/* Copyright (C) 2012-2021 Carl Hetherington This file is part of libdcp. libdcp 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 2 of the License, or (at your option) any later version. libdcp 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 libdcp. If not, see . In addition, as a special exception, the copyright holders give permission to link the code of portions of this program with the OpenSSL library under certain conditions as described in each individual source file, and distribute linked combinations including the two. You must obey the GNU General Public License in all respects for all of the code used other than OpenSSL. If you modify file(s) with this exception, you may extend this exception to your version of the file(s), but you are not obligated to do so. If you do not wish to do so, delete this exception statement from your version. If you delete this exception statement from all source files in the program, then also delete it here. */ /** @file src/smpte_text_asset.cc * @brief SMPTETextAsset class */ #include "compose.hpp" #include "crypto_context.h" #include "dcp_assert.h" #include "equality_options.h" #include "exceptions.h" #include "filesystem.h" #include "raw_convert.h" #include "smpte_load_font_node.h" #include "smpte_text_asset.h" #include "text_image.h" #include "util.h" #include "warnings.h" #include "xml.h" LIBDCP_DISABLE_WARNINGS #include #include #include #include LIBDCP_ENABLE_WARNINGS #include using std::string; using std::list; using std::vector; using std::map; using std::shared_ptr; using std::dynamic_pointer_cast; using std::make_shared; using boost::split; using boost::is_any_of; using boost::shared_array; using boost::optional; using boost::starts_with; using namespace dcp; static string const subtitle_smpte_ns_2007 = "http://www.smpte-ra.org/schemas/428-7/2007/DCST"; static string const subtitle_smpte_ns_2010 = "http://www.smpte-ra.org/schemas/428-7/2010/DCST"; static string const subtitle_smpte_ns_2014 = "http://www.smpte-ra.org/schemas/428-7/2014/DCST"; SMPTETextAsset::SMPTETextAsset(SubtitleStandard standard) : MXF(Standard::SMPTE) , _edit_rate (24, 1) , _time_code_rate (24) , _subtitle_standard(standard) , _xml_id (make_uuid()) { } SMPTETextAsset::SMPTETextAsset(boost::filesystem::path file) : TextAsset (file) { auto xml = make_shared("SubtitleReel"); Kumu::FileReaderFactory factory; auto reader = make_shared(factory); auto r = Kumu::RESULT_OK; { ASDCPErrorSuspender sus; r = reader->OpenRead(dcp::filesystem::fix_long_path(*_file).string().c_str()); } if (!ASDCP_FAILURE(r)) { /* MXF-wrapped */ ASDCP::WriterInfo info; reader->FillWriterInfo (info); _id = read_writer_info (info); if (!_key_id) { /* Not encrypted; read it in now */ string xml_string; reader->ReadTimedTextResource (xml_string); _raw_xml = xml_string; xml->read_string (xml_string); parse_xml (xml); read_mxf_descriptor (reader); read_mxf_resources(reader, std::make_shared(optional(), Standard::SMPTE)); } else { read_mxf_descriptor (reader); } } else { /* Plain XML */ try { _raw_xml = dcp::file_to_string (file); xml = make_shared("SubtitleReel"); xml->read_file(dcp::filesystem::fix_long_path(file)); parse_xml (xml); } catch (cxml::Error& e) { boost::throw_exception ( ReadError ( String::compose ( "Failed to read subtitle file %1; MXF failed with %2, XML failed with %3", file, static_cast(r), e.what() ) ) ); } /* Try to read PNG files from the same folder that the XML is in; the wisdom of this is debatable, at best... */ for (auto i: _texts) { auto im = dynamic_pointer_cast(i); if (im && im->png_image().size() == 0) { /* Even more dubious; allow .png or urn:uuid:.png */ auto p = file.parent_path() / String::compose("%1.png", im->id()); if (filesystem::is_regular_file(p)) { im->read_png_file (p); } else if (starts_with (im->id(), "urn:uuid:")) { p = file.parent_path() / String::compose("%1.png", remove_urn_uuid(im->id())); if (filesystem::is_regular_file(p)) { im->read_png_file (p); } } } } _standard = Standard::SMPTE; } /* Check that all required image data have been found */ for (auto i: _texts) { auto im = dynamic_pointer_cast(i); if (im && im->png_image().size() == 0) { throw MissingTextImageError (im->id()); } } } void SMPTETextAsset::parse_xml(shared_ptr xml) { if (xml->namespace_uri() == subtitle_smpte_ns_2007) { _subtitle_standard = SubtitleStandard::SMPTE_2007; } else if (xml->namespace_uri() == subtitle_smpte_ns_2010) { _subtitle_standard = SubtitleStandard::SMPTE_2010; } else if (xml->namespace_uri() == subtitle_smpte_ns_2014) { _subtitle_standard = SubtitleStandard::SMPTE_2014; } else { throw XMLError("Unrecognised subtitle namespace " + xml->namespace_uri()); } _xml_id = remove_urn_uuid(xml->string_child("Id")); _load_font_nodes = type_children (xml, "LoadFont"); _content_title_text = xml->string_child ("ContentTitleText"); _annotation_text = xml->optional_string_child ("AnnotationText"); _issue_date = LocalTime (xml->string_child ("IssueDate")); _reel_number = xml->optional_number_child ("ReelNumber"); _language = xml->optional_string_child ("Language"); /* This is supposed to be two numbers, but a single number has been seen in the wild */ auto const er = xml->string_child ("EditRate"); vector er_parts; split (er_parts, er, is_any_of (" ")); if (er_parts.size() == 1) { _edit_rate = Fraction (raw_convert (er_parts[0]), 1); } else if (er_parts.size() == 2) { _edit_rate = Fraction (raw_convert (er_parts[0]), raw_convert (er_parts[1])); } else { throw XMLError ("malformed EditRate " + er); } _time_code_rate = xml->number_child ("TimeCodeRate"); if (xml->optional_string_child ("StartTime")) { _start_time = Time (xml->string_child("StartTime"), _time_code_rate); } /* Now we need to drop down to xmlpp */ vector ps; for (auto i: xml->node()->get_children()) { auto const e = dynamic_cast(i); if (e && e->get_name() == "SubtitleList") { parse_texts(e, ps, _time_code_rate, Standard::SMPTE); } } /* Guess intrinsic duration */ _intrinsic_duration = latest_text_out().as_editable_units_ceil(_edit_rate.numerator / _edit_rate.denominator); } void SMPTETextAsset::read_mxf_resources(shared_ptr reader, shared_ptr dec) { ASDCP::TimedText::TimedTextDescriptor descriptor; reader->FillTimedTextDescriptor (descriptor); /* Load fonts and images */ for ( auto i = descriptor.ResourceList.begin(); i != descriptor.ResourceList.end(); ++i) { ASDCP::TimedText::FrameBuffer buffer; buffer.Capacity(32 * 1024 * 1024); auto const result = reader->ReadAncillaryResource(i->ResourceID, buffer, dec->context(), dec->hmac()); if (ASDCP_FAILURE(result)) { switch (i->Type) { case ASDCP::TimedText::MT_OPENTYPE: throw ReadError(String::compose("Could not read font from MXF file (%1)", static_cast(result))); case ASDCP::TimedText::MT_PNG: throw ReadError(String::compose("Could not read subtitle image from MXF file (%1)", static_cast(result))); default: throw ReadError(String::compose("Could not read resource from MXF file (%1)", static_cast(result))); } } char id[64]; Kumu::bin2UUIDhex (i->ResourceID, ASDCP::UUIDlen, id, sizeof(id)); switch (i->Type) { case ASDCP::TimedText::MT_OPENTYPE: { auto j = _load_font_nodes.begin(); while (j != _load_font_nodes.end() && (*j)->urn != id) { ++j; } if (j != _load_font_nodes.end ()) { _fonts.push_back(Font((*j)->id, (*j)->urn, ArrayData(buffer.RoData(), buffer.Size()))); } break; } case ASDCP::TimedText::MT_PNG: { auto j = _texts.begin(); while (j != _texts.end() && ((!dynamic_pointer_cast(*j)) || dynamic_pointer_cast(*j)->id() != id)) { ++j; } if (j != _texts.end()) { dynamic_pointer_cast(*j)->set_png_image(ArrayData(buffer.RoData(), buffer.Size())); } break; } default: break; } } } void SMPTETextAsset::read_mxf_descriptor(shared_ptr reader) { ASDCP::TimedText::TimedTextDescriptor descriptor; reader->FillTimedTextDescriptor (descriptor); _intrinsic_duration = descriptor.ContainerDuration; /* The thing which is called AssetID in the descriptor is also known as the * ResourceID of the MXF. We store that, at present just for verification * purposes. */ char id[64]; Kumu::bin2UUIDhex (descriptor.AssetID, ASDCP::UUIDlen, id, sizeof(id)); _resource_id = id; } void SMPTETextAsset::set_key(Key key) { /* See if we already have a key; if we do, and we have a file, we'll already have read that file. */ auto const had_key = static_cast(_key); auto const had_key_id = static_cast(_key_id); MXF::set_key (key); if (!had_key_id || !_file || had_key) { /* Either we don't have any data to read, it wasn't encrypted, or we've already read it, so we don't need to do anything else. */ return; } /* Our data was encrypted; now we can decrypt it */ Kumu::FileReaderFactory factory; auto reader = make_shared(factory); auto r = reader->OpenRead(dcp::filesystem::fix_long_path(*_file).string().c_str()); if (ASDCP_FAILURE (r)) { boost::throw_exception ( ReadError ( String::compose ("Could not read encrypted subtitle MXF (%1)", static_cast (r)) ) ); } auto dec = make_shared(key, Standard::SMPTE); string xml_string; reader->ReadTimedTextResource (xml_string, dec->context(), dec->hmac()); _raw_xml = xml_string; auto xml = make_shared("SubtitleReel"); xml->read_string (xml_string); parse_xml (xml); read_mxf_descriptor(reader); read_mxf_resources (reader, dec); } vector> SMPTETextAsset::load_font_nodes() const { vector> lf; copy (_load_font_nodes.begin(), _load_font_nodes.end(), back_inserter(lf)); return lf; } bool SMPTETextAsset::valid_mxf(boost::filesystem::path file) { Kumu::FileReaderFactory factory; ASDCP::TimedText::MXFReader reader(factory); Kumu::DefaultLogSink().UnsetFilterFlag(Kumu::LOG_ALLOW_ALL); auto r = reader.OpenRead(dcp::filesystem::fix_long_path(file).string().c_str()); Kumu::DefaultLogSink().SetFilterFlag(Kumu::LOG_ALLOW_ALL); return !ASDCP_FAILURE (r); } string SMPTETextAsset::xml_as_string() const { xmlpp::Document doc; auto root = doc.create_root_node ("SubtitleReel"); DCP_ASSERT (_xml_id); cxml::add_text_child(root, "Id", "urn:uuid:" + *_xml_id); cxml::add_text_child(root, "ContentTitleText", _content_title_text); if (_annotation_text) { cxml::add_text_child(root, "AnnotationText", _annotation_text.get()); } cxml::add_text_child(root, "IssueDate", _issue_date.as_string(false, false)); if (_reel_number) { cxml::add_text_child(root, "ReelNumber", raw_convert(_reel_number.get())); } if (_language) { cxml::add_text_child(root, "Language", _language.get()); } cxml::add_text_child(root, "EditRate", _edit_rate.as_string()); cxml::add_text_child(root, "TimeCodeRate", raw_convert(_time_code_rate)); if (_start_time) { cxml::add_text_child(root, "StartTime", _start_time.get().as_string(Standard::SMPTE)); } for (auto i: _load_font_nodes) { auto load_font = cxml::add_child(root, "LoadFont"); load_font->add_child_text ("urn:uuid:" + i->urn); load_font->set_attribute ("ID", i->id); } texts_as_xml(cxml::add_child(root, "SubtitleList"), _time_code_rate, Standard::SMPTE); return format_xml(doc, std::make_pair(string{}, schema_namespace())); } void SMPTETextAsset::write(boost::filesystem::path p) const { EncryptionContext enc (key(), Standard::SMPTE); ASDCP::WriterInfo writer_info; fill_writer_info (&writer_info, _id); ASDCP::TimedText::TimedTextDescriptor descriptor; descriptor.EditRate = ASDCP::Rational (_edit_rate.numerator, _edit_rate.denominator); descriptor.EncodingName = "UTF-8"; /* Font references */ for (auto i: _load_font_nodes) { auto j = _fonts.begin(); while (j != _fonts.end() && j->load_id != i->id) { ++j; } if (j != _fonts.end ()) { ASDCP::TimedText::TimedTextResourceDescriptor res; unsigned int c; Kumu::hex2bin (i->urn.c_str(), res.ResourceID, Kumu::UUID_Length, &c); DCP_ASSERT (c == Kumu::UUID_Length); res.Type = ASDCP::TimedText::MT_OPENTYPE; descriptor.ResourceList.push_back (res); } } /* Image subtitle references */ for (auto i: _texts) { auto si = dynamic_pointer_cast(i); if (si) { ASDCP::TimedText::TimedTextResourceDescriptor res; unsigned int c; Kumu::hex2bin (si->id().c_str(), res.ResourceID, Kumu::UUID_Length, &c); DCP_ASSERT (c == Kumu::UUID_Length); res.Type = ASDCP::TimedText::MT_PNG; descriptor.ResourceList.push_back (res); } } descriptor.NamespaceName = schema_namespace(); unsigned int c; DCP_ASSERT (_xml_id); Kumu::hex2bin (_xml_id->c_str(), descriptor.AssetID, ASDCP::UUIDlen, &c); DCP_ASSERT (c == Kumu::UUID_Length); descriptor.ContainerDuration = _intrinsic_duration; ASDCP::TimedText::MXFWriter writer; /* This header size is a guess. Empirically it seems that each subtitle reference is 90 bytes, and we need some extra. The default size is not enough for some feature-length PNG sub projects (see DCP-o-matic #1561). */ ASDCP::Result_t r = writer.OpenWrite(dcp::filesystem::fix_long_path(p).string().c_str(), writer_info, descriptor, _texts.size() * 90 + 16384); if (ASDCP_FAILURE (r)) { boost::throw_exception (FileError ("could not open subtitle MXF for writing", p.string(), r)); } _raw_xml = xml_as_string (); r = writer.WriteTimedTextResource (*_raw_xml, enc.context(), enc.hmac()); if (ASDCP_FAILURE (r)) { boost::throw_exception (MXFFileError ("could not write XML to timed text resource", p.string(), r)); } /* Font payload */ for (auto i: _load_font_nodes) { auto j = _fonts.begin(); while (j != _fonts.end() && j->load_id != i->id) { ++j; } if (j != _fonts.end ()) { ASDCP::TimedText::FrameBuffer buffer; ArrayData data_copy(j->data); buffer.SetData (data_copy.data(), data_copy.size()); buffer.Size (j->data.size()); r = writer.WriteAncillaryResource (buffer, enc.context(), enc.hmac()); if (ASDCP_FAILURE(r)) { boost::throw_exception (MXFFileError ("could not write font to timed text resource", p.string(), r)); } } } /* Image subtitle payload */ for (auto i: _texts) { if (auto si = dynamic_pointer_cast(i)) { ASDCP::TimedText::FrameBuffer buffer; buffer.SetData (si->png_image().data(), si->png_image().size()); buffer.Size (si->png_image().size()); r = writer.WriteAncillaryResource (buffer, enc.context(), enc.hmac()); if (ASDCP_FAILURE(r)) { boost::throw_exception (MXFFileError ("could not write PNG data to timed text resource", p.string(), r)); } } } writer.Finalize (); _file = p; } bool SMPTETextAsset::equals(shared_ptr other_asset, EqualityOptions const& options, NoteHandler note) const { if (!TextAsset::equals (other_asset, options, note)) { return false; } auto other = dynamic_pointer_cast(other_asset); if (!other) { note (NoteType::ERROR, "Subtitles/captions are in different standards"); return false; } auto i = _load_font_nodes.begin(); auto j = other->_load_font_nodes.begin(); while (i != _load_font_nodes.end ()) { if (j == other->_load_font_nodes.end ()) { note (NoteType::ERROR, " nodes differ"); return false; } if ((*i)->id != (*j)->id) { note (NoteType::ERROR, " nodes differ"); return false; } ++i; ++j; } if (_content_title_text != other->_content_title_text) { note (NoteType::ERROR, "Subtitle/caption content title texts differ"); return false; } if (_language != other->_language) { note (NoteType::ERROR, String::compose("Subtitle/caption languages differ (`%1' vs `%2')", _language.get_value_or("[none]"), other->_language.get_value_or("[none]"))); return false; } if (_annotation_text != other->_annotation_text) { note (NoteType::ERROR, "Subtitle/caption annotation texts differ"); return false; } if (_issue_date != other->_issue_date) { if (options.issue_dates_can_differ) { note (NoteType::NOTE, "Subtitle/caption issue dates differ"); } else { note (NoteType::ERROR, "Subtitle/caption issue dates differ"); return false; } } if (_reel_number != other->_reel_number) { note (NoteType::ERROR, "Subtitle/caption reel numbers differ"); return false; } if (_edit_rate != other->_edit_rate) { note (NoteType::ERROR, "Subtitle/caption edit rates differ"); return false; } if (_time_code_rate != other->_time_code_rate) { note (NoteType::ERROR, "Subtitle/caption time code rates differ"); return false; } if (_start_time != other->_start_time) { note (NoteType::ERROR, "Subtitle/caption start times differ"); return false; } return true; } void SMPTETextAsset::add_font(string load_id, dcp::ArrayData data) { string const uuid = make_uuid (); _fonts.push_back (Font(load_id, uuid, data)); _load_font_nodes.push_back (make_shared(load_id, uuid)); } void SMPTETextAsset::add(shared_ptr s) { TextAsset::add(s); _intrinsic_duration = latest_text_out().as_editable_units_ceil(_edit_rate.numerator / _edit_rate.denominator); } string SMPTETextAsset::schema_namespace() const { switch (_subtitle_standard) { case SubtitleStandard::SMPTE_2007: return subtitle_smpte_ns_2007; case SubtitleStandard::SMPTE_2010: return subtitle_smpte_ns_2010; case SubtitleStandard::SMPTE_2014: return subtitle_smpte_ns_2014; default: DCP_ASSERT(false); } DCP_ASSERT(false); }