Add eplanning adapter

[sebaperez]

Type of change

• Bugfix
• Feature
• New bidder adapter
• Code style update (formatting, local variables)
• Refactoring (no functional changes, no api changes)
• Build related changes
• CI related changes
• Other

Description of change

• test parameters for validating bids

{
  bidder: 'eplanning',
  params: {
    ci: '18f66',
    t: '1'
  }
}

• contact email of the adapter’s maintainer
  soporte@e-planning.net
• official adapter submission

Other information

[dbemiller]

LGTM

@sebaperez Just so you know, we're also looking to prevent JSONP in Prebid 1.0 (probably to be worked on starting in the second half of this year). This is fine for the 0.x releases though.

[protonate]

In test mode (with global param t === '1') neither the processEPlanningResponse or callback functions get called and no bid is added to pbjs._bidsReceived. Please have a look to see if test mode can be made to work similarly to the typical system flow.

[sebaperez]

@protonate I just tested it and works fine for me. Did you test it locally? Do you have the HTML file where it doesn't work?

[dbemiller]

An adapter test page got added in #1182 you can use.

[protonate]

This is the code that is returned by the bid request:

!function() {
    var c, e, a = window, b = a.adUnits, d = {
        sI: {
            k: "1f4b4"
        },
        sec: {
            k: "ROS"
        },
        sp: []
    };
    if (b && b.length)
        for (c = 0; c < b.length; c++)
            if (b[c].bids && b[c].bids.length)
                for (e = 0; e < b[c].bids.length; e++)
                    "eplanning" === b[c].bids[e].bidder && d.sp.push({
                        k: b[c].code,
                        a: [{
                            w: "object" == typeof b[c].sizes[0] ? b[c].sizes[0][0] : b[c].sizes[0],
                            h: "object" == typeof b[c].sizes[0] ? b[c].sizes[0][1] : b[c].sizes[1],
                            adm: "<p>test ad for " + b[c].code + "</p>",
                            pr: 1
                        }]
                    });
    hbpb.rH(d)
}();

The a variable is set to window and then b is set to an a.adUnits property. There isn't a window.adUnits property however. Are you getting this same code block? If so, how is window.adUnits getting set?

[sebaperez]

@protonate that's the response in test case (when t === '1') - If you do not set that variable, then the response is different. In both cases, it is handled by hbpb.rH().
I think the problem you were experiencing was because you tested it locally (file://) and the call to the adserver is made with a relative URL (//...). I just pushed some changes that forces http or https protocol, and never use "file" protocol.

[protonate]

The test case then doesn't return a bid "response". What is the expectation here?

The t param is set as String(1) in the PR description above. However this should be Number(0) to turn off test, as String(0) evaluates to true. It would be clearer if this was set as Number(1) for the test case.

I am testing from localhost and from a "prebidapp.com" domain with a hosts file entry to 127.0.0.1, and on port 9999 in both cases.

With t set to Number(0) the bidder makes this call:

http://ads.us.e-planning.net/hb/1/18f66/1/localhost/ROS?rnd=0.7947294411451566&e=/19968336/header-bid-tag-1:728x90,970x90+/19968336/header-bid-tag-0:300x250,300x600&ur=http%3A//localhost%3A9999/gpt.html&cb=hbpb.rH&r=pbjs

And the response is:

hbpb.rH();

So I'm not able to confirm a bid back in either test or non-test case. Any thoughts?

[dbemiller]

We chatted about it. Not sure what's happening on his system... but everything looks good to me. Screenshots below to prove it (using the hello_world.html file from #1182, substituting the params @sebaperez gave here).

@protonate said he's good with merging this for 0.25.0... so I'm going for it.