make Amazon.com translator work on book pages linked from other book pages
This commit is contained in:
parent
58a1fe2bf2
commit
6f885c9cb0
1 changed files with 11 additions and 3 deletions
14
scrapers.sql
14
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";
|
||||
}
|
||||
}
|
||||
}
|
||||
',
|
||||
|
|
Loading…
Reference in a new issue