Database structure for metaquery - database

Question

I have a database that has a the following structure:
+------+------+--------+-------+--------+-------+
| item | type | color | speed | length | width |
+------+------+--------+-------+--------+-------+
| 1 | 1 | 3 | 1 | 2 | 2 |
| 2 | 1 | 5 | 3 | 1 | 1 |
| 3 | 1 | 6 | 3 | 1 | 1 |
| 4 | 2 | 2 | 1 | 3 | 1 |
| 5 | 2 | 2 | 2 | 2 | 1 |
| 6 | 2 | 4 | 2 | 3 | 1 |
| 7 | 2 | 5 | 1 | 1 | 2 |
| 8 | 3 | 1 | 2 | 2 | 2 |
| 9 | 4 | 4 | 3 | 1 | 2 |
| 10 | 4 | 6 | 3 | 3 | 2 |
+------+------+--------+-------+--------+-------+
I would like to efficiently query what combination of fields are valid. So, for example, I'd like to query the database for the following:
What values of color are valid if type is 1?
ans: [3, 5, 6]
What values of speed are valid if type is 2 and color is 2?
ans: [1, 2]
What values of type are valid if length is 2 and width is 2?
ans: [1, 2]
The SQL equivalents are:
SELECT DISTINCT `color` FROM `cars` WHERE `type` =2
SELECT DISTINCT `speed` FROM `cars` WHERE `type` =2 AND `width` =2
SELECT DISTINCT `type` FROM `cars` WHERE `length` =2 AND `width` =2
I'm planning on using a cloud based database (Cloudant DBAAS - based on CouchDB). How would this best be implemented, keeping in mind that there may be thousands of items with tens of fields?

Answer 1

I haven't put too much thought into this question, so there may be errors in the approach, but one option is to represent each row with a document:
{
"_id": "1db91338150bfcfe5fcadbd98fe77d56",
"_rev": "1-83daafc1596c2dabd4698742c2d8b0cf",
"item": 1,
"type": 1,
"color": 3,
"speed": 1,
"length": 2,
"width": 2
}
Note the _id and _rev fields have been automatically generated by Cloudant for this example.
You could then create a secondary index on the type field:
function(doc) {
if(doc.type)
emit(doc.type);
}
To search using the type field:
https://accountname.cloudant.com/dashboard.html#database/so/_design/ddoc/_view/col_for_type?key=1&include_docs=true
A secondary index on the type and width fields:
function(doc) {
if( doc.type && doc.width)
emit([doc.type, doc.width]);
}
To search using the type and width fields:
https://accountname.cloudant.com/dashboard.html#database/so/_design/ddoc/_view/speed_for_type_and_width?key=[1,2]&include_docs=true
A secondary index on the length and width fields:
function(doc) {
if (doc.length && doc.width)
emit([doc.length, doc.width]);
}
To search using the length and width fields:
https://accountname.cloudant.com/dashboard.html#/database/so/_design/ddoc/_view/type_for_length_and_width?key=[2,2]&include_docs=true
The complete design document is here:
{
"_id": "_design\/ddoc",
"_rev": "3-c87d7c3cd44dcef35a030e23c1c91711",
"views": {
"col_for_type": {
"map": "function(doc) {\n if(doc.type)\n emit(doc.type);\n}"
},
"speed_for_type_and_width": {
"map": "function(doc) {\n if( doc.type && doc.width)\n emit([doc.type, doc.width]);\n}"
},
"type_for_length_and_width": {
"map": "function(doc) {\n if (doc.length && doc.width)\n emit([doc.length, doc.width]);\n}"
}
},
"language": "javascript"
}