CouchDB for Fixed Categories Queries

I have documents like this in my CouchDB:

{
  "_id": "0cb35be3cc73d6859c303fa3200011d2",
  "_rev": "1-f6e356bbf6ab09290aae11132af50d66",
  "adresse": "Bohrgaß 10 /",
  "plz": 56814,
  "ort": "Faid /",
  "kw": 2.32,
  "traeger": "SOL"
  ...
}

There are predefined categories for certain attributes e.g. traeger: "SOL", "BIO", "WAS"; kw: <2, 2-5, 5-20, 20-100; plz: 56814, plz: 56815; ...

I have to be able to efficiently query the total number of docs for every category and the total number of docs and the docs itself under certain conditions. E.g.

1. How many docs are in the category kw <2 (and all other kw categories) under the condition traeger = "SOL"
2. How many docs are in the category traeger = "SOL" (and all other traeger categories) under the conditions plz=56814 AND kw < 2

The user can sele开发者_开发问答ct which catagories he likes to combine. The categories are fix. There also will be more attributes and catagories.

How would map/ reduce functions for this look like?

Marcel

Since you are going to count documents, your reduce function is simply the built-in count. Your map function needs to emit the appropriate keys your users are going to search for. Finally, when the view is queried, the appropriate group level has to be picked.

Example: You can create a view with a composite key ["traeger", "kw"]. If you query that view with group_level = 2, you get the number of documents for each combination of traeger and kw.

If you only care about the traeger "SOL", you can restrict the output with the start_key and end_key parameters.

If you want to know the number of documents in each "traeger" category no matter their "kw", you can query that view with group_level 1.

For your second example, you can create a view with the key ["plz","kw","traeger"] and query it using start_key and end_key to restrict the results to plz=56814 AND kw < 2 and set group_level to 3.

Querying options for views are listed here:

http://wiki.apache.org/couchdb/HTTP_view_API#Querying_Options