New ContentType task.

Fix the problems I encountered while trying to use the new task.

include/dindexer-machinery/scantask/contenttype.hpp

@@ -0,0 +1,48 @@
+/* Copyright 2015, 2016, Michele Santullo
+ * This file is part of "dindexer".
+ *
+ * "dindexer" 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.
+ *
+ * "dindexer" 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 "dindexer".  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef id148DBED10A0B45238E810683656BA7D5
+#define id148DBED10A0B45238E810683656BA7D5
+
+#include "dindexer-machinery/scantask/base.hpp"
+#include "dindexer-machinery/guess_content_type.hpp"
+#include "dindexer-machinery/mediatypes.hpp"
+#include <memory>
+#include <vector>
+
+namespace mchlib {
+	struct FileRecordData;
+
+	namespace scantask {
+		class ContentType : public Base<mchlib::ContentTypes> {
+		public:
+			using DirTreeTaskPtr = std::shared_ptr<Base<std::vector<FileRecordData>>>;
+			using MediaTypeTaskPtr = std::shared_ptr<Base<mchlib::MediaTypes>>;
+
+			ContentType ( DirTreeTaskPtr parDirTree, MediaTypeTaskPtr parMediaType );
+
+		private:
+			virtual void on_data_destroy ( mchlib::ContentTypes& parData ) override;
+			virtual void on_data_create ( mchlib::ContentTypes& parData ) override;
+
+			DirTreeTaskPtr m_dir_tree;
+			MediaTypeTaskPtr m_media_type;
+		};
+	} //namespace scantask
+} //namespace mchlib
+
+#endif

include/dindexer-machinery/scantask/mediatype.hpp

@@ -24,7 +24,7 @@
 
 namespace mchlib {
 	namespace scantask {
-		class MediaType : Base<MediaTypes> {
+		class MediaType : public Base<mchlib::MediaTypes> {
 		public:
 			MediaType ( char parDefault, bool parForce, std::string parSearchPath );
 			virtual ~MediaType ( void ) noexcept = default;

src/machinery/CMakeLists.txt

@@ -19,6 +19,7 @@ add_library(${PROJECT_NAME} SHARED
 	scantask/dirtree.cpp
 	scantask/mediatype.cpp
 	scantask/hashing.cpp
+	scantask/contenttype.cpp
 )
 
 #target_include_directories(${PROJECT_NAME}

src/machinery/guess_content_type.cpp

@@ -155,7 +155,7 @@ namespace mchlib {
 		assert(std::equal(parContent.begin(), parContent.end(), SetListing(std::vector<FileRecordData>(parContent)).sorted_list().begin()));
 
 		//TODO: assert that the first item in the list is the shortest string
-		std::shared_ptr<PathName> pathname(new PathName(parContent.front().abs_path));
+		std::shared_ptr<PathName> pathname(new PathName(""));
 		ConstSetListingView view(parContent.begin(), parContent.end(), pathname->atom_count(), pathname);
 		assert(parContent.size() >= 1);
 		return guess_content_type(parMediaType, view, parContent.size() - 1);

src/machinery/scantask/contenttype.cpp

@@ -0,0 +1,41 @@
+/* Copyright 2015, 2016, Michele Santullo
+ * This file is part of "dindexer".
+ *
+ * "dindexer" 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.
+ *
+ * "dindexer" 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 "dindexer".  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "dindexer-machinery/scantask/contenttype.hpp"
+#include <cassert>
+
+namespace mchlib {
+	namespace scantask {
+		ContentType::ContentType (DirTreeTaskPtr parDirTree, MediaTypeTaskPtr parMediaType) :
+			m_dir_tree(parDirTree),
+			m_media_type(parMediaType)
+		{
+			assert(m_dir_tree);
+			assert(m_media_type);
+		}
+
+		void ContentType::on_data_destroy (mchlib::ContentTypes& parData) {
+			parData = ContentType_Unknown;
+		}
+
+		void ContentType::on_data_create (mchlib::ContentTypes& parData) {
+			auto media_type = m_media_type->get_or_create();
+			const auto& tree = m_dir_tree->get_or_create();
+			parData = mchlib::guess_content_type(media_type, tree);
+		}
+	} //namespace scantask
+} //namespace mchlib

src/machinery/set_listing.cpp

@@ -99,7 +99,7 @@ namespace mchlib {
 		{
 			assert(parBasePath);
 			assert(m_base_path or m_current == m_end);
-			assert(m_current == m_end or m_base_path->atom_count() == PathName(m_current->path).atom_count());
+			assert(m_current == m_end or m_base_path->atom_count() == PathName(m_current->path).atom_count() + parLevelOffset);
 			assert(m_current == m_end or m_base_path->atom_count() == m_current->level + m_level_offset);
 
 			//Look for the point where the children of this entry start

src/scan/main.cpp

@@ -29,6 +29,7 @@
 #include "dindexer-machinery/scantask/dirtree.hpp"
 #include "dindexer-machinery/scantask/mediatype.hpp"
 #include "dindexer-machinery/scantask/hashing.hpp"
+#include "dindexer-machinery/scantask/contenttype.hpp"
 #include <iostream>
 #include <iomanip>
 #include <ciso646>
@@ -78,12 +79,15 @@ int main (int parArgc, char* parArgv[]) {
 
 	const std::string search_path(vm["search-path"].as<std::string>());
 	std::shared_ptr<mchlib::scantask::DirTree> scan_dirtree(new mchlib::scantask::DirTree(search_path));
-	std::shared_ptr<mchlib::scantask::MediaType> media_type(new mchlib::scantask::MediaType((vm.count("type") ? vm["type"].as<char>() : 'O'), not vm.count("type"), search_path));
+	std::shared_ptr<mchlib::scantask::MediaType> media_type(new mchlib::scantask::MediaType((vm.count("type") ? vm["type"].as<char>() : 'O'), vm.count("type"), search_path));
 	std::shared_ptr<mchlib::scantask::Hashing> hashing(new mchlib::scantask::Hashing(scan_dirtree, true));
+	std::shared_ptr<mchlib::scantask::ContentType> content_type(new mchlib::scantask::ContentType(scan_dirtree, media_type));
+
+	std::cout << "Content type: " << mchlib::content_type_to_char(content_type->get_or_create()) << std::endl;
 
 	const auto& hashes = hashing->get_or_create();
 	for (const auto& hash : hashes) {
-		std::cout << '"' << hash.path << "\" -> " << mchlib::tiger_to_string(hash.hash) << std::endl;
+		std::cout << '"' << hash.path << "\" -> " << mchlib::tiger_to_string(hash.hash) << "\n";
 	}
 
 	return 0;