Run patch monkey in async mode #575
Description
Follow up this discussion https://groups.google.com/forum/?fromgroups=#!topic/qlio/2whvdxaTFiQ
I want to implement simple html scraper into ql.io. My idea is when you write query like code below, ql.io will fetch the page, create browser like environment (JSDom), include jQuery, execute query and fetch result. But my problem is JSDom run in async mode but ql.io 's patch monkey run in sync mode, so ql.io return result before JSDom finish running. In example below, result always is blank array. Any chance to fix that?
I think because of this code:
query
select * from www.scraper
where url = "http://abcdef.com/xyz"
and title = '[{"selector":"ul li a.title"}]'
and url = '[{"selector":"ul li a.title", "prop": "href"}]';
-- definition of www.scraper table
create table www.scraper
on select get from "{url}"
using patch 'www.scraper.js'
resultset 'results'
var jsdom = require('jsdom');
exports['parse response'] = function(args) {
var result = [];
// get body
var body = '';
_.each(args.body, function(buf) {
body += buf.toString('UTF-8');
});
// create JSDom environment, include jQuery
jsdom.env({
html: body,
scripts: ['http://ajax.googleapis.com/ajax/libs/jquery/1.8.3/jquery.min.js'],
done: function(err, window) {
// run CSS query using jQuery
// jQuery(selector).each(function() { result.push($(element).text()) });
}
});
// return result as new response
return {
type: 'application/json',
content: JSON.stringify(result)
}
}