From ca29f11383ff2035fa802228c42e26dddcb347d9 Mon Sep 17 00:00:00 2001 From: Michael Berkowitz Date: Wed, 9 Apr 2008 16:56:30 +0000 Subject: [PATCH] -Adds BusinessWeek translator. --- scrapers.sql | 47 ++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 46 insertions(+), 1 deletion(-) diff --git a/scrapers.sql b/scrapers.sql index 16e159840..a3f7ea367 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', '2008-04-08 20:30:00')); +REPLACE INTO version VALUES ('repository', STRFTIME('%s', '2008-04-09 18:00:00')); 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) { @@ -1087,6 +1087,51 @@ REPLACE INTO translators VALUES ('88915634-1af6-c134-0171-56fd198235ed', '1.0.0b Zotero.wait(); }'); +REPLACE INTO translators VALUES ('fb342bae-7727-483b-a871-c64c663c2fae', '1.0.0b4.r5', '', '2008-04-09 18:00:00', '0', '100', '4', 'BusinessWeek', 'Michael Berkowitz', 'http://(www\.)?businessweek.com', +'function detectWeb(doc, url) { + if (doc.title == "BusinessWeek Search Results") { + return "multiple"; + } else if (doc.evaluate(''//meta[@name="headline"]'', doc, null, XPathResult.ANY_TYPE, null).iterateNext()) { + return "magazineArticle"; + } +}', +'function doWeb(doc, url) { + var articles = new Array(); + if (detectWeb(doc, url) == "multiple") { + var results = doc.evaluate(''//div[@class="result"]/h3[@class="story"]/a'', doc, null, XPathResult.ANY_TYPE, null); + var result; + var items = new Object(); + while (result = results.iterateNext()) { + items[result.href] = Zotero.Utilities.trimInternal(result.textContent); + } + items = Zotero.selectItems(items); + for (var i in items) { + articles.push(i); + } + } else { + articles = [url]; + } + Zotero.debug(articles); + Zotero.Utilities.processDocuments(articles, function(newDoc) { + var metaTags = new Object(); + var metas = newDoc.evaluate(''//meta'', newDoc, null, XPathResult.ANY_TYPE, null); + var meta; + while (meta = metas.iterateNext()) { + metaTags[meta.name] = meta.content; + } + Zotero.debug(metaTags); + var item = new Zotero.Item("magazineArticle"); + item.title = metaTags[''headline'']; + item.abstractNote = metaTags[''abstract'']; + item.tags = metaTags[''keywords''].split(/\s*,\s*/); + item.creators.push(Zotero.Utilities.cleanAuthor(metaTags[''author''], "author")); + item.publicationTitle = "BusinessWeek: " + metaTags[''channel'']; + item.url = newDoc.location.href; + item.complete(); + }, function() {Zotero.done;}); + Zotero.wait(); +}'); + REPLACE INTO translators VALUES ('409c520b-0720-4011-8fce-70fcd9806493', '1.0.0b4.r5', '', '2008-04-09 00:45:00', '0', '100', '4', 'InderScience', 'Michael Berkowitz', 'http://www.inderscience.com/', 'function detectWeb(doc, url) { if (doc.evaluate(''/html/body/table/tbody/tr/td[2]/table[tbody/tr/td[3]][2]'', doc, null, XPathResult.ANY_TYPE, null).iterateNext()