Issue #1580: use short float repr where possible.

- incorporate and adapt David Gay's dtoa and strtod
   into the Python core
 - on platforms where we can use Gay's code (almost
   all!), repr(float) is based on the shortest
   sequence of decimal digits that rounds correctly.
 - add sys.float_repr_style attribute to indicate
   whether we're using Gay's code or not
 - add autoconf magic to detect and enable SSE2
   instructions on x86/gcc
 - slight change to repr and str:  repr switches
   to exponential notation at 1e16 instead of
   1e17, str switches at 1e11 instead of 1e12

Python/sysmodule.c

@@ -1025,6 +1025,7 @@ platform -- platform identifier\n\
 executable -- pathname of this Python interpreter\n\
 prefix -- prefix used to find the Python library\n\
 exec_prefix -- prefix used to find the machine-specific Python library\n\
+float_repr_style -- string indicating the style of repr() output for floats\n\
 "
 )
 #ifdef MS_WINDOWS
@@ -1428,6 +1429,15 @@ _PySys_Init(void)
 	FlagsType.tp_init = NULL;
 	FlagsType.tp_new = NULL;
 
+	/* float repr style: 0.03 (short) vs 0.029999999999999999 (legacy) */
+#ifndef PY_NO_SHORT_FLOAT_REPR
+	SET_SYS_FROM_STRING("float_repr_style",
+			    PyUnicode_FromString("short"));
+#else
+	SET_SYS_FROM_STRING("float_repr_style",
+			    PyUnicode_FromString("legacy"));
+#endif
+
 #undef SET_SYS_FROM_STRING
 	if (PyErr_Occurred())
 		return NULL;