René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Remove remnants of support for non-IEEE 754 systems from cmathmodule.c","articleBody":"# Feature or enhancement\n\n### Proposal:\n\nProposed patch:\n```diff\ndiff --git a/Modules/cmathmodule.c b/Modules/cmathmodule.c\nindex bf86a211bc..591442334e 100644\n--- a/Modules/cmathmodule.c\n+++ b/Modules/cmathmodule.c\n@@ -185,15 +185,8 @@ cmath_acos_impl(PyObject *module, Py_complex z)\n     if (fabs(z.real) \u003e CM_LARGE_DOUBLE || fabs(z.imag) \u003e CM_LARGE_DOUBLE) {\n         /* avoid unnecessary overflow for large arguments */\n         r.real = atan2(fabs(z.imag), z.real);\n-        /* split into cases to make sure that the branch cut has the\n-           correct continuity on systems with unsigned zeros */\n-        if (z.real \u003c 0.) {\n-            r.imag = -copysign(log(hypot(z.real/2., z.imag/2.)) +\n-                               M_LN2*2., z.imag);\n-        } else {\n-            r.imag = copysign(log(hypot(z.real/2., z.imag/2.)) +\n-                              M_LN2*2., -z.imag);\n-        }\n+        r.imag = -copysign(log(hypot(z.real/2., z.imag/2.)) +\n+                           M_LN2*2., z.imag);\n     } else {\n         s1.real = 1.-z.real;\n         s1.imag = -z.imag;\n@@ -386,11 +379,7 @@ cmath_atanh_impl(PyObject *module, Py_complex z)\n         */\n         h = hypot(z.real/2., z.imag/2.);  /* safe from overflow */\n         r.real = z.real/4./h/h;\n-        /* the two negations in the next line cancel each other out\n-           except when working with unsigned zeros: they're there to\n-           ensure that the branch cut has the correct continuity on\n-           systems that don't support signed zeros */\n-        r.imag = -copysign(Py_MATH_PI/2., -z.imag);\n+        r.imag = copysign(Py_MATH_PI/2., z.imag);\n         errno = 0;\n     } else if (z.real == 1. \u0026\u0026 ay \u003c CM_SQRT_DBL_MIN) {\n         /* C99 standard says:  atanh(1+/-0.) should be inf +/- 0i */\n```\n\nRemoving this legacy was [proposed](https://github.com/python/cpython/pull/102067#issuecomment-1437850117) in #102067, but *before* merging #31790.  Maybe now that change does make sense?\n\n\n### Has this already been discussed elsewhere?\n\nNo response given\n\n### Links to previous discussion of this feature:\n\nhttps://github.com/python/cpython/pull/102067#issuecomment-1437850117\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-122716\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/skirpichev","@type":"Person","name":"skirpichev"},"datePublished":"2024-07-02T13:57:28.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":3},"url":"https://github.com/121268/cpython/issues/121268"}