Move scan's core code into new machinery lib.

include/dindexer-machinery/filesearcher.hpp

@@ -0,0 +1,62 @@
+/* Copyright 2015, 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 idB6D385B7779240308449D6081CB790F1
+#define idB6D385B7779240308449D6081CB790F1
+
+#include "filestats.hpp"
+#include <vector>
+#include <functional>
+#include <boost/utility/string_ref.hpp>
+
+namespace fastf {
+  struct StringWithLength {
+    StringWithLength ( const char* parStr, int parLen ) :
+      str(parStr),
+      len(parLen)
+    {
+    }
+
+    const char* const str;
+    const int len;
+  };
+
+  class FileSearcher {
+  public:
+    typedef std::vector<StringWithLength> ConstCharVecType;
+	typedef std::function<bool(const char*, const FileStats&)> CallbackType;
+
+    explicit FileSearcher ( boost::string_ref parBaseDir );
+    ~FileSearcher ( void ) noexcept;
+
+    void Search ( const ConstCharVecType& parExtensions, const ConstCharVecType& parIgnorePaths );
+
+    void SetFollowSymlinks ( bool parFollow ) noexcept { followSymlinks_ = parFollow; }
+    bool FollowSymlinks ( void ) const noexcept { return followSymlinks_; }
+    void SetRemainInFilesystem ( bool parRemain ) noexcept { remainInFilesystem_ = parRemain; }
+    bool RemainInFilesystem ( void ) const noexcept { return remainInFilesystem_; }
+	void SetCallback ( CallbackType parCallback );
+
+  private:
+  	CallbackType callback_;
+    const std::string baseDir_;
+    bool followSymlinks_;
+    bool remainInFilesystem_;
+  };
+} //namespace fastf
+
+#endif

include/dindexer-machinery/filestats.hpp

@@ -0,0 +1,33 @@
+/* Copyright 2015, 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 id4A7D7AB671954418939FC0BDA19C5B3F
+#define id4A7D7AB671954418939FC0BDA19C5B3F
+
+#include <ctime>
+
+namespace fastf {
+	struct FileStats {
+		int level;
+		std::time_t atime;
+		std::time_t mtime;
+		bool is_dir;
+		bool is_symlink;
+	};
+} //namespace fastf
+
+#endif

include/dindexer-machinery/indexer.hpp

@@ -0,0 +1,77 @@
+/* Copyright 2015, 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 idE555EF56730442C1ADDC7B2AE7A9340E
+#define idE555EF56730442C1ADDC7B2AE7A9340E
+
+#include <memory>
+#include <string>
+#include <vector>
+
+#if !defined(NDEBUG)
+#	define INDEXER_VERBOSE
+#endif
+
+#if defined(WITH_PROGRESS_FEEDBACK)
+namespace std {
+	class condition_variable;
+} //namespace std
+#endif
+
+namespace fastf {
+	struct FileStats;
+} //namespace fastf
+
+namespace dinlib {
+	struct Settings;
+} //namespace dinlib
+
+namespace din {
+	struct FileRecordData;
+
+	class Indexer {
+	public:
+		Indexer ( void );
+		Indexer ( Indexer&& ) = default;
+		Indexer ( const Indexer& ) = delete;
+		~Indexer ( void ) noexcept;
+
+		bool add_path ( const char* parPath, const fastf::FileStats& parStats );
+#if defined(INDEXER_VERBOSE)
+		void dump ( void ) const;
+#endif
+
+		std::size_t total_items ( void ) const;
+		std::string operator[] ( std::size_t parIndex ) const;
+#if defined(WITH_PROGRESS_FEEDBACK)
+		std::size_t processed_items ( void ) const;
+		std::string current_item ( void ) const;
+		std::condition_variable& step_notify ( void );
+#endif
+		void calculate_hash ( void );
+		bool empty ( void ) const;
+		void ignore_read_errors ( bool parIgnore );
+		const std::vector<FileRecordData>& record_data ( void ) const;
+
+	private:
+		struct LocalData;
+
+		std::unique_ptr<LocalData> m_local_data;
+	};
+} //namespace din
+
+#endif

include/dindexer-machinery/mediatype.hpp

@@ -0,0 +1,43 @@
+/* Copyright 2015, 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 id66D41389BC59433CA58E325395A6197B
+#define id66D41389BC59433CA58E325395A6197B
+
+#include <string>
+#include <stdexcept>
+#include "dindexer-common/mediatypes.hpp"
+
+namespace din {
+#if defined(WITH_MEDIA_AUTODETECT)
+	dinlib::MediaTypes guess_media_type ( std::string&& parPath );
+
+	class UnknownMediaTypeException : std::runtime_error {
+	public:
+		UnknownMediaTypeException ( const std::string& parWhat );
+		UnknownMediaTypeException ( const char* parWhat );
+	};
+
+	class CantAutodetectException : std::runtime_error {
+	public:
+		CantAutodetectException ( const std::string& parWhat );
+		CantAutodetectException ( const char* parWhat );
+	};
+#endif
+} //namespace din
+
+#endif

include/dindexer-machinery/recorddata.hpp

@@ -0,0 +1,86 @@
+/* Copyright 2015, 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 id3CD7F105AC314540A864487E981E5A7E
+#define id3CD7F105AC314540A864487E981E5A7E
+
+#include "tiger.hpp"
+#include <string>
+#include <boost/utility/string_ref.hpp>
+#include <cstdint>
+#include <ctime>
+#include <boost/flyweight.hpp>
+#include <boost/flyweight/no_locking.hpp>
+
+namespace din {
+	struct FileRecordData {
+		struct MimeStringTagStruct { };
+		typedef boost::flyweights::tag<MimeStringTagStruct> MimeStringTag;
+		typedef boost::flyweight<std::string, boost::flyweights::no_locking, MimeStringTag> mime_string;
+
+		FileRecordData ( void ) = default;
+		FileRecordData ( const char* parPath, std::time_t parATime, std::time_t parMTime, uint64_t parLevel, bool parIsDir, bool parIsSymLink ) :
+			hash {},
+			path(parPath),
+			mime_full(),
+			atime(parATime),
+			mtime(parMTime),
+			mime_type(),
+			mime_charset(),
+			size(0),
+			level(parLevel),
+			is_directory(parIsDir),
+			is_symlink(parIsSymLink),
+			unreadable(false),
+			hash_valid(false)
+		{
+		}
+
+		FileRecordData ( const FileRecordData& ) = delete;
+		FileRecordData ( FileRecordData&& ) = default;
+		FileRecordData& operator= ( const FileRecordData& ) = delete;
+		FileRecordData& operator= ( FileRecordData&& ) = default;
+		bool operator== ( const FileRecordData& ) const = delete;
+
+		TigerHash hash;
+		std::string path;
+		mime_string mime_full;
+		std::time_t atime;
+		std::time_t mtime;
+		boost::string_ref mime_type;
+		boost::string_ref mime_charset;
+		uint64_t size;
+		uint16_t level;
+		bool is_directory;
+		bool is_symlink;
+		bool unreadable;
+		bool hash_valid;
+	};
+
+	struct SetRecordDataFull {
+		std::string name;
+		uint32_t disk_number;
+		char type;
+	};
+
+	struct SetRecordData {
+		const boost::string_ref name;
+		const char type;
+	};
+} //namespace din
+
+#endif

include/dindexer-machinery/tiger.hpp

@@ -0,0 +1,49 @@
+/* Copyright 2015, 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 idBE93AF97FA4343ECA2BC8FB1FD3E5E60
+#define idBE93AF97FA4343ECA2BC8FB1FD3E5E60
+
+#include <cstdint>
+#include <string>
+#include <vector>
+
+namespace din {
+	struct TigerHash {
+		TigerHash ( void ) = default;
+
+		union {
+			struct {
+				uint64_t part_a;
+				uint64_t part_b;
+				uint64_t part_c;
+			};
+			uint64_t data[3];
+			uint8_t byte_data[sizeof(uint64_t) * 3];
+		};
+	};
+
+	static_assert(sizeof(TigerHash) == 24, "Wrong struct size");
+
+	void tiger_file ( const std::string& parPath, TigerHash& parHashFile, TigerHash& parHashDir, uint64_t& parSizeOut );
+	void tiger_init_hash ( TigerHash& parHash );
+	std::string tiger_to_string ( const TigerHash& parHash, bool parUpcase=false );
+	void tiger_data ( const std::string& parData, TigerHash& parHash );
+	void tiger_data ( const std::vector<char>& parData, TigerHash& parHash );
+} //namespace din
+
+#endif