Continue function call speed enhancements. Neal Norwitz patch review. Zombie frames? TimPeters will continue.
Faster parsing of strings and bytes into int, long, and back. TimPeters: patch ready - needs clean-up
Buffer for use with network I/O MartinBlais working in blair-bytebuf branch --> sandbox? [Performance comparisons first]
Patch list review DONE (see ListOfPerformanceRelatedPatches)
Implement portions of the decimal module in C: GeorgBrandl and JackDeiderich
Create a PyInt64 type (for 32 bit machines) Evaluate source impact ...GvR: Not in 2.5! [/Deferred?]
- Build-out struct module to support fast, high-volume binary conversions [perhaps with regexp analogs such struct.compile() and struct.finditer()] Offset arg? DONE
Optimize methods in stringobject.c FredrikLundh + AndrewDalke
Allow selective removal of unused features such as profiling support DONE [/DeferredSuccesses?]
Expose/develop pre-allocation arguments/methods for dict/list? API? [/DeferredSucceses]
- String benchmark DONE
Improve interpreter startup time, like in patch [SF]921466? GeorgBrandl
Patch that Skip wrote for a possible implementation of str.partition() which returns a string subclass with pointers to the slices instead of a full copy. Any subsequent string ops lazily evaluate the slice into a full string. [--> RaymondHettinger]
- Unicode improvements per CCP requirements: DONE
Build-out the collections module for optimized data structures: ordered dictionary: API? Use case?! red/black? pivot tbl? skip list: contribution? radix/crit-bit trees? ChristianTismer + RichardEmslie
Struct benchmark DONE - needs to be incorporated into pybench BobIppolito
- Improve the API for timeit.py [SH to search python-dev and call Raymond Hettinger for details if necessary]
Update PythonSpeed/PerformanceTips Everyone can make suggestions SeanReifschneider + JohnBenediktsson
- Evaluate the PEPs for optimizing global and attribute lookups
- PEP 266 PEP 267 PEP 280
* Psyco improvements:
- Support for generator expressions Support for nested scopes Support for more dictionary operations Speedup float arithmetic Support for more built-ins (e.g. int(), long(), float(), etc.) Upgrade for python 2.5 Better tools for profiling psyco-ness of application Investigate usefulness of IVM (with aim to producing a more streamlined dispatch loop) LLVM backend Virtualized longs (for long longs) Virtualized slots (Ability to cache getattribute() values)
* Py3000: Make a wishlist for possible performance gains in Py3.0