A sequence is an unbounded array of eXtremeDB-supported scalar data elements. (Please see the Sequences page for further explanation of the sequence data type.)
Sequences can be stored, iterated, transformed, and aggregated using SQL or native language APIs. Please use the links below to view the guide for your development environment:
Using Sequences with SQL.Using Sequences in Python.Using Sequences in C.Using Sequences in C++.Using Sequences in Java.Using Sequences in C#.Notes on Sequences and Sequence Iterators
Note the following regarding the implementation of sequences and sequence iterators:
- The eXtremeDB runtime does not require a third-party arithmetic library. A build can use processor vector instructions to optimize compatible operations.
- Most operations are lazy: the result iterator retains and consumes its operand iterators. Do not attach one input iterator to several independent pipelines. Create equivalent source iterators or use the API's tee operation when two consumers are required.
- An iterator, its operands, and any transaction, object handle, cursor, allocator, callback context, or application buffer they reference must remain valid until traversal finishes.
- Use each iterator's installed reset operation rather than assuming a common implementation. Resetting a result also resets the operand subtree required by that result.
- Typed traversal always works with logical elements. Direct tile traversal exposes physical storage details, including RLE runs and null bitmaps, and is intended for advanced code.
C applications should start with Using Sequences in C, then use the generated field and C sequence function references linked from that guide.