HMAC.__init__(). Adapted from SF patch 895445 "hmac.HMAC.copy() speedup"
by Trevor Perrin, who reported that this approach increased throughput
of his hmac-intensive app by 30%.
> ----------------------------
> revision 1.20.4.4
> date: 2003/06/12 09:14:17; author: anthonybaxter; state: Exp; lines: +13 -6
> preamble is None when missing, not ''.
> Handle a couple of bogus formatted messages - now parses my main testsuite.
> Handle message/external-body.
> ----------------------------
> revision 1.20.4.3
> date: 2003/06/12 07:16:40; author: anthonybaxter; state: Exp; lines: +6 -4
> epilogue-processing is now the same as the old parser - the newline at the
> end of the line with the --endboundary-- is included as part of the epilogue.
> Note that any whitespace after the boundary is _not_ part of the epilogue.
> ----------------------------
> revision 1.20.4.2
> date: 2003/06/12 06:39:09; author: anthonybaxter; state: Exp; lines: +6 -4
> message/delivery-status fixed.
> HeaderParser fixed.
> ----------------------------
> revision 1.20.4.1
> date: 2003/06/12 06:08:56; author: anthonybaxter; state: Exp; lines: +163 -129
> A work-in-progress snapshot of the new parser. A couple of known problems:
>
> - first (blank) line of MIME epilogues is being consumed
> - message/delivery-status isn't quite right
>
> It still needs a lot of cleanup, but right now it parses a whole lot of
> badness that the old parser failed on. I also need to think about adding
> back the old 'strict' flag in some way.
> =============================================================================
* The default __reversed__ performed badly, so reintroduced a custom
reverse iterator.
* Added length transparency to improve speed with map(), list(), etc.
This gives another 30% speedup for operations such as
map(func, d.iteritems()) or list(d.iteritems()) which can both take
advantage of length information when provided.
* Split into three separate types that share everything except the
code for iternext. Saves run time decision making and allows
each iternext function to be specialized.
* Inlined PyDict_Next(). In addition to saving a function call, this
allows a redundant test to be eliminated and further specialization
of the code for the unique needs of each iterator type.
* Created a reusable result tuple for iteritems(). Saves the malloc
time for tuples when the previous result was not kept by client code
(this is the typical use case for iteritems). If the client code
does keep the reference, then a new tuple is created.
Results in a 20% to 30% speedup depending on the size and sparsity
of the dictionary.
* Factored constant structure references out of the inner loops for
PyDict_Next(), dict_keys(), dict_values(), and dict_items().
Gave measurable speedups to each (the improvement varies depending
on the sparseness of the dictionary being measured).
* Added a freelist scheme styled after that for tuples. Saves around
80% of the calls to malloc and free. About 10% of the time, the
previous dictionary was completely empty; in those cases, the
dictionary initialization with memset() can be skipped.
scheme in situations that likely won't benefit from it. This further
improves memory utilization from Py2.3 which always over-allocates
except for PyList_New().
Situations expected to benefit from over-allocation:
list.insert(), list.pop(), list.append(), and list.extend()
Situations deemed unlikely to benefit:
list_inplace_repeat, list_ass_slice, list_ass_subscript
The most gray area was for listextend_internal() which only runs
when the argument is a list or a tuple. This could be viewed as
a one-time fixed length addition or it could be viewed as wrapping
a series of appends. I left its over-allocation turned on but
could be convinced otherwise.
array.extend() now accepts iterable arguments implements as a series
of appends. Besides being a user convenience and matching the behavior
for lists, this the saves memory and cycles that would be used to
create a temporary array object.
lists. Speeds append() operations and reduces memory requirements
(because of more conservative overallocation).
Paves the way for the feature request for array.extend() to support
arbitrary iterable arguments.
- there were no accessor functions for the global per-database fields
- packages and their dependencies were installed in order in stead
of in reverse order.
Brett Cannon
Armin Rigo
Guido van Rossum
Thomas Wouters
Tim Peters
Raymond Hettinger
Fred Drake
Skip Montanaro
Hye-Shik Chang
Gregory P. Smith
Kurt B. Kaiser
Walter Dörwald
Jack Jansen