 |
The
simple API can be learned almost immediately, comprising
only a few basic data manipulation operations. InfinityDB
is based on an advanced B-Tree, providing direct,
optimized, low-level data access.
The
flexible data model is explained in "ItemSpace
Data Structures", which shows how all kinds of
relational, 'Entity-Attribute-Value,' text, array, LOB,
or high-performance customized data structures can be easily
formatted and stored. These structures form a 'representation
layer' on top of the 'engine layer.'
Runtime extension of schemas is possible because
empty structures take no space and are not preallocated,
unlike in SQL-based DBMS, where upgrade scripts proliferate.
All
primitive types are represented, as well as String,
Date, byte and char arrays, huge arrays of anything, and
even byte strings. Keys may be composite or even heterogenous.
Values may be composite, heterogenous, unlimited-cardinality
multi-valued, or unlimited-length Character- or Binary Long
Objects.
|
|
