Use Python's LRU cache instead of custom cache implementation

[ralittles]

Switched over to using Python's built-in LRU cache functions for memoization

[ralittles]

Unit test results:

 ◰³ venv  ~/mike_alexander_gitlab/Version1fromGithub/VersionOne.SDK.Python   use-lru-cache-instead-of-custom-cache *…  python setup.py test   Thu 13 Dec 2018 02:50:14 PM PST
running test
running egg_info
writing top-level names to v1pysdk.egg-info/top_level.txt
writing requirements to v1pysdk.egg-info/requires.txt
writing dependency_links to v1pysdk.egg-info/dependency_links.txt
writing v1pysdk.egg-info/PKG-INFO
reading manifest file 'v1pysdk.egg-info/SOURCES.txt'
reading manifest template 'MANIFEST.in'
warning: no files found matching 'product.json'
writing manifest file 'v1pysdk.egg-info/SOURCES.txt'
running build_ext
test_create_story (tests.creation_tests.TestV1Create)
tests.creation_tests.TestV1Create.test_create_story ... ok
test_connect (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_connect ... ok
test_connect_fails_when_invalid (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_connect_fails_when_invalid ... ok
test_meta_connect_instance_and_address (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_meta_connect_instance_and_address ... ok
test_meta_connect_instance_url (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_meta_connect_instance_url ... ok
test_meta_connect_instance_url_overrides_separate (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_meta_connect_instance_url_overrides_separate ... ok
test_meta_connect_oauth (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_meta_connect_oauth ... ok
test_meta_connect_oauth_ignores_username (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_meta_connect_oauth_ignores_username ... ok
test_reconnect_fails_when_invalid (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_reconnect_fails_when_invalid ... ok
test_reconnect_succeeds_after_invalid (tests.connect_tests.TestV1Connection)
tests.connect_tests.TestV1Connection.test_reconnect_succeeds_after_invalid ... ok
test_find_query (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_find_query ... ok
test_non_default_query (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_non_default_query ... ok
test_select_epic (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_select_epic ... ok
test_select_scope (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_select_scope ... ok
test_select_story (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_select_story ... ok
test_select_story_as_generic_asset (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_select_story_as_generic_asset ... ok
test_select_task (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_select_task ... ok
test_sum_query (tests.query_tests.TestV1Query)
tests.query_tests.TestV1Query.test_sum_query ... ok
test_split_attribute (tests.string_utils_tests.TestStringUtils)
tests.string_utils_tests.TestStringUtils.test_split_attribute ... ok
test_quick_close_and_reopen (tests.operation_tests.TestV1Operations)
tests.operation_tests.TestV1Operations.test_quick_close_and_reopen ... ok

----------------------------------------------------------------------
Ran 20 tests in 57.362s

OK
 ◰³ venv  ~/mike_a

[mtalexan]

@ralittles Does this work under Python 2.7 as well as Python 3.x?

[ralittles]

@mtalexan No, the lru_cache function is only available in the Python 3 version of functools.

I tried to use the Python 2 repoze.lru library as instructed here (https://stackoverflow.com/questions/11815873/memoization-library-for-python-2-7), but the syntax was too different to use it as a drop-in replacement (at least when handling the clear cache part -- see here: http://docs.repoze.org/lru/narr.html#cleaning-cache-of-decorated-function).

Do you think we could drop support for Python 2 at this point? Otherwise the code is going to require several if/else statements for version checks to make sure the right syntax gets processed.