Staging
v0.8.1
v0.8.1
https://github.com/python/cpython
Revision 7c2a85b2d44851c2442ade579b760f86447bf848 authored by Tim Peters on 31 August 2004, 02:19:55 UTC, committed by Tim Peters on 31 August 2004, 02:19:55 UTC
ldexp. Both methods are exact, and return the same results. Turns out
multiplication is a few (but just a few) percent faster on my box.
They're both significantly faster than using struct with a Q format
to convert bytes to a 64-bit long (struct.unpack() appears to lose due
to the tuple creation/teardown overhead), and calling _hexlify is
significantly faster than doing bytes.encode('hex'). So we appear to
have hit a local minimum (wrt speed) here.
1 parent 3fa19d7
Tip revision: 7c2a85b2d44851c2442ade579b760f86447bf848 authored by Tim Peters on 31 August 2004, 02:19:55 UTC
HardwareRandom: Go back to multiplying by 2**-BPF instead of using
HardwareRandom: Go back to multiplying by 2**-BPF instead of using
Tip revision: 7c2a85b
| File | Mode | Size |
|---|---|---|
| VC6 | ||
| bdist_wininst | ||
| example_nt | ||
| os2emx | ||
| os2vacpp | ||
| .cvsignore | -rw-r--r-- | 30 bytes |
| WinMain.c | -rw-r--r-- | 447 bytes |
| _winreg.c | -rw-r--r-- | 42.7 KB |
| config.c | -rw-r--r-- | 4.1 KB |
| dl_nt.c | -rw-r--r-- | 908 bytes |
| dllbase_nt.txt | -rw-r--r-- | 3.5 KB |
| frozen_dllmain.c | -rw-r--r-- | 3.6 KB |
| getpathp.c | -rw-r--r-- | 18.0 KB |
| icons.mak | -rw-r--r-- | 213 bytes |
| icons.rc | -rw-r--r-- | 59 bytes |
| import_nt.c | -rw-r--r-- | 2.4 KB |
| make_versioninfo.c | -rw-r--r-- | 1.0 KB |
| msvcrtmodule.c | -rwxr-xr-x | 4.9 KB |
| py.ico | -rw-r--r-- | 766 bytes |
| pyc.ico | -rw-r--r-- | 766 bytes |
| pycon.ico | -rw-r--r-- | 766 bytes |
| pyconfig.h | -rw-r--r-- | 15.5 KB |
| python.mk | -rw-r--r-- | 234 bytes |
| python_exe.rc | -rw-r--r-- | 49 bytes |
| python_nt.rc | -rw-r--r-- | 1.8 KB |
| readme.txt | -rw-r--r-- | 4.2 KB |
| testpy.py | -rw-r--r-- | 890 bytes |
| w9xpopen.c | -rw-r--r-- | 3.6 KB |
| winsound.c | -rw-r--r-- | 7.1 KB |
Computing file changes ...
