X-Git-Url: https://git.carlh.net/gitweb/?a=blobdiff_plain;f=src%2Fstl_binary_reader.cc;h=9a7a150fcf53b767b232ab9b7ed2375797649772;hb=1b0d3a25622031bff3c0dce186d5901c27ddfdcf;hp=022ed29a6d53b51018c568acef5f370271351733;hpb=e409a8b24d78684246e6dbf7092e1e1ac75da8ed;p=libsub.git diff --git a/src/stl_binary_reader.cc b/src/stl_binary_reader.cc index 022ed29..9a7a150 100644 --- a/src/stl_binary_reader.cc +++ b/src/stl_binary_reader.cc @@ -1,5 +1,5 @@ /* - Copyright (C) 2014 Carl Hetherington + Copyright (C) 2014-2020 Carl Hetherington This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -17,13 +17,15 @@ */ -#include -#include -#include #include "stl_binary_reader.h" #include "exceptions.h" #include "iso6937.h" +#include "stl_util.h" #include "compose.hpp" +#include +#include +#include +#include using std::map; using std::vector; @@ -34,81 +36,203 @@ using boost::lexical_cast; using boost::algorithm::replace_all; using boost::is_any_of; using boost::locale::conv::utf_to_utf; +using boost::shared_ptr; using namespace sub; -STLBinaryReader::STLBinaryReader (istream& in) - : _buffer (new unsigned char[1024]) +namespace sub { + +class InputReader : public boost::noncopyable { - in.read ((char *) _buffer, 1024); - if (in.gcount() != 1024) { - throw STLError ("Could not read GSI block from binary STL file"); +public: + InputReader () + : _buffer (new unsigned char[1024]) + { + } - code_page_number = atoi (get_string (0, 3).c_str ()); - - string const dfc = get_string (3, 8); - if (dfc == "STL24.01") { - frame_rate = 24; - } else if (dfc == "STL25.01") { - frame_rate = 25; - } else if (dfc == "STL30.01") { - frame_rate = 30; - } else { - throw STLError (String::compose ("Unknown disk format code %1 in binary STL file", dfc)); + virtual ~InputReader () + { + delete[] _buffer; } - display_standard = _tables.display_standard_file_to_enum (get_string (11, 1)); - language_group = _tables.language_group_file_to_enum (get_string (12, 2)); - language = _tables.language_file_to_enum (get_string (14, 2)); - original_programme_title = get_string (16, 32); - original_episode_title = get_string (48, 32); - translated_programme_title = get_string (80, 32); - translated_episode_title = get_string (112, 32); - translator_name = get_string (144, 32); - translator_contact_details = get_string (176, 32); - subtitle_list_reference_code = get_string (208, 16); - creation_date = get_string (224, 6); - revision_date = get_string (230, 6); - revision_number = get_string (236, 2); - - tti_blocks = atoi (get_string (238, 5).c_str ()); - number_of_subtitles = atoi (get_string (243, 5).c_str ()); - subtitle_groups = atoi (get_string (248, 3).c_str ()); - maximum_characters = atoi (get_string (251, 2).c_str ()); - maximum_rows = atoi (get_string (253, 2).c_str ()); - timecode_status = _tables.timecode_status_file_to_enum (get_string (255, 1)); - start_of_programme = get_string (256, 8); - first_in_cue = get_string (256, 8); - disks = atoi (get_string (272, 1).c_str ()); - disk_sequence_number = atoi (get_string (273, 1).c_str ()); - country_of_origin = get_string (274, 3); - publisher = get_string (277, 32); - editor_name = get_string (309, 32); - editor_contact_details = get_string (341, 32); + virtual void read (int size, string what) = 0; - for (int i = 0; i < tti_blocks; ++i) { - - in.read ((char *) _buffer, 128); - if (in.gcount() != 128) { - throw STLError ("Could not read TTI block from binary STL file"); + string get_string (int offset, int length) const + { + string s; + for (int i = 0; i < length; ++i) { + s += _buffer[offset + i]; + } + + return s; + } + + int get_int (int offset, int length) const + { + int v = 0; + for (int i = 0; i < length; ++i) { + v |= _buffer[offset + i] << (8 * i); + } + + return v; + } + + Time get_timecode (int offset, int frame_rate) const + { + return Time::from_hmsf (_buffer[offset], _buffer[offset + 1], _buffer[offset + 2], _buffer[offset + 3], Rational (frame_rate, 1)); + } + +protected: + unsigned char* _buffer; +}; + + +class StreamInputReader : public InputReader +{ +public: + StreamInputReader (istream& in) + : _in (in) + { + + } + + void read (int size, string what) + { + _in.read (reinterpret_cast(_buffer), size); + if (_in.gcount() != size) { + throw STLError (String::compose("Could not read %1 block from binary STL file", what)); } + } + +private: + std::istream& _in; +}; + +class FILEInputReader : public InputReader +{ +public: + FILEInputReader (FILE* in) + : _in (in) + { + + } + + void read (int size, string what) + { + size_t const N = fread (_buffer, 1, size, _in); + if (static_cast(N) != size) { + throw STLError (String::compose("Could not read %1 block from binary STL file", what)); + } + } + +private: + FILE* _in; +}; + +} - if (_tables.comment_file_to_enum (get_int (15, 1)) == COMMENT_YES) { +STLBinaryReader::STLBinaryReader (istream& in) +{ + read (shared_ptr(new StreamInputReader(in))); +} + +STLBinaryReader::STLBinaryReader (FILE* in) +{ + read (shared_ptr(new FILEInputReader(in))); +} + +void STLBinaryReader::read (shared_ptr reader) +{ + reader->read (1024, "GSI"); + + code_page_number = atoi (reader->get_string(0, 3).c_str()); + frame_rate = stl_dfc_to_frame_rate (reader->get_string(3, 8)); + display_standard = _tables.display_standard_file_to_enum (reader->get_string(11, 1)); + language_group = _tables.language_group_file_to_enum (reader->get_string(12, 2)); + language = _tables.language_file_to_enum (reader->get_string(14, 2)); + original_programme_title = reader->get_string(16, 32); + original_episode_title = reader->get_string(48, 32); + translated_programme_title = reader->get_string(80, 32); + translated_episode_title = reader->get_string(112, 32); + translator_name = reader->get_string(144, 32); + translator_contact_details = reader->get_string(176, 32); + subtitle_list_reference_code = reader->get_string(208, 16); + creation_date = reader->get_string(224, 6); + revision_date = reader->get_string(230, 6); + revision_number = reader->get_string(236, 2); + + tti_blocks = atoi (reader->get_string(238, 5).c_str()); + number_of_subtitles = atoi (reader->get_string(243, 5).c_str()); + subtitle_groups = atoi (reader->get_string(248, 3).c_str()); + maximum_characters = atoi (reader->get_string(251, 2).c_str()); + maximum_rows = atoi (reader->get_string(253, 2).c_str()); + + if (maximum_rows == 99) { + /* https://tech.ebu.ch/docs/tech/tech3360.pdf says + "It is recommended that for files with a large MNR value (e.g. '99') the + font size (height) should be defined as ~ 1/15 of the 'Subtitle Safe Area' + and a lineHeight of 120% is used to achieve a row height of ~ 1/12 of the height + of the 'Subtitle Safe Area'. + */ + maximum_rows = 12; + } + + timecode_status = _tables.timecode_status_file_to_enum (reader->get_string(255, 1)); + start_of_programme = reader->get_string(256, 8); + first_in_cue = reader->get_string(264, 8); + disks = atoi (reader->get_string(272, 1).c_str()); + disk_sequence_number = atoi (reader->get_string(273, 1).c_str()); + country_of_origin = reader->get_string(274, 3); + publisher = reader->get_string(277, 32); + editor_name = reader->get_string(309, 32); + editor_contact_details = reader->get_string(341, 32); + + for (int i = 0; i < tti_blocks; ++i) { + + reader->read (128, "TTI"); + + if (_tables.comment_file_to_enum (reader->get_int(15, 1)) == COMMENT_YES) { continue; } - string const whole = get_string (16, 112); + string const whole = reader->get_string(16, 112); /* Split the text up into lines (8Ah is a new line) */ vector lines; split (lines, whole, is_any_of ("\x8a")); + /* Italic / underline specifications can span lines, so we need to track them + outside the lines loop. + */ + bool italic = false; + bool underline = false; + for (size_t i = 0; i < lines.size(); ++i) { RawSubtitle sub; - sub.from.set_frame (get_timecode (5)); - sub.to.set_frame (get_timecode (9)); - sub.vertical_position.line = get_int (13, 1) + i; + sub.from = reader->get_timecode(5, frame_rate); + sub.to = reader->get_timecode(9, frame_rate); + sub.vertical_position.line = reader->get_int(13, 1) + i; + sub.vertical_position.lines = maximum_rows; + sub.vertical_position.reference = TOP_OF_SCREEN; + sub.italic = italic; + sub.underline = underline; + + /* XXX: not sure what to do with JC = 0, "unchanged presentation" */ + int const h = reader->get_int(14, 1); + switch (h) { + case 0: + case 2: + sub.horizontal_position.reference = HORIZONTAL_CENTRE_OF_SCREEN; + break; + case 1: + sub.horizontal_position.reference = LEFT_OF_SCREEN; + break; + case 3: + sub.horizontal_position.reference = RIGHT_OF_SCREEN; + break; + } + /* Loop over characters */ string text; for (size_t j = 0; j < lines[i].size(); ++j) { @@ -128,66 +252,36 @@ STLBinaryReader::STLBinaryReader (istream& in) switch (c) { case 0x80: - sub.italic = true; + italic = true; break; case 0x81: - sub.italic = false; + italic = false; break; case 0x82: - sub.underline = true; + underline = true; break; case 0x83: - sub.underline = false; + underline = false; break; default: text += lines[i][j]; break; } + + sub.italic = italic; + sub.underline = underline; } if (!text.empty ()) { sub.text = utf_to_utf (iso6937_to_utf16 (text.c_str())); _subs.push_back (sub); } - + /* XXX: justification */ } } } -STLBinaryReader::~STLBinaryReader () -{ - delete[] _buffer; -} - -string -STLBinaryReader::get_string (int offset, int length) const -{ - string s; - for (int i = 0; i < length; ++i) { - s += _buffer[offset + i]; - } - - return s; -} - -int -STLBinaryReader::get_int (int offset, int length) const -{ - int v = 0; - for (int i = 0; i < length; ++i) { - v |= _buffer[offset + i] << (8 * i); - } - - return v; -} - -FrameTime -STLBinaryReader::get_timecode (int offset) const -{ - return FrameTime (_buffer[offset], _buffer[offset + 1], _buffer[offset + 2], _buffer[offset + 3]); -} - map STLBinaryReader::metadata () const {