diff --git a/scrapers.sql b/scrapers.sql index 1b7eaf5fb..9ea0ff50a 100644 --- a/scrapers.sql +++ b/scrapers.sql @@ -1,7 +1,7 @@ -- 12 -- Set the following timestamp to the most recent scraper update date -REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-06-23 12:17:00')); +REPLACE INTO "version" VALUES ('repository', STRFTIME('%s', '2006-06-23 13:34:00')); REPLACE INTO "scrapers" VALUES('96b9f483-c44d-5784-cdad-ce21b984fe01', '2006-06-22 22:58:00', 'Amazon.com Scraper', 'Simon Kornblith', '^http://www\.amazon\.com/(?:gp/(?:product|search)/|exec/obidos/search-handle-url/)', NULL, 'var prefixRDF = ''http://www.w3.org/1999/02/22-rdf-syntax-ns#''; var prefixDC = ''http://purl.org/dc/elements/1.1/''; @@ -96,10 +96,7 @@ if(m) { } utilities.processDocuments(browser, null, uris, function(browser) { scrape(browser.contentDocument) }, - function() { - utilities.debugPrint("look, done"); - done(); - }, function() {}); + function() { done(); }, function() {}); wait(); } else { @@ -511,7 +508,7 @@ if(month && year) { model.addStatement(uri, prefixRDF + "type", prefixDummy + "journal", false); '); -REPLACE INTO "scrapers" VALUES('4fd6b89b-2316-2dc4-fd87-61a97dd941e8', '2006-06-23 10:11:00', 'InnoPAC Scraper', 'Simon Kornblith', '^http://[^/]+/(?:search/|record=)', +REPLACE INTO "scrapers" VALUES('4fd6b89b-2316-2dc4-fd87-61a97dd941e8', '2006-06-23 12:49:00', 'InnoPAC Scraper', 'Simon Kornblith', '^http://[^/]+/(?:search/|record=)', '// First, check to see if the URL alone reveals InnoPAC, since some sites don''t reveal the MARC button var matchRegexp = new RegExp(''^(http://[^/]+/search/[^/]+/[^/]+/1\%2C[^/]+/)frameset(.+)$''); if(matchRegexp.test(doc.location.href)) { @@ -591,7 +588,7 @@ if(newUri) { var urls = new Array(); var availableItems = new Array(); - var tableRows = utilities.gatherElementsOnXPath(doc, doc, ''//tr[@class="browseEntry"]'', nsResolver); + var tableRows = utilities.gatherElementsOnXPath(doc, doc, ''//table[@class="browseScreen"]//tr[td/input[@name="save"]]'', nsResolver); // Go through table rows for(var i=0; i