---
- branch: MAIN
  date: Mon Nov 14 14:31:18 UTC 2016
  files:
  - new: '1.23'
    old: '1.22'
    path: pkgsrc/devel/py-cffi/Makefile
    pathrev: pkgsrc/devel/py-cffi/Makefile@1.23
    type: modified
  - new: '1.22'
    old: '1.21'
    path: pkgsrc/devel/py-cffi/distinfo
    pathrev: pkgsrc/devel/py-cffi/distinfo@1.22
    type: modified
  - new: '1.2'
    old: '1.1'
    path: pkgsrc/devel/py-cffi/patches/patch-c_malloc__closure.h
    pathrev: pkgsrc/devel/py-cffi/patches/patch-c_malloc__closure.h@1.2
    type: modified
  id: 20161114T143118Z.84b941dfa30aec79f282f1dc7d050093ceb22743
  log: "Updated py-cffi to 1.9.1.\n\nv1.9\n\n    Structs with variable-sized arrays
    as their last field: now we\n    track the length of the array after ffi.new()
    is called, just\n    like we always tracked the length of ffi.new(\"int[]\", 42).
    This\n    lets us detect out-of-range accesses to array items. This also\n    lets
    us display a better repr(), and have the total size returned\n    by ffi.sizeof()
    and ffi.buffer(). Previously both functions\n    would return a result based on
    the size of the declared structure\n    type, with an assumed empty array. (Thanks
    andrew for starting\n    this refactoring.)\n    Add support in cdef()/set_source()
    for unspecified-length arrays\n    in typedefs: typedef int foo_t[...];. It was
    already supported\n    for global variables or structure fields.\n    I turned
    in v1.8 a warning from cffi/model.py into an error:\n    'enum xxx' has no values
    explicitly defined: refusing to guess\n    which integer type it is meant to be
    (unsigned/signed, int/long).\n    Now Iç©\x83 turning it back to a warning again;
    it seems that\n    guessing that the enum has size int is a 99%-safe bet. (But\n
    \   not 100%, so it stays as a warning.)\n    Fix leaks in the code handling FILE
    * arguments. In CPython 3\n    there is a remaining issue that is hard to fix:
    if you pass a\n    Python file object to a FILE * argument, then os.dup() is used\n
    \   and the new file descriptor is only closed when the GC reclaims\n    the Python
    file object秣nd not at the earlier time when you\n    call close(), which only
    closes the original file descriptor.\n    If this is an issue, you should avoid
    this automatic convertion\n    of Python file objects: instead, explicitly manipulate
    file\n    descriptors and call fdopen() from C (...via cffi).\n"
  module: pkgsrc
  subject: 'CVS commit: pkgsrc/devel/py-cffi'
  unixtime: '1479133878'
  user: wiz