summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorArmin Rigo <arigo@tunes.org>2013-12-08 12:28:39 +0100
committerArmin Rigo <arigo@tunes.org>2013-12-08 12:28:39 +0100
commit9805748b1af1ee029f207eb6fbb9921d564d4022 (patch)
treebccf020b66447222e9043fd998fea4924cc0f5b6
parent995ca309e3580b7b382a70d85f37c11de8b25116 (diff)
downloadcffi-9805748b1af1ee029f207eb6fbb9921d564d4022.tar.gz
Simplify a bit this paragraph.
-rw-r--r--doc/source/index.rst11
1 files changed, 4 insertions, 7 deletions
diff --git a/doc/source/index.rst b/doc/source/index.rst
index 32f0209..9fbd5ae 100644
--- a/doc/source/index.rst
+++ b/doc/source/index.rst
@@ -1272,13 +1272,10 @@ Note that ``from_handle()`` conceptually works like this: it searches in
the list of cdata objects made by ``new_handle()`` the one which has got
the same ``void *`` value; and then it fetches in that cdata object the
corresponding Python object. The cdata object keeps the Python object
-alive, but if the cdata object *itself* is not alive any more, then it
-will crash! (Obviously, the real
-implementation is more efficient than suggested here.) In other words,
-the result of ``new_handle()`` has *ownership* (similarly to
-``ffi.new()`` or ``ffi.gc()``) in the sense that the association ``void
-* -> python_object`` is only valid as long as *this* exact cdata
-returned by ``new_handle()`` is alive.
+alive, similar to how ``ffi.new()`` returns a cdata object that keeps a
+piece of memory alive. If the cdata object *itself* is not alive any
+more, then the association ``void * -> python_object`` is dead and
+``from_handle()`` will crash.
.. "versionadded:: 0.7" --- inlined in the previous paragraph