arrayfrombuffer allows a programmer to use Numerical Python arrays whose contents are stored in buffer objects, including memory-mapped files. Loading an array from a file is easy, requiring only a module import and a single function call. Loading the array is quick because it doesn't require any copying from one part of memory to another. Arrays created with arrayfrombuffer are also highly memory-efficient, since only the array data that you are currently using gets loaded into memory. When an array is modified, only the modified parts get written back out to disk. These arrays can also be bigger than physical memory.
python-ldap provides an object-oriented API to access LDAP directory servers from Python programs. Mainly, it wraps the OpenLDAP 2.3+ libs for that purpose. Additionally, the package contains modules for other LDAP-related stuff (such as processing LDIF, LDAPURLs, and LDAPv3 schema).
Kiwi is an object-oriented framework for developing graphical applications using Python. It has a set of classes that provide a solid application architecture, and a set of high level graphical components that can be used as drop-in replacements for some of the widgets in PyGTK, including CList, CTree, Label, OptionMenu, and more. It includes a complete developer's manual, with examples and screenshots, and full API documentation.
PyStem is a fast Python module with the the Porter stemming algorithm (a process for removing the commoner morphological and inflexional endings from words in English; its main use is as part of a term normalisation process that is usually done when setting up Information Retrieval systems).
QuantLib is a cross-platform, quantitative finance C++ library for modeling, pricing, trading, and risk management in real-life. It is also wrapped as Python/Ruby/Scheme modules. Extensions for Excel, R, and Mathematica are available. Other such extensions are under consideration. QuantLib offers tools that are useful both for practical implementation and for advanced modeling. It features market conventions, yield curve models, solvers, PDEs, Monte Carlo (low-discrepancy included), exotic options, VAR, and so on.
py-xmlrpc is an extremely fast implementation of the xmlrpc spec for Python (written in C). It supports both blocking and non-blocking clients and servers on Windows and POSIX platforms. Version 0.8.1 is 100% compliant with the xmlrpc validator found at http://validator.xmlrpc.com.
Libxslt is a C library for GNOME which allows developers to work with XSLT. It is based on libxml for XML parsing, tree manipulation, and XPath support. Also included is 'xsltproc', a command line XSLT processor. The library is written in plain C, making as few assumptions as possible, and sticking closely to ANSI C/POSIX for easy embedding. It should work on Linux, Unix, and Windows. Though not designed primarily with performances in mind, libxslt seems to be a relatively fast processor. It also include full support for the EXSLT set of extension functions as well as some common extensions present in other XSLT engines.