Introduce InputReader to read data from the file and do low-level parsing.
[libsub.git] / src / stl_binary_reader.cc
1 /*
2     Copyright (C) 2014-2020 Carl Hetherington <cth@carlh.net>
3
4     This program is free software; you can redistribute it and/or modify
5     it under the terms of the GNU General Public License as published by
6     the Free Software Foundation; either version 2 of the License, or
7     (at your option) any later version.
8
9     This program is distributed in the hope that it will be useful,
10     but WITHOUT ANY WARRANTY; without even the implied warranty of
11     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
12     GNU General Public License for more details.
13
14     You should have received a copy of the GNU General Public License
15     along with this program; if not, write to the Free Software
16     Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
17
18 */
19
20 #include "stl_binary_reader.h"
21 #include "exceptions.h"
22 #include "iso6937.h"
23 #include "stl_util.h"
24 #include "compose.hpp"
25 #include <boost/lexical_cast.hpp>
26 #include <boost/algorithm/string.hpp>
27 #include <boost/locale.hpp>
28 #include <iostream>
29
30 using std::map;
31 using std::vector;
32 using std::cout;
33 using std::string;
34 using std::istream;
35 using boost::lexical_cast;
36 using boost::algorithm::replace_all;
37 using boost::is_any_of;
38 using boost::locale::conv::utf_to_utf;
39 using namespace sub;
40
41 class InputReader : public boost::noncopyable
42 {
43 public:
44         InputReader (istream& in)
45                 : _in (in)
46                 , _buffer (new unsigned char[1024])
47         {
48
49         }
50
51         ~InputReader ()
52         {
53                 delete[] _buffer;
54         }
55
56
57         void read (int size, string what)
58         {
59                 _in.read (reinterpret_cast<char *>(_buffer), size);
60                 if (_in.gcount() != size) {
61                         throw STLError (String::compose("Could not read %1 block from binary STL file", what));
62                 }
63         }
64
65         string get_string (int offset, int length) const
66         {
67                 string s;
68                 for (int i = 0; i < length; ++i) {
69                         s += _buffer[offset + i];
70                 }
71
72                 return s;
73         }
74
75         int get_int (int offset, int length) const
76         {
77                 int v = 0;
78                 for (int i = 0; i < length; ++i) {
79                         v |= _buffer[offset + i] << (8 * i);
80                 }
81
82                 return v;
83         }
84
85         Time get_timecode (int offset, int frame_rate) const
86         {
87                 return Time::from_hmsf (_buffer[offset], _buffer[offset + 1], _buffer[offset + 2], _buffer[offset + 3], Rational (frame_rate, 1));
88         }
89
90 private:
91         std::istream& _in;
92         unsigned char* _buffer;
93 };
94
95 STLBinaryReader::STLBinaryReader (istream& in)
96 {
97         InputReader reader (in);
98         reader.read (1024, "GSI");
99
100         code_page_number = atoi (reader.get_string(0, 3).c_str());
101         frame_rate = stl_dfc_to_frame_rate (reader.get_string(3, 8));
102         display_standard = _tables.display_standard_file_to_enum (reader.get_string(11, 1));
103         language_group = _tables.language_group_file_to_enum (reader.get_string(12, 2));
104         language = _tables.language_file_to_enum (reader.get_string(14, 2));
105         original_programme_title = reader.get_string(16, 32);
106         original_episode_title = reader.get_string(48, 32);
107         translated_programme_title = reader.get_string(80, 32);
108         translated_episode_title = reader.get_string(112, 32);
109         translator_name = reader.get_string(144, 32);
110         translator_contact_details = reader.get_string(176, 32);
111         subtitle_list_reference_code = reader.get_string(208, 16);
112         creation_date = reader.get_string(224, 6);
113         revision_date = reader.get_string(230, 6);
114         revision_number = reader.get_string(236, 2);
115
116         tti_blocks = atoi (reader.get_string(238, 5).c_str());
117         number_of_subtitles = atoi (reader.get_string(243, 5).c_str());
118         subtitle_groups = atoi (reader.get_string(248, 3).c_str());
119         maximum_characters = atoi (reader.get_string(251, 2).c_str());
120         maximum_rows = atoi (reader.get_string(253, 2).c_str());
121         timecode_status = _tables.timecode_status_file_to_enum (reader.get_string(255, 1));
122         start_of_programme = reader.get_string(256, 8);
123         first_in_cue = reader.get_string(264, 8);
124         disks = atoi (reader.get_string(272, 1).c_str());
125         disk_sequence_number = atoi (reader.get_string(273, 1).c_str());
126         country_of_origin = reader.get_string(274, 3);
127         publisher = reader.get_string(277, 32);
128         editor_name = reader.get_string(309, 32);
129         editor_contact_details = reader.get_string(341, 32);
130
131         for (int i = 0; i < tti_blocks; ++i) {
132
133                 reader.read (128, "TTI");
134
135                 if (_tables.comment_file_to_enum (reader.get_int(15, 1)) == COMMENT_YES) {
136                         continue;
137                 }
138
139                 string const whole = reader.get_string(16, 112);
140
141                 /* Split the text up into lines (8Ah is a new line) */
142                 vector<string> lines;
143                 split (lines, whole, is_any_of ("\x8a"));
144
145                 /* Italic / underline specifications can span lines, so we need to track them
146                    outside the lines loop.
147                 */
148                 bool italic = false;
149                 bool underline = false;
150
151                 for (size_t i = 0; i < lines.size(); ++i) {
152                         RawSubtitle sub;
153                         sub.from = reader.get_timecode(5, frame_rate);
154                         sub.to = reader.get_timecode(9, frame_rate);
155                         sub.vertical_position.line = reader.get_int(13, 1) + i;
156                         sub.vertical_position.lines = maximum_rows;
157                         sub.vertical_position.reference = TOP_OF_SCREEN;
158                         sub.italic = italic;
159                         sub.underline = underline;
160
161                         /* XXX: not sure what to do with JC = 0, "unchanged presentation" */
162                         int const h = reader.get_int(14, 1);
163                         switch (h) {
164                         case 0:
165                         case 2:
166                                 sub.horizontal_position.reference = HORIZONTAL_CENTRE_OF_SCREEN;
167                                 break;
168                         case 1:
169                                 sub.horizontal_position.reference = LEFT_OF_SCREEN;
170                                 break;
171                         case 3:
172                                 sub.horizontal_position.reference = RIGHT_OF_SCREEN;
173                                 break;
174                         }
175
176                         /* Loop over characters */
177                         string text;
178                         for (size_t j = 0; j < lines[i].size(); ++j) {
179
180                                 unsigned char const c = static_cast<unsigned char> (lines[i][j]);
181
182                                 if (c == 0x8f) {
183                                         /* Unused space i.e. end of line */
184                                         break;
185                                 }
186
187                                 if (c >= 0x80 && c <= 0x83) {
188                                         /* Italic or underline control code */
189                                         sub.text = utf_to_utf<char> (iso6937_to_utf16 (text.c_str()));
190                                         _subs.push_back (sub);
191                                         text.clear ();
192                                 }
193
194                                 switch (c) {
195                                 case 0x80:
196                                         italic = true;
197                                         break;
198                                 case 0x81:
199                                         italic = false;
200                                         break;
201                                 case 0x82:
202                                         underline = true;
203                                         break;
204                                 case 0x83:
205                                         underline = false;
206                                         break;
207                                 default:
208                                         text += lines[i][j];
209                                         break;
210                                 }
211
212                                 sub.italic = italic;
213                                 sub.underline = underline;
214                         }
215
216                         if (!text.empty ()) {
217                                 sub.text = utf_to_utf<char> (iso6937_to_utf16 (text.c_str()));
218                                 _subs.push_back (sub);
219                         }
220
221                         /* XXX: justification */
222                 }
223         }
224 }
225
226 map<string, string>
227 STLBinaryReader::metadata () const
228 {
229         map<string, string> m;
230
231         m["Code page number"] = lexical_cast<string> (code_page_number);
232         m["Frame rate"] = lexical_cast<string> (frame_rate);
233         m["Display standard"] = _tables.display_standard_enum_to_description (display_standard);
234         m["Language group"] = _tables.language_group_enum_to_description (language_group);
235         m["Language"] = _tables.language_enum_to_description (language);
236         m["Original programme title"] = original_programme_title;
237         m["Original episode title"] = original_episode_title;
238         m["Translated programme title"] = translated_programme_title;
239         m["Translated episode title"] = translated_episode_title;
240         m["Translator name"] = translator_name;
241         m["Translator contact details"] = translator_contact_details;
242         m["Subtitle list reference code"] = subtitle_list_reference_code;
243         m["Creation date"] = creation_date;
244         m["Revision date"] = revision_date;
245         m["Revision number"] = revision_number;
246         m["TTI blocks"] = lexical_cast<string> (tti_blocks);
247         m["Number of subtitles"] = lexical_cast<string> (number_of_subtitles);
248         m["Subtitle groups"] = lexical_cast<string> (subtitle_groups);
249         m["Maximum characters"] = lexical_cast<string> (maximum_characters);
250         m["Maximum rows"] = lexical_cast<string> (maximum_rows);
251         m["Timecode status"] = _tables.timecode_status_enum_to_description (timecode_status);
252         m["Start of programme"] = start_of_programme;
253         m["First in cue"] = first_in_cue;
254         m["Disks"] = lexical_cast<string> (disks);
255         m["Disk sequence number"] = lexical_cast<string> (disk_sequence_number);
256         m["Country of origin"] = country_of_origin;
257         m["Publisher"] = publisher;
258         m["Editor name"] = editor_name;
259         m["Editor contact details"] = editor_contact_details;
260
261         return m;
262 }