This is a Python implementation of the Data Access Protocol, a
scientific protocol for data access developed by the OPeNDAP team
(http://opendap.org). This implementation is developed from scratch,
following the latest specification of the protocol (DAP 2.0 Draft
Community Standard 2005/04/27) and based on my experience with OPeNDAP
servers on the wild.

Using this module one can access hundreds of scientific datasets from
Python programs, accessing data in an efficient, transparent and
pythonic way. Arrays are manipulated like normal multi-dimensional
arrays (like numpy.array, e.g.), with the fundamental difference that
data is downloaded on-the-fly when a variable is sliced. Sequential
data can be filtered on the server side before being downloaded, saving
bandwith and time.

The module also implements a DAP server, allowing datasets from a
multitude of formats (netCDF, Matlab, CSV, GrADS/GRIB files, SQL RDBMS)
to be served on the internet. The server specifies a plugin API for
supporting new data formats in an easy way. The DAP server is
implemented as a WSGI application (see PEP 333), running on a variery
of servers, and can be combined with WSGI middleware to support
authentication, gzip compression and much more.

• Eugeny A. Rostovtsev (REAL)

• python-module-dap