/************************************************************************ ** ** Copyright (C) 2016 - 2019 Kevin B. Hendricks, Stratford, Ontario, Canada ** Copyright (C) 2012 John Schember ** Copyright (C) 2009, 2010, 2011 Strahinja Markovic ** ** This file is part of Sigil. ** ** Sigil 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. ** ** Sigil 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 Sigil. If not, see . ** *************************************************************************/ #ifdef _WIN32 #define NOMINMAX #endif #include "unzip.h" #ifdef _WIN32 #include "iowin32.h" #endif #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "BookManipulation/FolderKeeper.h" #include "BookManipulation/CleanSource.h" #include "Importers/ImportEPUB.h" #include "Misc/FontObfuscation.h" #include "Misc/HTMLEncodingResolver.h" #include "Misc/QCodePage437Codec.h" #include "Misc/SettingsStore.h" #include "Misc/Utility.h" #include "ResourceObjects/CSSResource.h" #include "ResourceObjects/HTMLResource.h" #include "ResourceObjects/OPFResource.h" #include "ResourceObjects/NCXResource.h" #include "ResourceObjects/Resource.h" #include "ResourceObjects/OPFParser.h" #include "SourceUpdates/UniversalUpdates.h" #include "sigil_constants.h" #include "sigil_exception.h" #ifndef MAX_PATH // Set Max length to 256 because that's the max path size on many systems. #define MAX_PATH 256 #endif // This is the same read buffer size used by Java and Perl. #define BUFF_SIZE 8192 const QString DUBLIN_CORE_NS = "http://purl.org/dc/elements/1.1/"; static const QString OEBPS_MIMETYPE = "application/oebps-package+xml"; static const QString UPDATE_ERROR_STRING = "SG_ERROR"; const QString NCX_MIMETYPE = "application/x-dtbncx+xml"; static const QString NCX_EXTENSION = "ncx"; const QString ADOBE_FONT_ALGO_ID = "http://ns.adobe.com/pdf/enc#RC"; const QString IDPF_FONT_ALGO_ID = "http://www.idpf.org/2008/embedding"; static const QString CONTAINER_XML = "\n" "\n" " \n" " \n" " \n" "\n"; static QCodePage437Codec *cp437 = 0; // Constructor; // The parameter is the file to be imported ImportEPUB::ImportEPUB(const QString &fullfilepath) : Importer(fullfilepath), m_ExtractedFolderPath(m_TempFolder.GetPath()), m_HasSpineItems(false), m_NCXNotInManifest(false), m_NavResource(NULL) { } // Reads and parses the file // and returns the created Book QSharedPointer ImportEPUB::GetBook(bool extract_metadata) { QList non_well_formed; SettingsStore ss; if (!Utility::IsFileReadable(m_FullFilePath)) { throw (EPUBLoadParseError(QString(QObject::tr("Cannot read EPUB: %1")).arg(QDir::toNativeSeparators(m_FullFilePath)).toStdString())); } // These read the EPUB file ExtractContainer(); QHash encrypted_files = ParseEncryptionXml(); if (BookContentEncrypted(encrypted_files)) { throw (FileEncryptedWithDrm("")); } QApplication::setOverrideCursor(Qt::WaitCursor); // These mutate the m_Book object LocateOPF(); m_opfDir = QFileInfo(m_OPFFilePath).dir(); // These mutate the m_Book object ReadOPF(); AddObfuscatedButUndeclaredFonts(encrypted_files); AddNonStandardAppleXML(); LoadInfrastructureFiles(); // Check for files missing in the Manifest and create warning QStringList notInManifest; foreach(QString file_path, m_ZipFilePaths) { // skip mimetype and anything in META-INF and the opf itself if (file_path == "mimetype") continue; if (file_path.startsWith("META-INF")) continue; if (m_OPFFilePath.contains(file_path)) continue; if (!m_ManifestFilePaths.contains(file_path)) { notInManifest << file_path; } } if (!notInManifest.isEmpty()) { Utility::DisplayStdWarningDialog(tr("Files exist in epub that are not listed in the manifest, they will be ignored"), notInManifest.join("\n")); } const QHash updates = LoadFolderStructure(); const QList resources = m_Book->GetFolderKeeper()->GetResourceList(); // We're going to check all html files until we find one that isn't well formed then we'll prompt // the user if they want to auto fix or not. // // If we have non-well formed content and they shouldn't be auto fixed we'll pass that on to // the universal update function so it knows to skip them. Otherwise we won't include them and // let it modify the file. for (int i=0; iType() == Resource::HTMLResourceType) { HTMLResource *hresource = dynamic_cast(resources.at(i)); if (!hresource) { continue; } // Load the content into the HTMLResource so we can perform a well formed check. try { hresource->SetText(HTMLEncodingResolver::ReadHTMLFile(hresource->GetFullPath())); } catch (...) { if (ss.cleanOn() & CLEANON_OPEN) { non_well_formed << hresource; continue; } } if (ss.cleanOn() & CLEANON_OPEN) { if (!XhtmlDoc::IsDataWellFormed(hresource->GetText(),hresource->GetEpubVersion())) { non_well_formed << hresource; } } } } if (!non_well_formed.isEmpty()) { QApplication::restoreOverrideCursor(); if (QMessageBox::Yes == QMessageBox::warning(QApplication::activeWindow(), tr("Sigil"), tr("This EPUB has HTML files that are not well formed. " "Sigil can attempt to automatically fix these files, although this " "can result in data loss.\n\n" "Do you want to automatically fix the files?"), QMessageBox::Yes|QMessageBox::No) ) { non_well_formed.clear(); } QApplication::setOverrideCursor(Qt::WaitCursor); } const QStringList load_errors = UniversalUpdates::PerformUniversalUpdates(false, resources, updates, non_well_formed); Q_FOREACH (QString err, load_errors) { AddLoadWarning(QString("%1").arg(err)); } ProcessFontFiles(resources, updates, encrypted_files); if (m_PackageVersion.startsWith('3')) { HTMLResource * nav_resource = NULL; if (m_NavResource) { if (m_NavResource->Type() == Resource::HTMLResourceType) { nav_resource = dynamic_cast(m_NavResource); } } if (!nav_resource) { // we need to create a nav file here because one was not found // it will automatically be added to the content.opf nav_resource = m_Book->CreateEmptyNavFile(true); Resource * res = dynamic_cast(nav_resource); m_Book->GetOPF()->SetItemRefLinear(res, false); } m_Book->GetOPF()->SetNavResource(nav_resource); } if (m_NCXNotInManifest) { // We manually created an NCX file because there wasn't one in the manifest. // Need to create a new manifest id for it. m_NCXId = m_Book->GetOPF()->AddNCXItem(m_NCXFilePath); } // Ensure that our spine has a element on it now in case it was missing. m_Book->GetOPF()->UpdateNCXOnSpine(m_NCXId); // Make sure the for the NCX in the manifest reflects correct href path m_Book->GetOPF()->UpdateNCXLocationInManifest(m_Book->GetNCX()); // If spine was not present or did not contain any items, recreate the OPF from scratch // preserving any important metadata elements and making a new reading order. if (!m_HasSpineItems) { QList originalMetadata = m_Book->GetOPF()->GetDCMetadata(); m_Book->GetOPF()->AutoFixWellFormedErrors(); if (extract_metadata) { m_Book->GetOPF()->SetDCMetadata(originalMetadata); } AddLoadWarning(QObject::tr("The OPF file does not contain a valid spine.") % "\n" % QObject::tr("Sigil has created a new one for you.")); } // If we have modified the book to add spine attribute, manifest item or NCX mark as changed. m_Book->SetModified(GetLoadWarnings().count() > 0); QApplication::restoreOverrideCursor(); return m_Book; } QHash ImportEPUB::ParseEncryptionXml() { QString encrpytion_xml_path = m_ExtractedFolderPath + "/META-INF/encryption.xml"; if (!QFileInfo(encrpytion_xml_path).exists()) { return QHash(); } QXmlStreamReader encryption(Utility::ReadUnicodeTextFile(encrpytion_xml_path)); QHash encrypted_files; QString encryption_algo; QString uri; while (!encryption.atEnd()) { encryption.readNext(); if (encryption.isStartElement()) { if (encryption.name() == "EncryptionMethod") { encryption_algo = encryption.attributes().value("", "Algorithm").toString(); } else if (encryption.name() == "CipherReference") { uri = Utility::URLDecodePath(encryption.attributes().value("", "URI").toString()); encrypted_files[ uri ] = encryption_algo; } } } if (encryption.hasError()) { const QString error = QString(QObject::tr("Error parsing encryption xml.\nLine: %1 Column %2 - %3")) .arg(encryption.lineNumber()) .arg(encryption.columnNumber()) .arg(encryption.errorString()); throw (EPUBLoadParseError(error.toStdString())); } return encrypted_files; } bool ImportEPUB::BookContentEncrypted(const QHash &encrypted_files) { foreach(QString algorithm, encrypted_files.values()) { if (algorithm != ADOBE_FONT_ALGO_ID && algorithm != IDPF_FONT_ALGO_ID) { return true; } } return false; } // This is basically a workaround for old versions of InDesign not listing the fonts it // embedded in the OPF manifest, even though the specs say it has to. // It does list them in the encryption.xml, so we use that. void ImportEPUB::AddObfuscatedButUndeclaredFonts(const QHash &encrypted_files) { if (encrypted_files.empty()) { return; } QDir opf_dir = QFileInfo(m_OPFFilePath).dir(); foreach(QString filepath, encrypted_files.keys()) { if (!FONT_EXTENSIONS.contains(QFileInfo(filepath).suffix().toLower())) { continue; } // Only add the path to the manifest if it is not already included. QMapIterator valueSearch(m_Files); if (!valueSearch.findNext(opf_dir.relativeFilePath(filepath))) { m_Files[ Utility::CreateUUID() ] = opf_dir.relativeFilePath(filepath); } } } // Another workaround for non-standard Apple files // At present it only handles com.apple.ibooks.display-options.xml, but any // further iBooks aberrations should be handled here as well. void ImportEPUB::AddNonStandardAppleXML() { QDir opf_dir = QFileInfo(m_OPFFilePath).dir(); QStringList aberrant_Apple_filenames; aberrant_Apple_filenames.append(m_ExtractedFolderPath + "/META-INF/com.apple.ibooks.display-options.xml"); for (int i = 0; i < aberrant_Apple_filenames.size(); ++i) { if (QFile::exists(aberrant_Apple_filenames.at(i))) { m_Files[ Utility::CreateUUID() ] = opf_dir.relativeFilePath(aberrant_Apple_filenames.at(i)); } } } // Each resource can provide us with its new path. encrypted_files provides // a mapping from old resource paths to the obfuscation algorithms. // So we use the updates hash which provides a mapping from old paths to new // paths to match the resources to their algorithms. void ImportEPUB::ProcessFontFiles(const QList &resources, const QHash &updates, const QHash &encrypted_files) { if (encrypted_files.empty()) { return; } QList font_resources = m_Book->GetFolderKeeper()->GetResourceTypeList(); if (font_resources.empty()) { return; } QHash new_font_paths_to_algorithms; foreach(QString old_update_path, updates.keys()) { if (!FONT_EXTENSIONS.contains(QFileInfo(old_update_path).suffix().toLower())) { continue; } QString new_update_path = updates[ old_update_path ]; foreach(QString old_encrypted_path, encrypted_files.keys()) { if (old_update_path == old_encrypted_path) { new_font_paths_to_algorithms[ new_update_path ] = encrypted_files[ old_encrypted_path ]; } } } foreach(FontResource * font_resource, font_resources) { QString match_path = "../" + font_resource->GetRelativePathToOEBPS(); QString algorithm = new_font_paths_to_algorithms.value(match_path); if (algorithm.isEmpty()) { continue; } font_resource->SetObfuscationAlgorithm(algorithm); // Actually we are de-obfuscating, but the inverse operations of the obfuscation methods // are the obfuscation methods themselves. For the math oriented, the obfuscation methods // are involutary [ f( f( x ) ) = x ]. if (algorithm == ADOBE_FONT_ALGO_ID) { FontObfuscation::ObfuscateFile(font_resource->GetFullPath(), algorithm, m_UuidIdentifierValue); } else { FontObfuscation::ObfuscateFile(font_resource->GetFullPath(), algorithm, m_UniqueIdentifierValue); } } } void ImportEPUB::ExtractContainer() { int res = 0; if (!cp437) { cp437 = new QCodePage437Codec(); } #ifdef Q_OS_WIN32 zlib_filefunc64_def ffunc; fill_win32_filefunc64W(&ffunc); unzFile zfile = unzOpen2_64(Utility::QStringToStdWString(QDir::toNativeSeparators(m_FullFilePath)).c_str(), &ffunc); #else unzFile zfile = unzOpen64(QDir::toNativeSeparators(m_FullFilePath).toUtf8().constData()); #endif if (zfile == NULL) { throw (EPUBLoadParseError(QString(QObject::tr("Cannot unzip EPUB: %1")).arg(QDir::toNativeSeparators(m_FullFilePath)).toStdString())); } res = unzGoToFirstFile(zfile); if (res == UNZ_OK) { do { // Get the name of the file in the archive. char file_name[MAX_PATH] = {0}; unz_file_info64 file_info; unzGetCurrentFileInfo64(zfile, &file_info, file_name, MAX_PATH, NULL, 0, NULL, 0); QString qfile_name; QString cp437_file_name; qfile_name = QString::fromUtf8(file_name); if (!(file_info.flag & (1<<11))) { // General purpose bit 11 says the filename is utf-8 encoded. If not set then // IBM 437 encoding might be used. cp437_file_name = cp437->toUnicode(file_name); } // If there is no file name then we can't do anything with it. if (!qfile_name.isEmpty()) { // for security reasons against maliciously crafted zip archives // we need the file path to always be inside the target folder // and not outside, so we will remove all illegal backslashes // and all relative upward paths segments "/../" from the zip's local // file name/path before prepending the target folder to create // the final path QString original_path = qfile_name; bool evil_or_corrupt_epub = false; if (qfile_name.contains("\\")) evil_or_corrupt_epub = true; qfile_name = "/" + qfile_name.replace("\\",""); if (qfile_name.contains("/../")) evil_or_corrupt_epub = true; qfile_name = qfile_name.replace("/../","/"); while(qfile_name.startsWith("/")) { qfile_name = qfile_name.remove(0,1); } if (cp437_file_name.contains("\\")) evil_or_corrupt_epub = true; cp437_file_name = "/" + cp437_file_name.replace("\\",""); if (cp437_file_name.contains("/../")) evil_or_corrupt_epub = true; cp437_file_name = cp437_file_name.replace("/../","/"); while(cp437_file_name.startsWith("/")) { cp437_file_name = cp437_file_name.remove(0,1); } if (evil_or_corrupt_epub) { unzCloseCurrentFile(zfile); unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Possible evil or corrupt epub file name: %1")).arg(original_path).toStdString())); } // We use the dir object to create the path in the temporary directory. // Unfortunately, we need a dir ojbect to do this as it's not a static function. QDir dir(m_ExtractedFolderPath); // Full file path in the temporary directory. QString file_path = m_ExtractedFolderPath + "/" + qfile_name; QFileInfo qfile_info(file_path); // Is this entry a directory? if (file_info.uncompressed_size == 0 && qfile_name.endsWith('/')) { dir.mkpath(qfile_name); continue; } else { dir.mkpath(qfile_info.path()); // add it to the list of files found inside the zip if (cp437_file_name.isEmpty()) { m_ZipFilePaths << qfile_name; } else { m_ZipFilePaths << cp437_file_name; } } // Open the file entry in the archive for reading. if (unzOpenCurrentFile(zfile) != UNZ_OK) { unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Cannot extract file: %1")).arg(qfile_name).toStdString())); } // Open the file on disk to write the entry in the archive to. QFile entry(file_path); if (!entry.open(QIODevice::WriteOnly | QIODevice::Truncate)) { unzCloseCurrentFile(zfile); unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Cannot extract file: %1")).arg(qfile_name).toStdString())); } // Buffered reading and writing. char buff[BUFF_SIZE] = {0}; int read = 0; while ((read = unzReadCurrentFile(zfile, buff, BUFF_SIZE)) > 0) { entry.write(buff, read); } entry.close(); // Read errors are marked by a negative read amount. if (read < 0) { unzCloseCurrentFile(zfile); unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Cannot extract file: %1")).arg(qfile_name).toStdString())); } // The file was read but the CRC did not match. // We don't check the read file size vs the uncompressed file size // because if they're different there should be a CRC error. if (unzCloseCurrentFile(zfile) == UNZ_CRCERROR) { unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Cannot extract file: %1")).arg(qfile_name).toStdString())); } if (!cp437_file_name.isEmpty() && cp437_file_name != qfile_name) { QString cp437_file_path = m_ExtractedFolderPath + "/" + cp437_file_name; QFile::copy(file_path, cp437_file_path); } } } while ((res = unzGoToNextFile(zfile)) == UNZ_OK); } if (res != UNZ_END_OF_LIST_OF_FILE) { unzClose(zfile); throw (EPUBLoadParseError(QString(QObject::tr("Cannot open EPUB: %1")).arg(QDir::toNativeSeparators(m_FullFilePath)).toStdString())); } unzClose(zfile); } void ImportEPUB::LocateOPF() { QString fullpath = m_ExtractedFolderPath + "/META-INF/container.xml"; QXmlStreamReader container; try { container.addData(Utility::ReadUnicodeTextFile(fullpath)); } catch (CannotOpenFile) { // Find the first OPF file. QString OPFfile; QDirIterator files(m_ExtractedFolderPath, QStringList() << "*.opf", QDir::NoFilter, QDirIterator::Subdirectories); while (files.hasNext()) { OPFfile = QDir(m_ExtractedFolderPath).relativeFilePath(files.next()); break; } if (OPFfile.isEmpty()) { std::string msg = fullpath.toStdString() + ": " + tr("Epub has missing or improperly specified OPF.").toStdString(); throw (CannotOpenFile(msg)); } // Create a default container.xml. QDir folder(m_ExtractedFolderPath); folder.mkdir("META-INF"); Utility::WriteUnicodeTextFile(CONTAINER_XML.arg(OPFfile), fullpath); container.addData(Utility::ReadUnicodeTextFile(fullpath)); } while (!container.atEnd()) { container.readNext(); if (container.isStartElement() && container.name() == "rootfile" ) { if (container.attributes().hasAttribute("media-type") && container.attributes().value("", "media-type") == OEBPS_MIMETYPE ) { m_OPFFilePath = m_ExtractedFolderPath + "/" + container.attributes().value("", "full-path").toString(); // As per OCF spec, the first rootfile element // with the OEBPS mimetype is considered the "main" one. break; } } } if (container.hasError()) { const QString error = QString( QObject::tr("Unable to parse container.xml file.\nLine: %1 Column %2 - %3")) .arg(container.lineNumber()) .arg(container.columnNumber()) .arg(container.errorString()); throw (EPUBLoadParseError(error.toStdString())); } if (m_OPFFilePath.isEmpty() || !QFile::exists(m_OPFFilePath)) { throw (EPUBLoadParseError(QString(QObject::tr("No appropriate OPF file found")).toStdString())); } } void ImportEPUB::ReadOPF() { QString opf_text = CleanSource::ProcessXML(PrepareOPFForReading(Utility::ReadUnicodeTextFile(m_OPFFilePath)),OEBPS_MIMETYPE); QXmlStreamReader opf_reader(opf_text); QString ncx_id_on_spine; while (!opf_reader.atEnd()) { opf_reader.readNext(); if (!opf_reader.isStartElement()) { continue; } if (opf_reader.name() == "package") { m_UniqueIdentifierId = opf_reader.attributes().value("", "unique-identifier").toString(); m_PackageVersion = opf_reader.attributes().value("", "version").toString(); if (m_PackageVersion == "1.0") m_PackageVersion = "2.0"; } else if (opf_reader.name() == "identifier") { ReadIdentifierElement(&opf_reader); } // epub3 look for linked metadata resources that are included inside the epub // but that are not and must not be included in the manifest else if (opf_reader.name() == "link") { ReadMetadataLinkElement(&opf_reader); } // Get the list of content files that // make up the publication else if (opf_reader.name() == "item") { ReadManifestItemElement(&opf_reader); } // We read this just to get the NCX id else if (opf_reader.name() == "spine") { ncx_id_on_spine = opf_reader.attributes().value("", "toc").toString(); } else if (opf_reader.name() == "itemref") { m_HasSpineItems = true; } } if (opf_reader.hasError()) { const QString error = QString(QObject::tr("Unable to read OPF file.\nLine: %1 Column %2 - %3")) .arg(opf_reader.lineNumber()) .arg(opf_reader.columnNumber()) .arg(opf_reader.errorString()); throw (EPUBLoadParseError(error.toStdString())); } // Ensure we have an NCX available LocateOrCreateNCX(ncx_id_on_spine); } void ImportEPUB::ReadIdentifierElement(QXmlStreamReader *opf_reader) { QString id = opf_reader->attributes().value("", "id").toString(); QString scheme = opf_reader->attributes().value("", "scheme").toString(); QString value = opf_reader->readElementText(); if (id == m_UniqueIdentifierId) { m_UniqueIdentifierValue = value; } if (m_UuidIdentifierValue.isEmpty() && (value.contains("urn:uuid:") || scheme.toLower() == "uuid")) { m_UuidIdentifierValue = value; } } void ImportEPUB::ReadMetadataLinkElement(QXmlStreamReader *opf_reader) { QString relation = opf_reader->attributes().value("", "rel").toString(); QString mtype = opf_reader->attributes().value("", "media-type").toString(); QString props = opf_reader->attributes().value("", "properties").toString(); QString href = opf_reader->attributes().value("", "href").toString(); if (!href.isEmpty()) { QUrl url = QUrl(href); if (url.isRelative()) { // we have a local unmanifested metadata file to handle // attempt to map deprecated record types into proper media-types if (relation == "marc21xml-record") { mtype = "application/marcxml+xml"; } else if (relation == "mods-record") { mtype = "application/mods+xml"; } else if (relation == "onix-record") { mtype = "application/xml;onix"; } else if (relation == "xmp-record") { mtype = "application/xml;xmp"; } else if (relation == "record") { if (props == "onix") mtype = "application/xml;onix"; if (props == "xmp") mtype = "application/xml;xmp"; } QDir opf_dir = QFileInfo(m_OPFFilePath).dir(); QString path = opf_dir.absolutePath() + "/" + url.path(); if (QFile::exists(path)) { QString id = Utility::CreateUUID(); m_Files[ id ] = opf_dir.relativeFilePath(path); m_FileMimetypes[ id ] = mtype; } } } } void ImportEPUB::ReadManifestItemElement(QXmlStreamReader *opf_reader) { QString id = opf_reader->attributes().value("", "id").toString(); QString href = opf_reader->attributes().value("", "href").toString(); QString type = opf_reader->attributes().value("", "media-type").toString(); QString properties = opf_reader->attributes().value("", "properties").toString(); // Paths are percent encoded in the OPF, we use "normal" paths internally. href = Utility::URLDecodePath(href); QString extension = QFileInfo(href).suffix().toLower(); // find the epub root relative file path from the opf location and the item href QString file_path = m_opfDir.absolutePath() + "/" + href; file_path = file_path.remove(0, m_ExtractedFolderPath.length() + 1); if (type != NCX_MIMETYPE && extension != NCX_EXTENSION) { if (!m_ManifestFilePaths.contains(file_path)) { if (m_Files.contains(id)) { // We have an error situation with a duplicate id in the epub. // We must warn the user, but attempt to use another id so the epub can still be loaded. QString base_id = QFileInfo(href).fileName(); QString new_id(base_id); int duplicate_index = 0; while (m_Files.contains(new_id)) { duplicate_index++; new_id = QString("%1%2").arg(base_id).arg(duplicate_index); } const QString load_warning = QObject::tr("The OPF manifest contains duplicate ids for: %1").arg(id) + " - " + QObject::tr("A temporary id has been assigned to load this EPUB. You should edit your OPF file to remove the duplication."); id = new_id; AddLoadWarning(load_warning); } m_Files[ id ] = href; m_FileMimetypes[ id ] = type; m_ManifestFilePaths << file_path; // store information about any nav document if (properties.contains("nav")) { m_NavId = id; m_NavHref = href; } } } else { m_NcxCandidates[ id ] = href; m_ManifestFilePaths << file_path; } } void ImportEPUB::LocateOrCreateNCX(const QString &ncx_id_on_spine) { QString load_warning; QString ncx_href = "not_found"; m_NCXId = ncx_id_on_spine; // Handle various failure conditions, such as: // - ncx not specified in the spine (search for a matching manifest item by extension) // - ncx specified in spine, but no matching manifest item entry (create a new one) // - ncx file not physically present (create a new one) // - ncx not in spine or manifest item (create a new one) if (!m_NCXId.isEmpty()) { ncx_href = m_NcxCandidates[ m_NCXId ]; } else { // Search for the ncx in the manifest by looking for files with // a .ncx extension. QHashIterator ncxSearch(m_NcxCandidates); while (ncxSearch.hasNext()) { ncxSearch.next(); if (QFileInfo(ncxSearch.value()).suffix().toLower() == NCX_EXTENSION) { m_NCXId = ncxSearch.key(); load_warning = QObject::tr("The OPF file did not identify the NCX file correctly.") + "\n" + " - " + QObject::tr("Sigil has used the following file as the NCX:") + QString(" %1").arg(m_NcxCandidates[ m_NCXId ]); ncx_href = m_NcxCandidates[ m_NCXId ]; break; } } } m_NCXFilePath = QFileInfo(m_OPFFilePath).absolutePath() % "/" % ncx_href; if (ncx_href.isEmpty() || !QFile::exists(m_NCXFilePath)) { m_NCXNotInManifest = m_NCXId.isEmpty() || ncx_href.isEmpty(); m_NCXId.clear(); // Things are really bad and no .ncx file was found in the manifest or // the file does not physically exist. We need to create a new one. m_NCXFilePath = QFileInfo(m_OPFFilePath).absolutePath() % "/" % NCX_FILE_NAME; // Create a new file for the NCX. NCXResource ncx_resource(m_ExtractedFolderPath, m_NCXFilePath, m_Book->GetFolderKeeper()); // We are relying on an identifier being set from the metadata. // It might not have one if the book does not have the urn:uuid: format. if (!m_UuidIdentifierValue.isEmpty()) { ncx_resource.SetMainID(m_UuidIdentifierValue); } ncx_resource.SaveToDisk(); if (m_PackageVersion.startsWith('3')) { load_warning = QObject::tr("Sigil has created a template NCX") + "\n" + QObject::tr("to support epub2 backwards compatibility."); } else { if (ncx_href.isEmpty()) { load_warning = QObject::tr("The OPF file does not contain an NCX file.") + "\n" + " - " + QObject::tr("Sigil has created a new one for you."); } else { load_warning = QObject::tr("The NCX file is not present in this EPUB.") + "\n" + " - " + QObject::tr("Sigil has created a new one for you."); } } } if (!load_warning.isEmpty()) { AddLoadWarning(load_warning); } } void ImportEPUB::LoadInfrastructureFiles() { // always SetEpubVersion before SetText in OPF as SetText will validate with it m_Book->GetOPF()->SetEpubVersion(m_PackageVersion); m_Book->GetOPF()->SetText(CleanSource::ProcessXML(PrepareOPFForReading(Utility::ReadUnicodeTextFile(m_OPFFilePath)),OEBPS_MIMETYPE)); QString OPFBookRelPath = m_OPFFilePath; OPFBookRelPath = OPFBookRelPath.remove(0,m_ExtractedFolderPath.length()+1); m_Book->GetOPF()->SetCurrentBookRelPath(OPFBookRelPath); m_Book->GetNCX()->SetText(CleanSource::ProcessXML(Utility::ReadUnicodeTextFile(m_NCXFilePath),"application/x-dtbncx+xml")); m_Book->GetNCX()->SetEpubVersion(m_PackageVersion); QString NCXBookRelPath = m_NCXFilePath; NCXBookRelPath = NCXBookRelPath.remove(0,m_ExtractedFolderPath.length()+1); m_Book->GetNCX()->SetCurrentBookRelPath(NCXBookRelPath); } QHash ImportEPUB::LoadFolderStructure() { QList keys = m_Files.keys(); int num_files = keys.count(); QFutureSynchronizer> sync; for (int i = 0; i < num_files; ++i) { QString id = keys.at(i); sync.addFuture(QtConcurrent::run( this, &ImportEPUB::LoadOneFile, m_Files.value(id), m_FileMimetypes.value(id))); } sync.waitForFinished(); QList>> futures = sync.futures(); int num_futures = futures.count(); QHash updates; for (int i = 0; i < num_futures; ++i) { std::tuple result = futures.at(i).result(); updates[std::get<0>(result)] = std::get<1>(result); } updates.remove(UPDATE_ERROR_STRING); return updates; } std::tuple ImportEPUB::LoadOneFile(const QString &path, const QString &mimetype) { QString fullfilepath = QDir::cleanPath(QFileInfo(m_OPFFilePath).absolutePath() + "/" + path); QString currentpath = fullfilepath; currentpath = currentpath.remove(0,m_ExtractedFolderPath.length()+1); try { Resource *resource = m_Book->GetFolderKeeper()->AddContentFileToFolder(fullfilepath, false, mimetype); if (path == m_NavHref) { m_NavResource = resource; } resource->SetCurrentBookRelPath(currentpath); QString newpath = "../" + resource->GetRelativePathToOEBPS(); return std::make_tuple(currentpath, newpath); } catch (FileDoesNotExist) { return std::make_tuple(UPDATE_ERROR_STRING, UPDATE_ERROR_STRING); } } QString ImportEPUB::PrepareOPFForReading(const QString &source) { QString source_copy(source); QString prefix = source_copy.left(XML_DECLARATION_SEARCH_PREFIX_SIZE); QRegularExpression version(VERSION_ATTRIBUTE); QRegularExpressionMatch mo = version.match(prefix); if (mo.hasMatch()) { // MASSIVE hack for XML 1.1 "support"; // this is only for people who specify // XML 1.1 when they actually only use XML 1.0 source_copy.replace(mo.capturedStart(1), mo.capturedLength(1), "1.0"); } return source_copy; }