Optionally use pickle5 (Redux) #370

jakirkham · 2020-05-20T05:19:59Z

Fixes #179

Thanks to @pierreglaser's work in PR ( #368 ), this is a rebased/simplified version of PR ( #364 ). Otherwise is the same in that it tries to use pickle5 on older Python versions to support out-of-band buffers.

codecov · 2020-05-20T05:36:19Z

Codecov Report

Merging #370 into master will increase coverage by 0.15%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #370      +/-   ##
==========================================
+ Coverage   91.25%   91.40%   +0.15%     
==========================================
  Files           2        3       +1     
  Lines         629      640      +11     
  Branches      132      134       +2     
==========================================
+ Hits          574      585      +11     
  Misses         34       34              
  Partials       21       21

Impacted Files	Coverage Δ
cloudpickle/cloudpickle.py	`86.44% <100.00%> (ø)`
cloudpickle/cloudpickle_fast.py	`96.64% <100.00%> (+0.01%)`	⬆️
cloudpickle/compat.py	`100.00% <100.00%> (ø)`

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 938553f...f17b31a. Read the comment docs.

jakirkham · 2020-05-20T05:37:40Z

We probably need the fix in PR ( pitrou/pickle5-backport#16 ). Asking about a patch release with that change in issue ( pitrou/pickle5-backport#17 ).

Edit: This has been released as pickle5 version 0.0.10. Have updated the requirement here accordingly.

Tries to use `pickle5` for `pickle` if available on older Python versions.

jakirkham · 2020-05-21T22:06:45Z

Alright I think I've cleared the out-of-band related errors and have managed to get Python 3.7 to pass. 🎉

There still seem to be some other errors for earlier Python versions. The errors seem to be things like pickling ABCs and Unions. Not sure how we address those. If anyone has thoughts, they would be most welcome 🙂

Has a fix for the module name `PickleBuffer` is in.

As NumPy 1.18.5 supports `pickle5` on Python 3.5 and we require that NumPy version for testing, include this out-of-band buffers test on Python 3.5 as well.

jakirkham · 2020-06-30T20:16:18Z

@pierreglaser, do you know what we need to do to run downstream CI tests here? Tried adding the label and restarting, but maybe something else is needed?

pierreglaser · 2020-06-30T20:32:05Z

pushing an empty commit should do the trick :)

This reverts commit 8178a2c.

pierreglaser · 2020-06-30T20:53:58Z

tests/cloudpickle_test.py

@@ -521,7 +521,7 @@ def test_module_locals_behavior(self):
        pickled_func_path = os.path.join(self.tmpdir, 'local_func_g.pkl')

        child_process_script = '''
-        import pickle
+        from cloudpickle.compat import pickle


using cloudpickle.compat.pickle or pickle should be equivalent in the load case right?

pierreglaser

LGTM.

A small benchmark with and without pickle5:

(cloudpickle_py37)  ~/repos/cloudpickle  (opt_use_pickle5_redux)❯_ pip show pickle5
WARNING: Package(s) not found: pickle5
(cloudpickle_py37)  ~/repos/cloudpickle  (opt_use_pickle5_redux)❯_ python -m timeit 'import cloudpickle; _ = cloudpickle.dumps(list(range(int(1e6))))'
1 loop, best of 5: 1.21 sec per loop
(cloudpickle_py37)  ~/repos/cloudpickle  (opt_use_pickle5_redux)❯_ pip install pickle5
Processing /home/pierreglaser/.cache/pip/wheels/7e/6a/00/67136a90d6aca437d806d1d3cedf98106e840c97a3e5188198/pickle5-0.0.11-cp37-cp37m-linux_x86_64.whl
Installing collected packages: pickle5
Successfully installed pickle5-0.0.11
WARNING: You are using pip version 20.0.2; however, version 20.1.1 is available.
You should consider upgrading via the '/home/pierreglaser/.virtualenvs/cloudpickle_py37/bin/python3.7 -m pip install --upgrade pip' command.
(cloudpickle_py37)  ~/repos/cloudpickle  (opt_use_pickle5_redux)❯_ python -m timeit 'import cloudpickle; _ = cloudpickle.dumps(list(range(int(1e6))))'
5 loops, best of 5: 46.7 msec per loop

cloudpickle/cloudpickle_fast.py

pierreglaser · 2020-06-30T20:56:09Z

dev-requirements.txt

 # To be able to test tornado coroutines
 tornado
 # To be able to test numpy specific things
 # but do not build numpy from source on Python nightly
-numpy; python_version <= '3.8'
+numpy >=1.18.5; python_version <= '3.8'


so we don't have to skip numpy + Python 3.5 after all? nice :)

Indeed :) PR ( numpy/numpy#16439 ) added Python 3.5 support

pierreglaser · 2020-06-30T20:56:51Z

tests/cloudpickle_test.py

@@ -1008,7 +1008,8 @@ def example():

        # choose "subprocess" rather than "multiprocessing" because the latter
        # library uses fork to preserve the parent environment.
-        command = ("import pickle, base64; "
+        command = ("import base64; "
+                   "from cloudpickle.compat import pickle; "


ditto: cloudpickle.compat.pickle and pickle are interchangeable in this situation right?

In both of these test cases, we are using the highest supported protocol with cloudpickle, which is protocol 5, to produce the pickled data. So we wound up needing to change these since we are trying to load the pickled data and then need to use pickle5 when pickle doesn't have protocol 5 support. We could alternatively skip the test, restrict the protocol based on pickle.HIGHEST_PROTOCOL, or something else. We could also leave as-is. Some options to consider 🙂

In this case I agree. Thanks.

Co-authored-by: Pierre Glaser <pierreglaser@msn.com>

cloudpickle/cloudpickle_fast.py

ogrisel

LGTM.

pierreglaser · 2020-07-01T15:49:54Z

@jakirkham I'm going to release now, feel free to voice any concern/remarks :)

jakirkham · 2020-07-01T16:43:17Z

Awesome! Thank you both 😀

No concerns from me. Was thinking the same thing in fact 😉

jakirkham · 2020-07-01T17:21:49Z

Just as an FYI, @suquark this is in cloudpickle 1.5.0

jorisvandenbossche · 2020-07-08T09:09:28Z

I encountered a failing dask test which I suppose is related to this: dask/dask#6374. It happens when you have pickle5 installed with the latest cloudpickle. The pickling roundtrip of a dask dataframe seems to no longer work (but I am not very knowledgeable about pickling/cloudpickle/pickle5, so not directly sure what the cause is / whether it's a bug).

pierreglaser · 2020-07-08T09:31:49Z

Duplicating my comment from dask/dask#6374:

The reason why this test is failing is that cp_dumps relies on pickle5 and thus uses pickle protocol 5 when dumping df. On the countrary, loads is the stdlib pickle.loads function, which, on Python 3.7 does not implement pickle protocol 5 yet, hence the error. So pickle files generated using cp_dumps when pickle5 is installed should/can only be read using pickle5.loads/cloudpickle.loads. This change was not avoidable to have pickle5 support on Python 3.6/7.

Although this is a breaking change:

pickle5 is only an optional dependency on cloudpickle
Libraries such as dask are working alongside with cloudpickle to adapt to this change (Optionally use pickle5 dask/distributed#3849)

But I'd be glad to hear your concerns.

suquark · 2020-07-08T18:59:12Z

@jakirkham thanks for notifying me about the release!

jakirkham · 2020-07-08T19:22:57Z

Of course! 😄

Also the Dask test failure should be fixed with PR ( dask/dask#6379 ).

2.0.0 ===== - Python 3.5 is no longer supported. - Support for registering modules to be serialised by value. This allows code defined in local modules to be serialised and executed remotely without those local modules installed on the remote machine. ([PR #417](cloudpipe/cloudpickle#417)) - Fix a side effect altering dynamic modules at pickling time. ([PR #426](cloudpipe/cloudpickle#426)) - Support for pickling type annotations on Python 3.10 as per [PEP 563]( https://www.python.org/dev/peps/pep-0563/) ([PR #400](cloudpipe/cloudpickle#400)) - Stricter parametrized type detection heuristics in _is_parametrized_type_hint to limit false positives. ([PR #409](cloudpipe/cloudpickle#409)) - Support pickling / depickling of OrderedDict KeysView, ValuesView, and ItemsView, following similar strategy for vanilla Python dictionaries. ([PR #423](cloudpipe/cloudpickle#423)) - Suppressed a source of non-determinism when pickling dynamically defined functions and handles the deprecation of co_lnotab in Python 3.10+. ([PR #428](cloudpipe/cloudpickle#428)) 1.6.0 ===== - `cloudpickle`'s pickle.Pickler subclass (currently defined as `cloudpickle.cloudpickle_fast.CloudPickler`) can and should now be accessed as `cloudpickle.Pickler`. This is the only officially supported way of accessing it. ([issue #366](cloudpipe/cloudpickle#366)) - `cloudpickle` now supports pickling `dict_keys`, `dict_items` and `dict_values`. ([PR #384](cloudpipe/cloudpickle#384)) 1.5.0 ===== - Fix a bug causing cloudpickle to crash when pickling dynamically created, importable modules. ([issue #360](cloudpipe/cloudpickle#354)) - Add optional dependency on `pickle5` to get improved performance on Python 3.6 and 3.7. ([PR #370](cloudpipe/cloudpickle#370)) - Internal refactoring to ease the use of `pickle5` in cloudpickle for Python 3.6 and 3.7. ([PR #368](cloudpipe/cloudpickle#368)) 1.4.1 ===== - Fix incompatibilities between cloudpickle 1.4.0 and Python 3.5.0/1/2 introduced by the new support of cloudpickle for pickling typing constructs. ([issue #360](cloudpipe/cloudpickle#360)) - Restore compat with loading dynamic classes pickled with cloudpickle version 1.2.1 that would reference the `types.ClassType` attribute. ([PR #359](cloudpipe/cloudpickle#359)) 1.4.0 ===== **This version requires Python 3.5 or later** - cloudpickle can now all pickle all constructs from the ``typing`` module and the ``typing_extensions`` library in Python 3.5+ ([PR #318](cloudpipe/cloudpickle#318)) - Stop pickling the annotations of a dynamic class for Python < 3.6 (follow up on #276) ([issue #347](cloudpipe/cloudpickle#347)) - Fix a bug affecting the pickling of dynamic `TypeVar` instances on Python 3.7+, and expand the support for pickling `TypeVar` instances (dynamic or non-dynamic) to Python 3.5-3.6 ([PR #350](cloudpipe/cloudpickle#350)) - Add support for pickling dynamic classes subclassing `typing.Generic` instances on Python 3.7+ ([PR #351](cloudpipe/cloudpickle#351)) 1.3.0 ===== - Fix a bug affecting dynamic modules occuring with modified builtins ([issue #316](cloudpipe/cloudpickle#316)) - Fix a bug affecting cloudpickle when non-modules objects are added into sys.modules ([PR #326](cloudpipe/cloudpickle#326)). - Fix a regression in cloudpickle and python3.8 causing an error when trying to pickle property objects. ([PR #329](cloudpipe/cloudpickle#329)). - Fix a bug when a thread imports a module while cloudpickle iterates over the module list ([PR #322](cloudpipe/cloudpickle#322)). - Add support for out-of-band pickling (Python 3.8 and later). https://docs.python.org/3/library/pickle.html#example ([issue #308](cloudpipe/cloudpickle#308)) - Fix a side effect that would redefine `types.ClassTypes` as `type` when importing cloudpickle. ([issue #337](cloudpipe/cloudpickle#337)) - Fix a bug affecting subclasses of slotted classes. ([issue #311](cloudpipe/cloudpickle#311)) - Dont pickle the abc cache of dynamically defined classes for Python 3.6- (This was already the case for python3.7+) ([issue #302](cloudpipe/cloudpickle#302)) 1.2.2 ===== - Revert the change introduced in ([issue #276](cloudpipe/cloudpickle#276)) attempting to pickle functions annotations for Python 3.4 to 3.6. It is not possible to pickle complex typing constructs for those versions (see [issue #193]( cloudpipe/cloudpickle#193)) - Fix a bug affecting bound classmethod saving on Python 2. ([issue #288](cloudpipe/cloudpickle#288)) - Add support for pickling "getset" descriptors ([issue #290](cloudpipe/cloudpickle#290)) 1.2.1 ===== - Restore (partial) support for Python 3.4 for downstream projects that have LTS versions that would benefit from cloudpickle bug fixes. 1.2.0 ===== - Leverage the C-accelerated Pickler new subclassing API (available in Python 3.8) in cloudpickle. This allows cloudpickle to pickle Python objects up to 30 times faster. ([issue #253](cloudpipe/cloudpickle#253)) - Support pickling of classmethod and staticmethod objects in python2. arguments. ([issue #262](cloudpipe/cloudpickle#262)) - Add support to pickle type annotations for Python 3.5 and 3.6 (pickling type annotations was already supported for Python 3.7, Python 3.4 might also work but is no longer officially supported by cloudpickle) ([issue #276](cloudpipe/cloudpickle#276)) - Internal refactoring to proactively detect dynamic functions and classes when pickling them. This refactoring also yields small performance improvements when pickling dynamic classes (~10%) ([issue #273](cloudpipe/cloudpickle#273)) 1.1.1 ===== - Minor release to fix a packaging issue (Markdown formatting of the long description rendered on pypi.org). The code itself is the same as 1.1.0. 1.1.0 ===== - Support the pickling of interactively-defined functions with positional-only arguments. ([issue #266](cloudpipe/cloudpickle#266)) - Track the provenance of dynamic classes and enums so as to preseve the usual `isinstance` relationship between pickled objects and their original class defintions. ([issue #246](cloudpipe/cloudpickle#246)) 1.0.0 ===== - Fix a bug making functions with keyword-only arguments forget the default values of these arguments after being pickled. ([issue #264](cloudpipe/cloudpickle#264)) 0.8.1 ===== - Fix a bug (already present before 0.5.3 and re-introduced in 0.8.0) affecting relative import instructions inside depickled functions ([issue #254](cloudpipe/cloudpickle#254)) 0.8.0 ===== - Add support for pickling interactively defined dataclasses. ([issue #245](cloudpipe/cloudpickle#245)) - Global variables referenced by functions pickled by cloudpickle are now unpickled in a new and isolated namespace scoped by the CloudPickler instance. This restores the (previously untested) behavior of cloudpickle prior to changes done in 0.5.4 for functions defined in the `__main__` module, and 0.6.0/1 for other dynamic functions. 0.7.0 ===== - Correctly serialize dynamically defined classes that have a `__slots__` attribute. ([issue #225](cloudpipe/cloudpickle#225)) 0.6.1 ===== - Fix regression in 0.6.0 which breaks the pickling of local function defined in a module, making it impossible to access builtins. ([issue #211](cloudpipe/cloudpickle#211)) 0.6.0 ===== - Ensure that unpickling a function defined in a dynamic module several times sequentially does not reset the values of global variables. ([issue #187](cloudpipe/cloudpickle#205)) - Restrict the ability to pickle annotations to python3.7+ ([issue #193]( cloudpipe/cloudpickle#193) and [issue #196]( cloudpipe/cloudpickle#196)) - Stop using the deprecated `imp` module under Python 3. ([issue #207](cloudpipe/cloudpickle#207)) - Fixed pickling issue with singleton types `NoneType`, `type(...)` and `type(NotImplemented)` ([issue #209](cloudpipe/cloudpickle#209)) 0.5.6 ===== - Ensure that unpickling a locally defined function that accesses the global variables of a module does not reset the values of the global variables if they are already initialized. ([issue #187](cloudpipe/cloudpickle#187)) 0.5.5 ===== - Fixed inconsistent version in `cloudpickle.__version__`. 0.5.4 ===== - Fixed a pickling issue for ABC in python3.7+ ([issue #180]( cloudpipe/cloudpickle#180)). - Fixed a bug when pickling functions in `__main__` that access global variables ([issue #187]( cloudpipe/cloudpickle#187)). 0.5.3 ===== - Fixed a crash in Python 2 when serializing non-hashable instancemethods of built-in types ([issue #144](cloudpipe/cloudpickle#144)). - itertools objects can also pickled ([PR #156](cloudpipe/cloudpickle#156)). - `logging.RootLogger` can be also pickled ([PR #160](cloudpipe/cloudpickle#160)). 0.5.2 ===== - Fixed a regression: `AttributeError` when loading pickles that hold a reference to a dynamically defined class from the `__main__` module. ([issue #131]( cloudpipe/cloudpickle#131)). - Make it possible to pickle classes and functions defined in faulty modules that raise an exception when trying to look-up their attributes by name. 0.5.1 ===== - Fixed `cloudpickle.__version__`. 0.5.0 ===== - Use `pickle.HIGHEST_PROTOCOL` by default. 0.4.4 ===== - `logging.RootLogger` can be also pickled ([PR #160](cloudpipe/cloudpickle#160)). 0.4.3 ===== - Fixed a regression: `AttributeError` when loading pickles that hold a reference to a dynamically defined class from the `__main__` module. ([issue #131]( cloudpipe/cloudpickle#131)). - Fixed a crash in Python 2 when serializing non-hashable instancemethods of built-in types. ([issue #144](cloudpipe/cloudpickle#144)) 0.4.2 ===== - Restored compatibility with pickles from 0.4.0. - Handle the `func.__qualname__` attribute. 0.4.1 ===== - Fixed a crash when pickling dynamic classes whose `__dict__` attribute was defined as a [`property`](https://docs.python.org/3/library/functions.html#property). Most notably, this affected dynamic [namedtuples](https://docs.python.org/2/library/collections.html#namedtuple-factory-function-for-tuples-with-named-fields) in Python 2. (cloudpipe/cloudpickle#113) - Cloudpickle now preserves the `__module__` attribute of functions (cloudpipe/cloudpickle#118). - Fixed a crash when pickling modules that don't have a `__package__` attribute (cloudpipe/cloudpickle#116). 0.4.0 ===== * Fix functions with empty cells * Allow pickling Logger objects * Fix crash when pickling dynamic class cycles * Ignore "None" mdoules added to sys.modules * Support WeakSets and ABCMeta instances * Remove non-standard `__transient__` support * Catch exception from `pickle.whichmodule()` 0.3.1 ===== * Fix version information and ship a changelog 0.3.0 ===== * Import submodules accessed by pickled functions * Support recursive functions inside closures * Fix `ResourceWarnings` and `DeprecationWarnings` * Assume modules with `__file__` attribute are not dynamic 0.2.2 ===== * Support Python 3.6 * Support Tornado Coroutines * Support builtin methods

pierreglaser added 10 commits May 14, 2020 10:57

backport python3.8 save_reduce into old CloudPickler

a860509

python 3.8 compat

ee722f1

use the same Cloudpickler class for all Pythons

d69ca86

increase flake8 max complexity

6264641

unused functions and imports

dc9ba84

pypy compat

d5ed4d9

be pickle5 friendly

c44ad88

duplicate top-level functions

fd67f4b

increase flake8 max-complexity

9740b22

install tornado in the CI

c657fe1

jakirkham mentioned this pull request May 20, 2020

0.0.10 release? pitrou/pickle5-backport#17

Closed

jakirkham force-pushed the opt_use_pickle5_redux branch 2 times, most recently from f356dbd to 6635ce1 Compare May 20, 2020 05:35

jakirkham mentioned this pull request May 20, 2020

deduplicate cloudpickle reducers. #368

Merged

jakirkham force-pushed the opt_use_pickle5_redux branch 8 times, most recently from 8a6cb2f to 4e66e39 Compare May 21, 2020 21:14

jakirkham added 3 commits May 21, 2020 14:40

Add compat

1f2410f

Tries to use `pickle5` for `pickle` if available on older Python versions.

Use compat to provide pickle or pickle5

ebaf7f6

Import pickle from compat in tests

1d6abbd

jakirkham force-pushed the opt_use_pickle5_redux branch from 4e66e39 to 47097e5 Compare May 21, 2020 21:40

jakirkham force-pushed the opt_use_pickle5_redux branch from 47097e5 to 85c566a Compare May 22, 2020 00:27

jakirkham added 2 commits June 30, 2020 13:03

Require pickle5 0.0.11

d588774

Has a fix for the module name `PickleBuffer` is in.

Include NumPy test on Python 3.5

7b3f1a7

As NumPy 1.18.5 supports `pickle5` on Python 3.5 and we require that NumPy version for testing, include this out-of-band buffers test on Python 3.5 as well.

jakirkham added the ci downstream Signal the CI to run the test suite of all registered cloudpickle downstream projects. label Jun 30, 2020

jakirkham changed the title ~~WIP: Optionally use pickle5 (Redux)~~ Optionally use pickle5 (Redux) Jun 30, 2020

jakirkham added 2 commits June 30, 2020 13:35

Add blank line

8178a2c

Revert "Add blank line"

dc8bd28

This reverts commit 8178a2c.

pierreglaser reviewed Jun 30, 2020

View reviewed changes

pierreglaser approved these changes Jun 30, 2020

View reviewed changes

Fix typo

8a890f5

Co-authored-by: Pierre Glaser <pierreglaser@msn.com>

pierreglaser reviewed Jul 1, 2020

View reviewed changes

cloudpickle/cloudpickle_fast.py Outdated Show resolved Hide resolved

pierreglaser added 2 commits July 1, 2020 10:09

Update cloudpickle/cloudpickle_fast.py

5cdb5e1

CI trigger

268778b

ogrisel approved these changes Jul 1, 2020

View reviewed changes

Update CHANGES.md

f17b31a

ogrisel merged commit 508cdf1 into cloudpipe:master Jul 1, 2020

jakirkham deleted the opt_use_pickle5_redux branch July 1, 2020 17:20

AnesBenmerzoug mentioned this pull request Jul 6, 2020

Fix sklearn deserialization with latest cloudpickle when pickle5 is installed mlflow/mlflow#3054

Merged

24 tasks

jorisvandenbossche mentioned this pull request Jul 8, 2020

DataFrame pickling test fails with latest cloudpickle when pickle5 is installed dask/dask#6374

Closed

yannikschaelte mentioned this pull request Nov 27, 2020

cloudpickle 1.5.0 breaks on python 3.6+7 without pickle5 #401

Closed

lesteve mentioned this pull request Sep 7, 2022

Fix import of pickle module on PYPY 3.8 (#455) #461

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Optionally use pickle5 (Redux) #370

Optionally use pickle5 (Redux) #370

jakirkham commented May 20, 2020 •

edited

Loading

codecov bot commented May 20, 2020 •

edited

Loading

jakirkham commented May 20, 2020 •

edited

Loading

jakirkham commented May 21, 2020

jakirkham commented Jun 30, 2020

pierreglaser commented Jun 30, 2020

pierreglaser Jun 30, 2020

pierreglaser left a comment

pierreglaser Jun 30, 2020

jakirkham Jun 30, 2020

pierreglaser Jun 30, 2020

jakirkham Jun 30, 2020

pierreglaser Jul 1, 2020

ogrisel left a comment

pierreglaser commented Jul 1, 2020

jakirkham commented Jul 1, 2020

jakirkham commented Jul 1, 2020

jorisvandenbossche commented Jul 8, 2020

pierreglaser commented Jul 8, 2020

suquark commented Jul 8, 2020

jakirkham commented Jul 8, 2020

Optionally use pickle5 (Redux) #370

Optionally use pickle5 (Redux) #370

Conversation

jakirkham commented May 20, 2020 • edited Loading

codecov bot commented May 20, 2020 • edited Loading

Codecov Report

jakirkham commented May 20, 2020 • edited Loading

jakirkham commented May 21, 2020

jakirkham commented Jun 30, 2020

pierreglaser commented Jun 30, 2020

pierreglaser Jun 30, 2020

Choose a reason for hiding this comment

pierreglaser left a comment

Choose a reason for hiding this comment

pierreglaser Jun 30, 2020

Choose a reason for hiding this comment

jakirkham Jun 30, 2020

Choose a reason for hiding this comment

pierreglaser Jun 30, 2020

Choose a reason for hiding this comment

jakirkham Jun 30, 2020

Choose a reason for hiding this comment

pierreglaser Jul 1, 2020

Choose a reason for hiding this comment

ogrisel left a comment

Choose a reason for hiding this comment

pierreglaser commented Jul 1, 2020

jakirkham commented Jul 1, 2020

jakirkham commented Jul 1, 2020

jorisvandenbossche commented Jul 8, 2020

pierreglaser commented Jul 8, 2020

suquark commented Jul 8, 2020

jakirkham commented Jul 8, 2020

jakirkham commented May 20, 2020 •

edited

Loading

codecov bot commented May 20, 2020 •

edited

Loading

jakirkham commented May 20, 2020 •

edited

Loading