Multi-dimensional data structure to efficiently search data items
The disclosed embodiments include a computer-implemented method that can include generating a sub-map for each of multiple data items, which each includes keys, key values, and a data item value, and where each sub-map maps keys to key values. The method can further include creating a composite key for each sub-map, where a particular composite key is a composite of key values of a particular data item. The method can further include generating, based on the sub-maps, a multi-dimensional data structure including a dimension for each key of the sub-maps where each dimension maps any key values of the dimension's key to any composite key that includes the key value, and where the multi-dimensional data structure is searchable instead of searching the plurality of data items.
1. A method comprising:
receiving a query for a plurality of data items, the query specifying a key and a key value, each data item including a plurality of keys and corresponding key values, and a data value;
applying the query to a multi-dimensional dataset instead of querying the plurality of data items, the multi-dimensional dataset including a dimension for each of the plurality of keys, each dimension mapping a key value of a respective key to any composite keys that include the key value, wherein applying the query to the multi-dimensional dataset further comprises:
interrogating the multi-dimensional dataset for a composite key by performing a first look-up operation for the queried key value in a corresponding dimension of the multi-dimensional dataset;
interrogating a main data map for a data item value by performing a second look-up based on the composite key, wherein the main data map maps composite keys to data item values of the plurality of data items;
performing a function based on the data item value to obtain the data derived from the data item value; and
returning the data-item value or data derived from the data-item value in response to the query; and
returning query results that satisfies the query,
wherein the function is a sum or average function including the data-item value.
2. The method of claim 1 , wherein each dimension of the multi-dimensional dataset has a minimal number of key values required to represent all key values for the key of the dimension.
3. The method of claim 1 , wherein the query is in a human-readable format.