diff --git a/scrapers.sql b/scrapers.sql index c7c91561a..7ef06f1e6 100644 --- a/scrapers.sql +++ b/scrapers.sql @@ -22,7 +22,7 @@ -- Set the following timestamp to the most recent scraper update date -REPLACE INTO version VALUES ('repository', STRFTIME('%s', '2008-03-28 18:15:00')); +REPLACE INTO version VALUES ('repository', STRFTIME('%s', '2008-03-28 20:00:00')); REPLACE INTO translators VALUES ('96b9f483-c44d-5784-cdad-ce21b984fe01', '1.0.0b4.r1', '', '2008-03-21 20:00:00', '1', '100', '4', 'Amazon.com', 'Sean Takats and Michael Berkowitz', '^https?://(?:www\.)?amazon', 'function detectWeb(doc, url) { @@ -1579,6 +1579,70 @@ function doWeb(doc, url) { } }'); +REPLACE INTO translators VALUES ('7cb0089b-9551-44b2-abca-eb03cbf586d9', '1.0.0b4.r5', '', '2008-03-28 20:00:00', '0', '100', '4', 'BioOne', 'Michael Berkowitz', 'http://[^/]*www.bioone.org[^/]*/', +'function detectWeb(doc, url) { + if (url.indexOf("searchtype") != -1) { + return "multiple"; + } else if (url.indexOf("get-document") != -1 || url.indexOf("get-abstract") != -1) { + return "journalArticle"; + } +}', +'function createCitationURL(str) { + str = str.match(/doi=([^&]+)/)[1]; + return "http://www.bioone.org/perlserv/?request=cite-builder&doi=" + str; +} + +function getPDFurl(item) { + var bits = new Array( + item.DOI.match(/\/([^(]+)\(/)[1], + item.volume, + item.issue, + item.pages.match(/^([^-]+)\-/)[1] + ); + return "http://www.bioone.org/archive/" + bits.slice(0,3).join("/") + "/pdf/i" + bits.join("-") + ".pdf"; +} + +function doWeb(doc, url) { + var articles = new Array(); + if (detectWeb(doc, url) == "multiple") { + var items = new Object(); + var results = doc.evaluate(''//div[@class="content"]/table/tbody/tr/td[3][@class="group"]'', doc, null, XPathResult.ANY_TYPE, null); + var next_result; + while (next_result = results.iterateNext()) { + var title = doc.evaluate(''.//span[@class="title"]'', next_result, null, XPathResult.ANY_TYPE, null).iterateNext().textContent; + var link = doc.evaluate(''.//tr[4]/td/a[1]'', next_result, null, XPathResult.ANY_TYPE, null).iterateNext().href; + items[link] = title; + } + items = Zotero.selectItems(items); + for (var i in items) { + articles.push(createCitationURL(i)); + } + } else { + articles = [createCitationURL(url)]; + } + Zotero.debug(articles); + Zotero.Utilities.processDocuments(articles, function(newDoc) { + var newlink = newDoc.evaluate(''//a[contains(@href, "refman")]'', newDoc, null, XPathResult.ANY_TYPE, null).iterateNext().href; + Zotero.Utilities.HTTP.doGet(newlink, function(text) { + var translator = Zotero.loadTranslator("import"); + translator.setTranslator("32d59d2d-b65a-4da4-b0a3-bdd3cfb979e7"); + translator.setString(text); + translator.setHandler("itemDone", function(obj, item) { + item.url = decodeURIComponent(item.url); + item.DOI = item.url.match(/http:\/\/dx\.doi\.org\/(.*)$/)[1]; + var pdfurl = getPDFurl(item); + item.attachments = [ + {url:item.url, title:item.title, mimeType:"text/html"}, + {url:pdfurl, title:"BioOne Full Text PDF", mimeType:"application/pdf"} + ]; + item.complete(); + }); + translator.translate(); + }); + }, function() {Zotero.done;}); + Zotero.wait(); +}'); + REPLACE INTO translators VALUES ('b8a86e36-c270-48c9-bdd1-22aaa167ef46', '1.0.0b4.r5', '', '2008-03-28 17:45:00', '0', '100', '4', 'Agencia del ISBN', 'Michael Berkowitz', 'http://www.mcu.es/cgi-brs/BasesHTML', 'function detectWeb(doc, url) { if (doc.evaluate(''//div[@id="formularios"]/div[@class="isbnResultado"]/div[@class="isbnResDescripcion"]'', doc, null, XPathResult.ANY_TYPE, null).iterateNext()) {