[3.12] GH-95079: document error behaviour for some unicode C APIs (GH-95080) (#124662)

(cherry picked from commit b79a21e)

Co-authored-by: Max Bachmann <kontakt@maxbachmann.de>

Author: miss-islington

Doc/c-api/unicode.rst

@@ -345,6 +345,8 @@ APIs:
    This is the recommended way to allocate a new Unicode object.  Objects
    created using this function are not resizable.
 
+   On error, set an exception and return ``NULL``.
+
    .. versionadded:: 3.3
 
 
@@ -598,6 +600,8 @@ APIs:
 
    Return the length of the Unicode object, in code points.
 
+   On error, set an exception and return ``-1``.
+
    .. versionadded:: 3.3
 
 
@@ -641,6 +645,8 @@ APIs:
    not out of bounds, and that the object can be modified safely (i.e. that it
    its reference count is one).
 
+   Return ``0`` on success, ``-1`` on error with an exception set.
+
    .. versionadded:: 3.3
 
 
@@ -650,6 +656,8 @@ APIs:
    Unicode object and the index is not out of bounds, in contrast to
    :c:func:`PyUnicode_READ_CHAR`, which performs no error checking.
 
+   Return character on success, ``-1`` on error with an exception set.
+
    .. versionadded:: 3.3
 
 
@@ -658,6 +666,7 @@ APIs:
 
    Return a substring of *unicode*, from character index *start* (included) to
    character index *end* (excluded).  Negative indices are not supported.
+   On error, set an exception and return ``NULL``.
 
    .. versionadded:: 3.3