Import packaging codes geocoordinates to MongoDB

[hangy]

In order to implementing MadeNearMe efficiently (related to #712), it would be useful to import the packaging codes into MongoDB to make them searchable with a geospatial index. ProductOpener::Display::search_and_map_products has the product to get all coordinates.

[hangy]

With the codes imported, we can run something like this:

db.products.aggregate([
  { $unwind: "$emb_codes_tags" },
  { $lookup:
    { from: "emb_codes",
      localField: "emb_codes_tags",
      foreignField: "emb_code",
      "as": "emb_codes_coordinates"
    }
  },
  { $unwind: "$emb_codes_coordinates" },
    { $match: {
     "emb_codes_coordinates.loc": {
       $geoWithin: {
          $geometry: {"type":"MultiPolygon","coordinates":[[[[-5.661949,54.554603],[-6.197885,53.867565],[-6.95373,54.073702],[-7.572168,54.059956],[-7.366031,54.595841],[-7.572168,55.131622],[-6.733847,55.17286],[-5.661949,54.554603]]],[[[-3.005005,58.635],[-4.073828,57.553025],[-3.055002,57.690019],[-1.959281,57.6848],[-2.219988,56.870017],[-3.119003,55.973793],[-2.085009,55.909998],[-2.005676,55.804903],[-1.114991,54.624986],[-0.430485,54.464376],[0.184981,53.325014],[0.469977,52.929999],[1.681531,52.73952],[1.559988,52.099998],[1.050562,51.806761],[1.449865,51.289428],[0.550334,50.765739],[-0.787517,50.774989],[-2.489998,50.500019],[-2.956274,50.69688],[-3.617448,50.228356],[-4.542508,50.341837],[-5.245023,49.96],[-5.776567,50.159678],[-4.30999,51.210001],[-3.414851,51.426009],[-3.422719,51.426848],[-4.984367,51.593466],[-5.267296,51.9914],[-4.222347,52.301356],[-4.770013,52.840005],[-4.579999,53.495004],[-3.093831,53.404547],[-3.09208,53.404441],[-2.945009,53.985],[-3.614701,54.600937],[-3.630005,54.615013],[-4.844169,54.790971],[-5.082527,55.061601],[-4.719112,55.508473],[-5.047981,55.783986],[-5.586398,55.311146],[-5.644999,56.275015],[-6.149981,56.78501],[-5.786825,57.818848],[-5.009999,58.630013],[-4.211495,58.550845],[-3.005005,58.635]]]]}
       }
     }
   }
  },
  { $group:
    {  _id:
      { emb_code: "$emb_codes_coordinates.emb_code" },
        count: { $sum: 1 }
      }
  }
]);

Which returns this on my test server:

{ "_id" : { "emb_code" : "uk-1005-ec" }, "count" : 1 }
{ "_id" : { "emb_code" : "uk-bb011-ec" }, "count" : 1 }

(The geometry is the UK as taken from https://github.com/johan/world.geo.json/blob/master/countries.geo.json and would be replaced with the Leaflet bounding box when we run the query dynamically.)