mohawk | building page-object like structures

Query

• its an aggregation update even if pipeline doesn't allow us to use stages like group lookup etc that here is used. (you can use $out and replace collection after or $merge to replace documents(similar to update))

• first map

  • for each trait(document member of traits), it makes it into array
    [["trait_type": "type"] ["value": "Male"] ["display_type": null] ...]
  • reduce on that array to contruct from them 1 document only
    {"type" "type","value" :"Male"} (does also that lowercase and "_")

• Now traits its like

Saving data on keys, makes it more complicated, because we have to do things like $objectToArray see this also

For each trait in traits array adds 1 field to the eternal document. {"trait_$value" : ["$value1" "$value2"]} if trait had many values of {"trait_$value" : "$value1"} if trait had 1 value only

("trait_$value" gets lower-case and replace spaces with "_" also)

Query (works on your data and produce results with data on keys)

If you look at the code example we have 2 collections, your collection, and one dummy c2=[{}], we need this to make the $lookup so we can use the stage operator like $group dont delete it, just add this dummy 1 empty document collection to the database.

Because we had to do things like $group it was not possible to do it with update, there are 2 solutions

• add an $out stage to the aggregation (output to new collection and replace it)
• add a $merge stage to the aggregation, that works like an update

Example data1 Test code here

Example data2 (i added 1 eye_color to make an array) Test code here

Query

• $lookup with a dummy collection [{}] reason to do that is to use stage operators inside 1 document to avoid things like unwind-group (we need to group that traits array)
• Inside the lookup pipeline (in the pipeline its just the ROOT doc)
• unwind traits (each traite separate document)
• group by trait_type, collect the values {"_id" : type1 "values" [value1 ...]} {"_id" : type2 "values" [value1 ...]}
• now on that we change {"_id" : type1 "values" [value1 ...]} to (object to array $$ROOT) [["traits_type1" [value1 ...]]] (if 1 value we dont have array) to (array to object $$pair) {"traits_type1" : [value1 ...]}
• the next is making that array "traits" output of lookup to object
• and add it back to the root

I'm not sure this is the exact answer to the OP but here is my two-cents. In my opinion it is better to keep the geographical data separate from the non geographical data. So first I created some function stateNGData (non geographical data). This function return an object with a method shufffle for easy shuffling of the state language. I then created the object data. In this way it is easy to retrieve|update what ever information about the state in question.

It is because your function executed before DOM fully loaded.

Write function in $(document).ready()