diff --git a/scrapers.sql b/scrapers.sql index a7d555b72..773c1683b 100644 --- a/scrapers.sql +++ b/scrapers.sql @@ -1,15 +1,23 @@ --- 69 +-- 70 -- Set the following timestamp to the most recent scraper update date REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-08-15 15:42:00')); -REPLACE INTO "translators" VALUES ('96b9f483-c44d-5784-cdad-ce21b984fe01', '2006-08-11 11:18:00', 4, 'Amazon.com', 'Simon Kornblith', '^http://www\.amazon\.com/(?:gp/(?:product|search)/|exec/obidos/search-handle-url/|s/)', +REPLACE INTO "translators" VALUES ('96b9f483-c44d-5784-cdad-ce21b984fe01', '2006-08-11 11:18:00', 4, 'Amazon.com', 'Simon Kornblith', '^http://www\.amazon\.com/', 'function detectWeb(doc, url) { var searchRe = new RegExp(''^http://www\.amazon\.com/(gp/search/|exec/obidos/search-handle-url/|s/)''); if(searchRe.test(doc.location.href)) { return "multiple"; } else { - return "book"; + var namespace = doc.documentElement.namespaceURI; + var nsResolver = namespace ? function(prefix) { + if (prefix == ''x'') return namespace; else return null; + } : null; + + var xpath = ''/html/body/table/tbody/tr/td[2]/table/tbody/tr/td[@class="bucket"]/div[@class="content"]/ul/li''; + if(doc.evaluate(xpath, doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext()) { + return "book"; + } } } ',