phantomjs not waiting for “full” page load

Question

I\'m using PhantomJS v1.4.1 to load some web pages. I don\'t have access to their server-side, I just getting links pointing to them. I\'m using obsolete version of Phantom

Answer 1

In my program, I use some logic to judge if it was onload: watching it's network request, if there was no new request on past 200ms, I treat it onload.

Use this, after onLoadFinish().

function onLoadComplete(page, callback){
    var waiting = [];  // request id
    var interval = 200;  //ms time waiting new request
    var timer = setTimeout( timeout, interval);
    var max_retry = 3;  //
    var counter_retry = 0;

    function timeout(){
        if(waiting.length && counter_retry < max_retry){
            timer = setTimeout( timeout, interval);
            counter_retry++;
            return;
        }else{
            try{
                callback(null, page);
            }catch(e){}
        }
    }

    //for debug, log time cost
    var tlogger = {};

    bindEvent(page, 'request', function(req){
        waiting.push(req.id);
    });

    bindEvent(page, 'receive', function (res) {
        var cT = res.contentType;
        if(!cT){
            console.log('[contentType] ', cT, ' [url] ', res.url);
        }
        if(!cT) return remove(res.id);
        if(cT.indexOf('application') * cT.indexOf('text') != 0) return remove(res.id);

        if (res.stage === 'start') {
            console.log('!!received start: ', res.id);
            //console.log( JSON.stringify(res) );
            tlogger[res.id] = new Date();
        }else if (res.stage === 'end') {
            console.log('!!received end: ', res.id, (new Date() - tlogger[res.id]) );
            //console.log( JSON.stringify(res) );
            remove(res.id);

            clearTimeout(timer);
            timer = setTimeout(timeout, interval);
        }

    });

    bindEvent(page, 'error', function(err){
        remove(err.id);
        if(waiting.length === 0){
            counter_retry = 0;
        }
    });

    function remove(id){
        var i = waiting.indexOf( id );
        if(i < 0){
            return;
        }else{
            waiting.splice(i,1);
        }
    }

    function bindEvent(page, evt, cb){
        switch(evt){
            case 'request':
                page.onResourceRequested = cb;
                break;
            case 'receive':
                page.onResourceReceived = cb;
                break;
            case 'error':
                page.onResourceError = cb;
                break;
            case 'timeout':
                page.onResourceTimeout = cb;
                break;
        }
    }
}