| 1 |
|
| 2 |
|
| 3 |
About |
| 4 |
===== |
| 5 |
|
| 6 |
_decimal.c is a wrapper for the libmpdec library. libmpdec is a fast C |
| 7 |
library for correctly-rounded arbitrary precision decimal floating point |
| 8 |
arithmetic. It is a complete implementation of Mike Cowlishaw/IBM's |
| 9 |
General Decimal Arithmetic Specification. |
| 10 |
|
| 11 |
|
| 12 |
Build process for the module |
| 13 |
============================ |
| 14 |
|
| 15 |
As usual, the build process for _decimal.so is driven by setup.py in the top |
| 16 |
level directory. setup.py autodetects the following build configurations: |
| 17 |
|
| 18 |
1) x64 - 64-bit Python, x86_64 processor (AMD, Intel) |
| 19 |
|
| 20 |
2) uint128 - 64-bit Python, compiler provides __uint128_t (gcc) |
| 21 |
|
| 22 |
3) ansi64 - 64-bit Python, ANSI C |
| 23 |
|
| 24 |
4) ppro - 32-bit Python, x86 CPU, PentiumPro or later |
| 25 |
|
| 26 |
5) ansi32 - 32-bit Python, ANSI C |
| 27 |
|
| 28 |
6) ansi-legacy - 32-bit Python, compiler without uint64_t |
| 29 |
|
| 30 |
7) universal - Mac OS only (multi-arch) |
| 31 |
|
| 32 |
|
| 33 |
It is possible to override autodetection by exporting: |
| 34 |
|
| 35 |
PYTHON_DECIMAL_WITH_MACHINE=value, where value is one of the above options. |
| 36 |
|
| 37 |
|
| 38 |
NOTE |
| 39 |
==== |
| 40 |
|
| 41 |
decimal.so is not built from a static libmpdec.a since doing so led to |
| 42 |
failures on AIX (user report) and Windows (mixing static and dynamic CRTs |
| 43 |
causes locale problems and more). |
| 44 |
|
| 45 |
|
| 46 |
|