bpo-45584: Clarify math.trunc documentation (GH-29183)

While floor/ceil 's documentation are very precise, `truncate` was not explained. I actually had to search online to understand the difference between `truncate` and `floor` (admittedly, once I remembered that numbers are signed, and that floating numbers actually uses a bit for negation symbol instead of two complement, it became obvious)

Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
Co-authored-by: Éric Araujo <merwok@netwok.org>
Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Co-authored-by: Jelle Zijlstra <jelle.zijlstra@gmail.com>

Doc/library/math.rst

@@ -32,8 +32,8 @@ Number-theoretic and representation functions
 .. function:: ceil(x)
 
    Return the ceiling of *x*, the smallest integer greater than or equal to *x*.
-   If *x* is not a float, delegates to ``x.__ceil__()``, which should return an
-   :class:`~numbers.Integral` value.
+   If *x* is not a float, delegates to :meth:`x.__ceil__ <object.__ceil__>`,
+   which should return an :class:`~numbers.Integral` value.
 
 
 .. function:: comb(n, k)
@@ -77,9 +77,9 @@ Number-theoretic and representation functions
 
 .. function:: floor(x)
 
-   Return the floor of *x*, the largest integer less than or equal to *x*.
-   If *x* is not a float, delegates to ``x.__floor__()``, which should return an
-   :class:`~numbers.Integral` value.
+   Return the floor of *x*, the largest integer less than or equal to *x*.  If
+   *x* is not a float, delegates to :meth:`x.__floor__ <object.__floor__>`, which
+   should return an :class:`~numbers.Integral` value.
 
 
 .. function:: fmod(x, y)
@@ -298,9 +298,11 @@ Number-theoretic and representation functions
 
 .. function:: trunc(x)
 
-   Return the :class:`~numbers.Real` value *x* truncated to an
-   :class:`~numbers.Integral` (usually an integer). Delegates to
-   :meth:`x.__trunc__() <object.__trunc__>`.
+   Return *x* with the fractional part
+   removed, leaving the integer part.  This rounds toward 0: ``trunc()`` is
+   equivalent to :func:`floor` for positive *x*, and equivalent to :func:`ceil`
+   for negative *x*. If *x* is not a float, delegates to :meth:`x.__trunc__
+   <object.__trunc__>`, which should return an :class:`~numbers.Integral` value.
 
 .. function:: ulp(x)