/*************************************************************************** copyright : (C) 2004-2006 by Robby Stephenson email : robby@periapsis.org ***************************************************************************/ /*************************************************************************** * * * This program is free software; you can redistribute it and/or modify * * it under the terms of version 2 of the GNU General Public License as * * published by the Free Software Foundation; * * * ***************************************************************************/ #include "amazonfetcher.h" #include "messagehandler.h" #include "../translators/xslthandler.h" #include "../translators/tellicoimporter.h" #include "../imagefactory.h" #include "../tellico_kernel.h" #include "../latin1literal.h" #include "../collection.h" #include "../document.h" #include "../entry.h" #include "../field.h" #include "../tellico_utils.h" #include "../tellico_debug.h" #include "../isbnvalidator.h" #include "../gui/combobox.h" #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include namespace { static const int AMAZON_RETURNS_PER_REQUEST = 10; static const int AMAZON_MAX_RETURNS_TOTAL = 20; static const char* AMAZON_ACCESS_KEY = "0834VQ4S71KYPVSYQD02"; static const char* AMAZON_ASSOC_TOKEN = "tellico-20"; // need to have these in the translation file static const char* linkText = I18N_NOOP("Amazon Link"); } using Tellico::Fetch::AmazonFetcher; // static const AmazonFetcher::SiteData& AmazonFetcher::siteData(int site_) { static SiteData dataVector[6] = { { i18n("Amazon (US)"), "http://webservices.amazon.com/onca/xml" }, { i18n("Amazon (UK)"), "http://webservices.amazon.co.uk/onca/xml" }, { i18n("Amazon (Germany)"), "http://webservices.amazon.de/onca/xml" }, { i18n("Amazon (Japan)"), "http://webservices.amazon.co.jp/onca/xml" }, { i18n("Amazon (France)"), "http://webservices.amazon.fr/onca/xml" }, { i18n("Amazon (Canada)"), "http://webservices.amazon.ca/onca/xml" } }; return dataVector[site_]; } AmazonFetcher::AmazonFetcher(Site site_, TQObject* tqparent_, const char* name_) : Fetcher(tqparent_, name_), m_xsltHandler(0), m_site(site_), m_imageSize(MediumImage), m_access(TQString::tqfromLatin1(AMAZON_ACCESS_KEY)), m_assoc(TQString::tqfromLatin1(AMAZON_ASSOC_TOKEN)), m_addLinkField(true), m_limit(AMAZON_MAX_RETURNS_TOTAL), m_countOffset(0), m_page(1), m_total(-1), m_numResults(0), m_job(0), m_started(false) { m_name = siteData(site_).title; } AmazonFetcher::~AmazonFetcher() { delete m_xsltHandler; m_xsltHandler = 0; } TQString AmazonFetcher::defaultName() { return i18n("Amazon.com Web Services"); } TQString AmazonFetcher::source() const { return m_name.isEmpty() ? defaultName() : m_name; } bool AmazonFetcher::canFetch(int type) const { return type == Data::Collection::Book || type == Data::Collection::ComicBook || type == Data::Collection::Bibtex || type == Data::Collection::Album || type == Data::Collection::Video || type == Data::Collection::Game; } void AmazonFetcher::readConfigHook(const KConfigGroup& config_) { TQString s = config_.readEntry("AccessKey"); if(!s.isEmpty()) { m_access = s; } s = config_.readEntry("AssocToken"); if(!s.isEmpty()) { m_assoc = s; } int imageSize = config_.readNumEntry("Image Size", -1); if(imageSize > -1) { m_imageSize = static_cast(imageSize); } m_fields = config_.readListEntry("Custom Fields", TQString::tqfromLatin1("keyword")); } void AmazonFetcher::search(FetchKey key_, const TQString& value_) { m_key = key_; m_value = value_.stripWhiteSpace(); m_started = true; m_page = 1; m_total = -1; m_countOffset = 0; m_numResults = 0; doSearch(); } void AmazonFetcher::continueSearch() { m_started = true; m_limit += AMAZON_MAX_RETURNS_TOTAL; doSearch(); } void AmazonFetcher::doSearch() { m_data.truncate(0); // myDebug() << "AmazonFetcher::doSearch() - value = " << m_value << endl; // myDebug() << "AmazonFetcher::doSearch() - getting page " << m_page << endl; const SiteData& data = siteData(m_site); KURL u = data.url; u.addQueryItem(TQString::tqfromLatin1("Service"), TQString::tqfromLatin1("AWSECommerceService")); u.addQueryItem(TQString::tqfromLatin1("AssociateTag"), m_assoc); u.addQueryItem(TQString::tqfromLatin1("AWSAccessKeyId"), m_access); u.addQueryItem(TQString::tqfromLatin1("Operation"), TQString::tqfromLatin1("ItemSearch")); u.addQueryItem(TQString::tqfromLatin1("ResponseGroup"), TQString::tqfromLatin1("Large")); u.addQueryItem(TQString::tqfromLatin1("ItemPage"), TQString::number(m_page)); u.addQueryItem(TQString::tqfromLatin1("Version"), TQString::tqfromLatin1("2007-10-29")); const int type = Kernel::self()->collectionType(); switch(type) { case Data::Collection::Book: case Data::Collection::ComicBook: case Data::Collection::Bibtex: u.addQueryItem(TQString::tqfromLatin1("SearchIndex"), TQString::tqfromLatin1("Books")); u.addQueryItem(TQString::tqfromLatin1("SortIndex"), TQString::tqfromLatin1("relevancerank")); break; case Data::Collection::Album: u.addQueryItem(TQString::tqfromLatin1("SearchIndex"), TQString::tqfromLatin1("Music")); break; case Data::Collection::Video: u.addQueryItem(TQString::tqfromLatin1("SearchIndex"), TQString::tqfromLatin1("Video")); u.addQueryItem(TQString::tqfromLatin1("SortIndex"), TQString::tqfromLatin1("relevancerank")); break; case Data::Collection::Game: u.addQueryItem(TQString::tqfromLatin1("SearchIndex"), TQString::tqfromLatin1("VideoGames")); break; case Data::Collection::Coin: case Data::Collection::Stamp: case Data::Collection::Wine: case Data::Collection::Base: case Data::Collection::Card: default: message(i18n("%1 does not allow searching for this collection type.").tqarg(source()), MessageHandler::Warning); stop(); return; } // I have not been able to find any documentation about what character set to use // when URL encoding the search term in the Amazon REST interface. But I do know // that utf8 DOES NOT WORK. So I'm arbitrarily using iso-8859-1, except for JP. // Why different for JP? Well, I've not received any bug reports from that direction yet // TQString value = KURL::decode_string(value_, 106); // TQString value = TQString::fromLocal8Bit(value_.utf8()); TQString value = m_value; // a mibenum of 106 is utf-8, 4 is iso-8859-1, 0 means use user's locale, int mib = m_site == AmazonFetcher::JP ? 106 : 4; switch(m_key) { case Title: u.addQueryItem(TQString::tqfromLatin1("Title"), value, mib); break; case Person: if(type == Data::Collection::Video) { u.addQueryItem(TQString::tqfromLatin1("Actor"), value, mib); u.addQueryItem(TQString::tqfromLatin1("Director"), value, mib); } else if(type == Data::Collection::Album) { u.addQueryItem(TQString::tqfromLatin1("Artist"), value, mib); } else if(type == Data::Collection::Game) { u.addQueryItem(TQString::tqfromLatin1("Manufacturer"), value, mib); } else { // books and bibtex TQString s = TQString::tqfromLatin1("author:%1 or publisher:%2").tqarg(value, value); // u.addQueryItem(TQString::tqfromLatin1("Author"), value, mib); // u.addQueryItem(TQString::tqfromLatin1("Publisher"), value, mib); u.addQueryItem(TQString::tqfromLatin1("Power"), s, mib); } break; case ISBN: { u.removeQueryItem(TQString::tqfromLatin1("Operation")); u.addQueryItem(TQString::tqfromLatin1("Operation"), TQString::tqfromLatin1("ItemLookup")); TQString s = m_value; // not encValue!!! s.remove('-'); // ISBN only get digits or 'X', and multiple values are connected with "; " TQStringList isbns = TQStringList::split(TQString::tqfromLatin1("; "), s); // Amazon isbn13 search is still very flaky, so if possible, we're going to convert // all of them to isbn10. If we run into a 979 isbn13, then we're forced to do an // isbn13 search bool isbn13 = false; for(TQStringList::Iterator it = isbns.begin(); it != isbns.end(); ) { if(m_value.startsWith(TQString::tqfromLatin1("979"))) { if(m_site == JP) { // never works for JP kdWarning() << "AmazonFetcher:doSearch() - ISBN-13 searching not implemented for Japan" << endl; isbns.remove(it); // automatically skips to next continue; } isbn13 = true; break; } ++it; } // if we want isbn10, then convert all if(!isbn13) { for(TQStringList::Iterator it = isbns.begin(); it != isbns.end(); ++it) { if((*it).length() > 12) { (*it) = ISBNValidator::isbn10(*it); (*it).remove('-'); } } // the default search is by ASIN, which prohibits SearchIndex u.removeQueryItem(TQString::tqfromLatin1("SearchIndex")); } // limit to first 10 while(isbns.size() > 10) { isbns.pop_back(); } u.addQueryItem(TQString::tqfromLatin1("ItemId"), isbns.join(TQString::tqfromLatin1(","))); if(isbn13) { u.addQueryItem(TQString::tqfromLatin1("IdType"), TQString::tqfromLatin1("EAN")); } } break; case UPC: { u.removeQueryItem(TQString::tqfromLatin1("Operation")); u.addQueryItem(TQString::tqfromLatin1("Operation"), TQString::tqfromLatin1("ItemLookup")); // US allows UPC, all others are EAN if(m_site == US) { u.addQueryItem(TQString::tqfromLatin1("IdType"), TQString::tqfromLatin1("UPC")); } else { u.addQueryItem(TQString::tqfromLatin1("IdType"), TQString::tqfromLatin1("EAN")); } TQString s = m_value; // not encValue!!! s.remove('-'); // limit to first 10 s.tqreplace(TQString::tqfromLatin1("; "), TQString::tqfromLatin1(",")); s = s.section(',', 0, 9); u.addQueryItem(TQString::tqfromLatin1("ItemId"), s); } break; case Keyword: u.addQueryItem(TQString::tqfromLatin1("Keywords"), m_value, mib); break; case Raw: { TQString key = value.section('=', 0, 0).stripWhiteSpace(); TQString str = value.section('=', 1).stripWhiteSpace(); u.addQueryItem(key, str, mib); } break; default: kdWarning() << "AmazonFetcher::search() - key not recognized: " << m_key << endl; stop(); return; } // myDebug() << "AmazonFetcher::search() - url: " << u.url() << endl; m_job = KIO::get(u, false, false); connect(m_job, TQT_SIGNAL(data(KIO::Job*, const TQByteArray&)), TQT_SLOT(slotData(KIO::Job*, const TQByteArray&))); connect(m_job, TQT_SIGNAL(result(KIO::Job*)), TQT_SLOT(slotComplete(KIO::Job*))); } void AmazonFetcher::stop() { if(!m_started) { return; } // myDebug() << "AmazonFetcher::stop()" << endl; if(m_job) { m_job->kill(); m_job = 0; } m_data.truncate(0); m_started = false; emit signalDone(this); } void AmazonFetcher::slotData(KIO::Job*, const TQByteArray& data_) { TQDataStream stream(m_data, IO_WriteOnly | IO_Append); stream.writeRawBytes(data_.data(), data_.size()); } void AmazonFetcher::slotComplete(KIO::Job* job_) { // myDebug() << "AmazonFetcher::slotComplete()" << endl; // since the fetch is done, don't worry about holding the job pointer m_job = 0; if(job_->error()) { job_->showErrorDialog(Kernel::self()->widget()); stop(); return; } if(m_data.isEmpty()) { myDebug() << "AmazonFetcher::slotComplete() - no data" << endl; stop(); return; } #if 0 kdWarning() << "Remove debug from amazonfetcher.cpp" << endl; TQFile f(TQString::tqfromLatin1("/tmp/test%1.xml").tqarg(m_page)); if(f.open(IO_WriteOnly)) { TQTextStream t(&f); t.setEncoding(TQTextStream::UnicodeUTF8); t << TQCString(m_data, m_data.size()+1); } f.close(); #endif TQStringList errors; if(m_total == -1) { TQDomDocument dom; if(!dom.setContent(m_data, false)) { kdWarning() << "AmazonFetcher::slotComplete() - server did not return valid XML." << endl; stop(); return; } // find TotalResults element // it's in the first level under the root element //ItemSearchResponse/Items/TotalResults TQDomNode n = dom.documentElement().namedItem(TQString::tqfromLatin1("Items")) .namedItem(TQString::tqfromLatin1("TotalResults")); TQDomElement e = n.toElement(); if(!e.isNull()) { m_total = e.text().toInt(); } n = dom.documentElement().namedItem(TQString::tqfromLatin1("Items")) .namedItem(TQString::tqfromLatin1("Request")) .namedItem(TQString::tqfromLatin1("Errors")); e = n.toElement(); if(!e.isNull()) { TQDomNodeList nodes = e.elementsByTagName(TQString::tqfromLatin1("Error")); for(uint i = 0; i < nodes.count(); ++i) { e = nodes.item(i).toElement().namedItem(TQString::tqfromLatin1("Code")).toElement(); if(!e.isNull() && e.text() == Latin1Literal("AWS.ECommerceService.NoExactMatches")) { // no exact match, not a real error, so skip continue; } // for some reason, Amazon will return an error simply when a valid ISBN is not found // I really want to ignore that, so check the IsValid element in the Request element TQDomNode isValidNode = n.tqparentNode().namedItem(TQString::tqfromLatin1("IsValid")); if(m_key == ISBN && isValidNode.toElement().text().lower() == Latin1Literal("true")) { continue; } e = nodes.item(i).toElement().namedItem(TQString::tqfromLatin1("Message")).toElement(); if(!e.isNull()) { errors << e.text(); } } } } if(!m_xsltHandler) { initXSLTHandler(); if(!m_xsltHandler) { // probably an error somewhere in the stylesheet loading stop(); return; } } // TQRegExp stripHTML(TQString::tqfromLatin1("<.*>"), true); // stripHTML.setMinimal(true); // assume amazon is always utf-8 TQString str = m_xsltHandler->applyStylesheet(TQString::fromUtf8(m_data, m_data.size())); Import::TellicoImporter imp(str); Data::CollPtr coll = imp.collection(); if(!coll) { myDebug() << "AmazonFetcher::slotComplete() - no collection pointer" << endl; stop(); return; } if(!m_addLinkField) { // remove amazon field if it's not to be added coll->removeField(TQString::tqfromLatin1("amazon")); } Data::EntryVec entries = coll->entries(); if(entries.isEmpty() && !errors.isEmpty()) { for(TQStringList::ConstIterator it = errors.constBegin(); it != errors.constEnd(); ++it) { myDebug() << "AmazonFetcher::" << *it << endl; } message(errors[0], MessageHandler::Error); stop(); return; } int count = 0; for(Data::EntryVec::Iterator entry = entries.begin(); m_numResults < m_limit && entry != entries.end(); ++entry, ++count) { if(count < m_countOffset) { continue; } if(!m_started) { // might get aborted break; } // special case book author // amazon is really bad about not putting spaces after periods if(coll->type() == Data::Collection::Book) { TQRegExp rx(TQString::tqfromLatin1("\\.([^\\s])")); TQStringList values = entry->fields(TQString::tqfromLatin1("author"), false); for(TQStringList::Iterator it = values.begin(); it != values.end(); ++it) { (*it).tqreplace(rx, TQString::tqfromLatin1(". \\1")); } entry->setField(TQString::tqfromLatin1("author"), values.join(TQString::tqfromLatin1("; "))); } // UK puts the year in the title for some reason if(m_site == UK && coll->type() == Data::Collection::Video) { TQRegExp rx(TQString::tqfromLatin1("\\[(\\d{4})\\]")); TQString t = entry->title(); if(t.tqfind(rx) > -1) { TQString y = rx.cap(1); t.remove(rx).simplifyWhiteSpace(); entry->setField(TQString::tqfromLatin1("title"), t); if(entry->field(TQString::tqfromLatin1("year")).isEmpty()) { entry->setField(TQString::tqfromLatin1("year"), y); } } } TQString desc; switch(coll->type()) { case Data::Collection::Book: case Data::Collection::ComicBook: case Data::Collection::Bibtex: desc = entry->field(TQString::tqfromLatin1("author")) + TQChar('/') + entry->field(TQString::tqfromLatin1("publisher")); if(!entry->field(TQString::tqfromLatin1("cr_year")).isEmpty()) { desc += TQChar('/') + entry->field(TQString::tqfromLatin1("cr_year")); } else if(!entry->field(TQString::tqfromLatin1("pub_year")).isEmpty()){ desc += TQChar('/') + entry->field(TQString::tqfromLatin1("pub_year")); } break; case Data::Collection::Video: desc = entry->field(TQString::tqfromLatin1("studio")) + TQChar('/') + entry->field(TQString::tqfromLatin1("director")) + TQChar('/') + entry->field(TQString::tqfromLatin1("year")) + TQChar('/') + entry->field(TQString::tqfromLatin1("medium")); break; case Data::Collection::Album: desc = entry->field(TQString::tqfromLatin1("artist")) + TQChar('/') + entry->field(TQString::tqfromLatin1("label")) + TQChar('/') + entry->field(TQString::tqfromLatin1("year")); break; case Data::Collection::Game: desc = entry->field(TQString::tqfromLatin1("platform")) + TQChar('/') + entry->field(TQString::tqfromLatin1("year")); break; default: break; } // strip HTML from comments, or plot in movies // tentatively don't do this, looks like ECS 4 cleaned everything up /* if(coll->type() == Data::Collection::Video) { TQString plot = entry->field(TQString::tqfromLatin1("plot")); plot.remove(stripHTML); entry->setField(TQString::tqfromLatin1("plot"), plot); } else if(coll->type() == Data::Collection::Game) { TQString desc = entry->field(TQString::tqfromLatin1("description")); desc.remove(stripHTML); entry->setField(TQString::tqfromLatin1("description"), desc); } else { TQString comments = entry->field(TQString::tqfromLatin1("comments")); comments.remove(stripHTML); entry->setField(TQString::tqfromLatin1("comments"), comments); } */ // myDebug() << "AmazonFetcher::slotComplete() - " << entry->title() << endl; SearchResult* r = new SearchResult(this, entry->title(), desc, entry->field(TQString::tqfromLatin1("isbn"))); m_entries.insert(r->uid, Data::EntryPtr(entry)); emit signalResultFound(r); ++m_numResults; } // we might have gotten aborted if(!m_started) { return; } // are there any additional results to get? m_hasMoreResults = m_page * AMAZON_RETURNS_PER_REQUEST < m_total; const int currentTotal = TQMIN(m_total, m_limit); if(m_page * AMAZON_RETURNS_PER_REQUEST < currentTotal) { int foundCount = (m_page-1) * AMAZON_RETURNS_PER_REQUEST + coll->entryCount(); message(i18n("Results from %1: %2/%3").tqarg(source()).tqarg(foundCount).tqarg(m_total), MessageHandler::tqStatus); ++m_page; m_countOffset = 0; doSearch(); } else if(m_value.tqcontains(';') > 9) { search(m_key, m_value.section(';', 10)); } else { m_countOffset = m_entries.count() % AMAZON_RETURNS_PER_REQUEST; if(m_countOffset == 0) { ++m_page; // need to go to next page } stop(); } } Tellico::Data::EntryPtr AmazonFetcher::fetchEntry(uint uid_) { Data::EntryPtr entry = m_entries[uid_]; if(!entry) { kdWarning() << "AmazonFetcher::fetchEntry() - no entry in dict" << endl; return 0; } TQStringList defaultFields = customFields().keys(); for(TQStringList::Iterator it = defaultFields.begin(); it != defaultFields.end(); ++it) { if(!m_fields.tqcontains(*it)) { entry->setField(*it, TQString()); } } // do what we can to remove useless keywords const int type = Kernel::self()->collectionType(); switch(type) { case Data::Collection::Book: case Data::Collection::ComicBook: case Data::Collection::Bibtex: { const TQString keywords = TQString::tqfromLatin1("keyword"); TQStringList oldWords = entry->fields(keywords, false); StringSet words; for(TQStringList::Iterator it = oldWords.begin(); it != oldWords.end(); ++it) { // the amazon2tellico stylesheet separates keywords with '/' TQStringList nodes = TQStringList::split('/', *it); for(TQStringList::Iterator it2 = nodes.begin(); it2 != nodes.end(); ++it2) { if(*it2 == Latin1Literal("General") || *it2 == Latin1Literal("Subjects") || *it2 == Latin1Literal("Par prix") || // french stuff *it2 == Latin1Literal("Divers") || // french stuff (*it2).startsWith(TQChar('(')) || (*it2).startsWith(TQString::tqfromLatin1("Authors"))) { continue; } words.add(*it2); } } entry->setField(keywords, words.toList().join(TQString::tqfromLatin1("; "))); } entry->setField(TQString::tqfromLatin1("comments"), Tellico::decodeHTML(entry->field(TQString::tqfromLatin1("comments")))); break; case Data::Collection::Video: { const TQString genres = TQString::tqfromLatin1("genre"); TQStringList oldWords = entry->fields(genres, false); StringSet words; // only care about genres that have "Genres" in the amazon response // and take the first word after that for(TQStringList::Iterator it = oldWords.begin(); it != oldWords.end(); ++it) { if((*it).tqfind(TQString::tqfromLatin1("Genres")) == -1) { continue; } // the amazon2tellico stylesheet separates words with '/' TQStringList nodes = TQStringList::split('/', *it); for(TQStringList::Iterator it2 = nodes.begin(); it2 != nodes.end(); ++it2) { if(*it2 != Latin1Literal("Genres")) { continue; } ++it2; if(it2 != nodes.end() && *it2 != Latin1Literal("General")) { words.add(*it2); } break; // we're done } } entry->setField(genres, words.toList().join(TQString::tqfromLatin1("; "))); // language tracks get duplicated, too TQStringList langs = entry->fields(TQString::tqfromLatin1("language"), false); words.clear(); for(TQStringList::ConstIterator it = langs.begin(); it != langs.end(); ++it) { words.add(*it); } entry->setField(TQString::tqfromLatin1("language"), words.toList().join(TQString::tqfromLatin1("; "))); } entry->setField(TQString::tqfromLatin1("plot"), Tellico::decodeHTML(entry->field(TQString::tqfromLatin1("plot")))); break; case Data::Collection::Album: { const TQString genres = TQString::tqfromLatin1("genre"); TQStringList oldWords = entry->fields(genres, false); StringSet words; // only care about genres that have "Styles" in the amazon response // and take the first word after that for(TQStringList::Iterator it = oldWords.begin(); it != oldWords.end(); ++it) { if((*it).tqfind(TQString::tqfromLatin1("Styles")) == -1) { continue; } // the amazon2tellico stylesheet separates words with '/' TQStringList nodes = TQStringList::split('/', *it); bool isStyle = false; for(TQStringList::Iterator it2 = nodes.begin(); it2 != nodes.end(); ++it2) { if(!isStyle) { if(*it2 == Latin1Literal("Styles")) { isStyle = true; } continue; } if(*it2 != Latin1Literal("General")) { words.add(*it2); } } } entry->setField(genres, words.toList().join(TQString::tqfromLatin1("; "))); } entry->setField(TQString::tqfromLatin1("comments"), Tellico::decodeHTML(entry->field(TQString::tqfromLatin1("comments")))); break; case Data::Collection::Game: entry->setField(TQString::tqfromLatin1("description"), Tellico::decodeHTML(entry->field(TQString::tqfromLatin1("description")))); break; } // clean up the title parseTitle(entry, type); // also sometimes table fields have rows but no values Data::FieldVec fields = entry->collection()->fields(); TQRegExp blank(TQString::tqfromLatin1("[\\s:;]+")); // only white space, column separators and row separators for(Data::FieldVec::Iterator fIt = fields.begin(); fIt != fields.end(); ++fIt) { if(fIt->type() != Data::Field::Table) { continue; } if(blank.exactMatch(entry->field(fIt))) { entry->setField(fIt, TQString()); } } KURL imageURL; switch(m_imageSize) { case SmallImage: imageURL = entry->field(TQString::tqfromLatin1("small-image")); break; case MediumImage: imageURL = entry->field(TQString::tqfromLatin1("medium-image")); break; case LargeImage: imageURL = entry->field(TQString::tqfromLatin1("large-image")); break; case NoImage: default: break; } // myDebug() << "AmazonFetcher::fetchEntry() - grabbing " << imageURL.prettyURL() << endl; if(!imageURL.isEmpty()) { TQString id = ImageFactory::addImage(imageURL, true); // FIXME: need to add cover image field to bibtex collection if(id.isEmpty()) { message(i18n("The cover image could not be loaded."), MessageHandler::Warning); } else { // amazon serves up 1x1 gifs occasionally, but that's caught in the image constructor // all relevant collection types have cover fields entry->setField(TQString::tqfromLatin1("cover"), id); } } // don't want to show image urls in the fetch dialog entry->setField(TQString::tqfromLatin1("small-image"), TQString()); entry->setField(TQString::tqfromLatin1("medium-image"), TQString()); entry->setField(TQString::tqfromLatin1("large-image"), TQString()); return entry; } void AmazonFetcher::initXSLTHandler() { TQString xsltfile = locate("appdata", TQString::tqfromLatin1("amazon2tellico.xsl")); if(xsltfile.isEmpty()) { kdWarning() << "AmazonFetcher::initXSLTHandler() - can not locate amazon2tellico.xsl." << endl; return; } KURL u; u.setPath(xsltfile); delete m_xsltHandler; m_xsltHandler = new XSLTHandler(u); if(!m_xsltHandler->isValid()) { kdWarning() << "AmazonFetcher::initXSLTHandler() - error in amazon2tellico.xsl." << endl; delete m_xsltHandler; m_xsltHandler = 0; return; } } void AmazonFetcher::updateEntry(Data::EntryPtr entry_) { // myDebug() << "AmazonFetcher::updateEntry()" << endl; int type = entry_->collection()->type(); if(type == Data::Collection::Book || type == Data::Collection::ComicBook || type == Data::Collection::Bibtex) { TQString isbn = entry_->field(TQString::tqfromLatin1("isbn")); if(!isbn.isEmpty()) { m_limit = 5; // no need for more search(Fetch::ISBN, isbn); return; } } else if(type == Data::Collection::Album) { TQString a = entry_->field(TQString::tqfromLatin1("artist")); if(!a.isEmpty()) { search(Fetch::Person, a); return; } } // optimistically try searching for title and rely on Collection::sameEntry() to figure things out TQString t = entry_->field(TQString::tqfromLatin1("title")); if(!t.isEmpty()) { search(Fetch::Title, t); return; } myDebug() << "AmazonFetcher::updateEntry() - insufficient info to search" << endl; emit signalDone(this); // always need to emit this if not continuing with the search } void AmazonFetcher::parseTitle(Data::EntryPtr entry, int collType) { Q_UNUSED(collType); // assume that everything in brackets or tqparentheses is extra TQRegExp rx(TQString::tqfromLatin1("[\\(\\[](.*)[\\)\\]]")); rx.setMinimal(true); TQString title = entry->field(TQString::tqfromLatin1("title")); int pos = rx.search(title); while(pos > -1) { if(parseTitleToken(entry, rx.cap(1))) { title.remove(pos, rx.matchedLength()); --pos; // search again there } pos = rx.search(title, pos+1); } entry->setField(TQString::tqfromLatin1("title"), title.stripWhiteSpace()); } bool AmazonFetcher::parseTitleToken(Data::EntryPtr entry, const TQString& token) { // if res = true, then the token gets removed from the title bool res = false; if(token.tqfind(TQString::tqfromLatin1("widescreen"), 0, false /* case-insensitive*/) > -1 || token.tqfind(i18n("Widescreen"), 0, false) > -1) { entry->setField(TQString::tqfromLatin1("widescreen"), TQString::tqfromLatin1("true")); // res = true; leave it in the title } else if(token.tqfind(TQString::tqfromLatin1("full screen"), 0, false) > -1) { // skip, but go ahead and remove from title res = true; } if(token.tqfind(TQString::tqfromLatin1("blu-ray"), 0, false) > -1) { entry->setField(TQString::tqfromLatin1("medium"), i18n("Blu-ray")); res = true; } else if(token.tqfind(TQString::tqfromLatin1("hd dvd"), 0, false) > -1) { entry->setField(TQString::tqfromLatin1("medium"), i18n("HD DVD")); res = true; } if(token.tqfind(TQString::tqfromLatin1("director's cut"), 0, false) > -1 || token.tqfind(i18n("Director's Cut"), 0, false) > -1) { entry->setField(TQString::tqfromLatin1("directors-cut"), TQString::tqfromLatin1("true")); // res = true; leave it in the title } return res; } Tellico::Fetch::ConfigWidget* AmazonFetcher::configWidget(TQWidget* tqparent_) const { return new AmazonFetcher::ConfigWidget(tqparent_, this); } AmazonFetcher::ConfigWidget::ConfigWidget(TQWidget* tqparent_, const AmazonFetcher* fetcher_/*=0*/) : Fetch::ConfigWidget(tqparent_) { TQGridLayout* l = new TQGridLayout(optionsWidget(), 4, 2); l->setSpacing(4); l->setColStretch(1, 10); int row = -1; TQLabel* label = new TQLabel(i18n("Co&untry: "), optionsWidget()); l->addWidget(label, ++row, 0); m_siteCombo = new GUI::ComboBox(optionsWidget()); m_siteCombo->insertItem(i18n("United States"), US); m_siteCombo->insertItem(i18n("United Kingdom"), UK); m_siteCombo->insertItem(i18n("Germany"), DE); m_siteCombo->insertItem(i18n("Japan"), JP); m_siteCombo->insertItem(i18n("France"), FR); m_siteCombo->insertItem(i18n("Canada"), CA); connect(m_siteCombo, TQT_SIGNAL(activated(int)), TQT_SLOT(slotSetModified())); connect(m_siteCombo, TQT_SIGNAL(activated(int)), TQT_SLOT(slotSiteChanged())); l->addWidget(m_siteCombo, row, 1); TQString w = i18n("Amazon.com provides data from several different localized sites. Choose the one " "you wish to use for this data source."); TQWhatsThis::add(label, w); TQWhatsThis::add(m_siteCombo, w); label->setBuddy(m_siteCombo); label = new TQLabel(i18n("&Image size: "), optionsWidget()); l->addWidget(label, ++row, 0); m_imageCombo = new GUI::ComboBox(optionsWidget()); m_imageCombo->insertItem(i18n("Small Image"), SmallImage); m_imageCombo->insertItem(i18n("Medium Image"), MediumImage); m_imageCombo->insertItem(i18n("Large Image"), LargeImage); m_imageCombo->insertItem(i18n("No Image"), NoImage); connect(m_imageCombo, TQT_SIGNAL(activated(int)), TQT_SLOT(slotSetModified())); l->addWidget(m_imageCombo, row, 1); w = i18n("The cover image may be downloaded as well. However, too many large images in the " "collection may degrade performance."); TQWhatsThis::add(label, w); TQWhatsThis::add(m_imageCombo, w); label->setBuddy(m_imageCombo); label = new TQLabel(i18n("&Associate's ID: "), optionsWidget()); l->addWidget(label, ++row, 0); m_assocEdit = new KLineEdit(optionsWidget()); connect(m_assocEdit, TQT_SIGNAL(textChanged(const TQString&)), TQT_SLOT(slotSetModified())); l->addWidget(m_assocEdit, row, 1); w = i18n("The associate's id identifies the person accessing the Amazon.com Web Services, and is included " "in any links to the Amazon.com site."); TQWhatsThis::add(label, w); TQWhatsThis::add(m_assocEdit, w); label->setBuddy(m_assocEdit); l->setRowStretch(++row, 10); if(fetcher_) { m_siteCombo->setCurrentData(fetcher_->m_site); m_assocEdit->setText(fetcher_->m_assoc); m_imageCombo->setCurrentData(fetcher_->m_imageSize); } else { // defaults m_assocEdit->setText(TQString::tqfromLatin1(AMAZON_ASSOC_TOKEN)); m_imageCombo->setCurrentData(MediumImage); } addFieldsWidget(AmazonFetcher::customFields(), fetcher_ ? fetcher_->m_fields : TQStringList()); KAcceleratorManager::manage(optionsWidget()); } void AmazonFetcher::ConfigWidget::saveConfig(KConfigGroup& config_) { int n = m_siteCombo->currentData().toInt(); config_.writeEntry("Site", n); TQString s = m_assocEdit->text().stripWhiteSpace(); if(!s.isEmpty()) { config_.writeEntry("AssocToken", s); } n = m_imageCombo->currentData().toInt(); config_.writeEntry("Image Size", n); saveFieldsConfig(config_); slotSetModified(false); } TQString AmazonFetcher::ConfigWidget::preferredName() const { return AmazonFetcher::siteData(m_siteCombo->currentData().toInt()).title; } void AmazonFetcher::ConfigWidget::slotSiteChanged() { emit signalName(preferredName()); } //static Tellico::StringMap AmazonFetcher::customFields() { StringMap map; map[TQString::tqfromLatin1("keyword")] = i18n("Keywords"); return map; } #include "amazonfetcher.moc"