Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@couchdb.apache.org
Received-SPF: pass (nike.apache.org: domain of bosak.tomas@gmail.com
 designates 209.85.212.52 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:date:message-id:subject:from:to:content-type;
        b=KMlOZ5Tx3EZa16dQN2W76TVyQbqmvKiCvDYJBCV0/C9iGfqeBHd4sYf75aGpe0Ez63
         btaMP1CJKOS5LRdj7fAM7swW7zOB60Raskt0ZXJKthT8WY8kDVDUKB5IztHwFXKA11t6
         MCxkiEpnAjewf1YmMXGtPYVFisLwq7r9Uw4yc=
MIME-Version: 1.0
Date: Sun, 19 Dec 2010 11:27:07 +0100
Message-ID: <AANLkTik94vLMucJ=2vT-ACv-vhRffueG8CoB-h_tqFOT@mail.gmail.com>
Subject: CouchDB querying pattern
From: Tomi <bosak.tomas@gmail.com>
To: user@couchdb.apache.org
Content-Type: text/plain; charset=UTF-8

Hi folks!

Is it a good idea to query couchdb like this:

- stored documents have format like this:

{
    "_id": ...,
    "_rev": ...,
    "Title": "Test title",
    "Accounts": ["joe", "johny"],
    "Labels": ["work", "school", "programming"],
    "Status": 0,
    "Type": "ticket"
}

- map function will generate rows where keys consists of document
field name and it's value (multiple times when it's an array):

function(doc) {
  if(doc.Type && (doc.Type == "ticket")) {
    emit(["_id", doc._id], null);
    emit(["title", doc.Title], null);
    for(var index in doc.Accounts) {
      emit(["accounts", doc.Accounts[index]], null);
    }
    for(var index2 in doc.Labels) {
      emit(["labels", doc.Labels[index2]], null);
    }
    emit(["status", doc.Status], null);
  }
}

- list function[1] will merge the map function result based on data
sent in POST like this:

curl -X POST -d
'{"keys":[["accounts","joe"],["labels","school"],["status",2]]}' -H
"Content-Type: application/json"
http://127.0.0.1:5984/testdb/_design/docs/_list/listTest/adhoc?include_docs=true

Result contains only "ticket" type documents which have "joe" in
Accounts field, "school" in Labels field and number 2 in Status field.
Is it a good idea to use this "pattern" for "ad-hoc querying" in
couchdb? I can imagine that database with millions of documents (or
with thousands of documents where each document have lots of fields or
big arrays) could have serious performance issues due to map and list
function from this scenario. Thanks for your help and time.

[1] http://www.vertigrated.com/blog/2010/04/generic-ad-hoc-queries-in-couchdb/