercesc::XMLString::transcode(err.getError()));
+		}
+		context->setContextItem(ptr);
 
 		XPathBatchResults retval;
 		for (const auto& xpath : parQueries) {
-			pugi::xpath_node_set xpathRes = doc.select_nodes(xpath.c_str());
+			AutoDelete<XQQuery> query(xqilla.parse(X(xpath.c_str())));
+			context->setContextPosition(1);
+			context->setContextSize(1);
+
+			Result result = query->execute(context);
+			Item::Ptr item;
 			std::vector<std::pair<std::string, std::string>> new_lst;
-			for (pugi::xpath_node_set::const_iterator itFind(xpathRes.begin()), itFindEND(xpathRes.end()); itFind != itFindEND; ++itFind) {
+			while(nullptr != (item = result->next(context))) {
-				const pugi::xpath_node& node = *itFind;
+				new_lst.push_back(std::make_pair(std::string(), UTF8(item->asString(context))));
-				std::pair<std::string, std::string> new_itm;
-				if (node.node()) {
-					new_itm.first = std::string(node.node().name());
-					new_itm.second = std::string(node.node().value());
-				}
-				else if (node.attribute()) {
-					new_itm.first = std::string(node.attribute().name());
-					new_itm.second = std::string(node.attribute().value());
-				}
-				new_lst.push_back(std::move(new_itm));
 			}
 			retval.push_back(std::move(new_lst));
 		}

src/xpath.hpp

@@ -23,6 +23,7 @@
 #include <vector>
 #include <exception>
 #include <utility>
+#include <xqilla/xqilla-simple.hpp>
 
 namespace duck {
 	typedef std::vector<std::vector<std::pair<std::string, std::string>>> XPathBatchResults;