Module sqlarray

source code

:mod:`sqlarray` --- Implementation of :class:`SQLarray`
=======================================================

:class:`SQLarray` is a thin wrapper around pysqlite SQL tables. The main
features ares that ``SELECT`` queries can return ``numpy.recarrays`` and the
:meth:`SQLarray.selection` method returns a new :class:`SQLarray` instance.  

numpy arrays can be stored in sql fields which allows advanced table
aggregate functions such as ``histogram``.

A number of additional SQL functions are defined.

:TODO:
   * Make object saveable (i.e. store the database on disk instead of
     memory or dump the memory db and provide a load() method
   * Use hooks for the pickling protocol to make this transparent. 

.. SeeAlso:: PyTables_ is a high-performance interface to table data. 

.. _PyTables: http://www.pytables.org

Module content
--------------
.. See the autogenerated content in the online docs or the source code.

Create a :class:`SQLarray` from *filename*.

Uses the filename suffix to detect the contents:
  rst, txt
      restructure text (see :mod:`recsql.rest_table`
  csv
      comma-separated (see :mod:`recsql.csv_table`)

:Arguments:
  *filename*
      name of the file that contains the data with the appropriate
      file extension
  *kwargs*
      - additional arguments for :class:`SQLarray`
      - additional arguments :class:`recsql.csv_table.Table2array` or
        :class:`recsql.rest_table.Table2array` such as *mode* or
        *autoncovert*.