Disk ARchive  2.4.10
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups
escape.hpp
Go to the documentation of this file.
1 /*********************************************************************/
2 // dar - disk archive - a backup/restoration program
3 // Copyright (C) 2002-2052 Denis Corbin
4 //
5 // This program is free software; you can redistribute it and/or
6 // modify it under the terms of the GNU General Public License
7 // as published by the Free Software Foundation; either version 2
8 // of the License, or (at your option) any later version.
9 //
10 // This program is distributed in the hope that it will be useful,
11 // but WITHOUT ANY WARRANTY; without even the implied warranty of
12 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 // GNU General Public License for more details.
14 //
15 // You should have received a copy of the GNU General Public License
16 // along with this program; if not, write to the Free Software
17 // Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
18 //
19 // to contact the author : http://dar.linux.free.fr/email.html
20 /*********************************************************************/
21 
37 
38 
39 #ifndef ESCAPE_HPP
40 #define ESCAPE_HPP
41 
42 #include "../my_config.h"
43 
44 extern "C"
45 {
46 #if HAVE_LIMITS_H
47 #include <limits.h>
48 #endif
49 }
50 
51 #include <set>
52 
53 #include "generic_file.hpp"
54 
55 #define ESCAPE_FIXED_SEQUENCE_NORMAL 0xAD
56 #define ESCAPE_FIXED_SEQUENCE_SPARSE_FILE 0xAE
57 
58 #define MAX_BUFFER_SIZE 204800
59 #ifdef SSIZE_MAX
60 #if SSIZE_MAX < MAX_BUFFER_SIZE
61 #undef MAX_BUFFER_SIZE
62 #define MAX_BUFFER_SIZE SSIZE_MAX
63 #endif
64 #endif
65 
66 namespace libdar
67 {
68 
71 
72  class escape : public generic_file
73  {
74  public:
75  enum sequence_type
76  {
77  seqt_undefined, //< not enough data to define the type of the escape sequence
78  seqt_not_a_sequence, //< to escape data corresponding to an escape sequence's fixed byte sequence
79  seqt_file, //< placed before inode information, eventually followed by file data
80  seqt_ea, //< placed before EA data
81  seqt_catalogue, //< placed before the archive's internal catalogue
82  seqt_data_name, //< placed before the archive data_name (at the beginning of the archive)
83  seqt_file_crc, //< placed before the CRC of file's data
84  seqt_ea_crc, //< placed before the CRC of file's EA
85  seqt_changed, //< placed before new copy of file's data if file's data changed while reading it for backup
86  seqt_dirty, //< placed after data CRC if file is dirty
87  seqt_failed_backup //< placed after inode information if the file could not be openned at backup time
88  };
89 
90  // the archive layout of marks is :
91  // ... <seqt_file> <inode> [<file data> [<seqt_changed> <new copy of data> [...] ] <seqt_file_crc> <CRC>[<seqt_dirty>]] [<seqt_ea> <EA> <sqt_ea_crc> <CRC>] ...
92  // this previous sequence that we will call <SEQ> is repeated for each file, then on the overall archive we have :
93  // <seqt_data_name> <data_name> <SEQ> ... <SEQ> <seqt_catalogue> <catalogue> <terminator>
94 
95  // the provided "below" object must exist during the whole live of the escape object, the escape object does not own this "below" object
96  // it must be destroyed by the caller/creator of the escape object.
97 
98 
99  // constructors & destructors
100 
101  escape(generic_file *below, //< "Below" is the generic file that holds the escaped data
102  const std::set<sequence_type> & x_unjumpable); //< a set of marks that can never been jumped over when skipping for the next mark of a any given type.
103  escape(const escape & ref) : generic_file(ref) { copy_from(ref); };
104  const escape & operator = (const escape & ref);
105  ~escape();
106 
107  // escape specific routines
108 
109  void add_mark_at_current_position(sequence_type t);
110 
112 
117  bool skip_to_next_mark(sequence_type t, bool jump);
118  bool next_to_read_is_mark(sequence_type t);
119  bool next_to_read_is_which_mark(sequence_type & t);
120 
121  void add_unjumpable_mark(sequence_type t) { if(is_terminated()) throw SRC_BUG; unjumpable.insert(t); };
122  void remove_unjumpable_mark(sequence_type t);
123  bool is_unjumpable_mark(sequence_type t) const { return unjumpable.find(t) != unjumpable.end(); };
124  void clear_all_unjumpable_marks() { unjumpable.clear(); };
125 
126  // generic_file inherited routines
127  // NOTA: Nothing is done to prevent skip* operation to put the read cursor in the middle of an escape sequence and
128  // thus incorrectly consider it as normal data. Such event should only occure upon archive corruption and will be detected
129  // by checksum mechanisms.
130 
131  bool skip(const infinint & pos);
132  bool skip_to_eof();
133  bool skip_relative(S_I x);
134  infinint get_position();
135 
136  generic_file *get_below() const { return x_below; };
137 
138  protected:
139  U_I inherited_read(char *a, U_I size);
140  void inherited_write(const char *a, U_I size);
141  void inherited_sync_write() { flush_write(); };
142  void inherited_terminate() { flush_or_clean(); };
143 
144  void change_fixed_escape_sequence(unsigned char value) { fixed_sequence[0] = value; };
145  bool has_escaped_data_since_last_skip() const { return escaped_data_count_since_last_skip > 0; };
146 
147  private:
148 
149  //-- constants
150 
152  static const U_I ESCAPE_SEQUENCE_LENGTH = 6;
153  static const U_I WRITE_BUFFER_SIZE = 2*ESCAPE_SEQUENCE_LENGTH;
154  static const U_I READ_BUFFER_SIZE = MAX_BUFFER_SIZE;
155 
157 
163  static const unsigned char usual_fixed_sequence[ESCAPE_SEQUENCE_LENGTH];
164 
165  //-- variables
166 
167  generic_file *x_below; //< the generic_file in which we read/write escaped data from/to the object is not owned by "this"
168  U_I write_buffer_size; //< amount of data in write transit not yet written to "below" (may have to be escaped)
169  char write_buffer[WRITE_BUFFER_SIZE]; //< data in write transit, all data is unescaped, up to the first real mark, after it, data is raw (may be escaped)
170  //< the first real mark is pointed to by escape_seq_offset_in_buffer
171  U_I read_buffer_size; //< amount of data in write transit, read from below, but not yet unescaped and returned to the upper layer
172  U_I already_read; //< data in buffer that has already returned to the upper layer
173  bool read_eof; //< whether we reached a escape sequence while reading data
174  U_I escape_seq_offset_in_buffer; //< location of the first escape sequence which is not a data sequence
175  char read_buffer[READ_BUFFER_SIZE]; //< data in read transit
176  std::set<sequence_type> unjumpable; //< list of mark that cannot be jumped over when searching for the next mark
177  unsigned char fixed_sequence[ESCAPE_SEQUENCE_LENGTH]; // the preambule of an escape sequence to use/search for
178  infinint escaped_data_count_since_last_skip;
179 
180  //-- routines
181 
182  void set_fixed_sequence_for(sequence_type t) { fixed_sequence[ESCAPE_SEQUENCE_LENGTH - 1] = type2char(t); };
183  void check_below() { if(x_below == NULL) throw SRC_BUG; };
184  void clean_data() { read_buffer_size = already_read = escape_seq_offset_in_buffer = 0; }; //< drops all in-transit data
185  void flush_write(); //< write down to "below" all in-transit data
186  void flush_or_clean()
187  {
188  switch(get_mode())
189  {
190  case gf_read_only:
191  clean_data();
192  break;
193  case gf_write_only:
194  flush_write();
195  break;
196  default:
197  throw SRC_BUG;
198  }
199  };
200  void copy_from(const escape & ref);
201  bool mini_read_buffer(); //< returns true if it could end having at least ESCAPE_SEQUENCE_LENGTH bytes in read_buffer, false else (EOF reached).
202 
203  //-- static routine(s)
204 
205  // some convertion routines
206  static char type2char(sequence_type x);
207  static sequence_type char2type(char x);
208 
210 
212 
215  static U_I trouve_amorce(const char *a, U_I size, const unsigned char escape_sequence[ESCAPE_SEQUENCE_LENGTH]);
216 
218 
225  static U_I remove_data_marks_and_stop_at_first_real_mark(char *a, U_I size, U_I & delta, const unsigned char escape_sequence[ESCAPE_SEQUENCE_LENGTH]);
226  };
227 
229 
230 } // end of namespace
231 
232 #endif
write only access
class generic_file is defined here as well as class fichierthe generic_file interface is widely used ...
bool is_terminated() const
generic_file(gf_mode m)
main constructor
gf_mode get_mode() const
retreive the openning mode for this object
read only access