X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/130b8d3147af5240ba3118c56b0aa1f19d2d5c59..c3f49b76d173818386f5c65db46b353e1d334d1e:/sdk/python/tests/test_collections.py diff --git a/sdk/python/tests/test_collections.py b/sdk/python/tests/test_collections.py index 5f86fa5fad..bc3c0bf1af 100644 --- a/sdk/python/tests/test_collections.py +++ b/sdk/python/tests/test_collections.py @@ -4,7 +4,6 @@ import arvados import copy -import hashlib import mock import os import pprint @@ -13,10 +12,9 @@ import tempfile import unittest import run_test_server +from arvados._ranges import Range, LocatorAndRange +from arvados.collection import Collection, CollectionReader import arvados_testutil as tutil -from arvados.ranges import Range, LocatorAndRange -from arvados import import_manifest, export_manifest -from arvados.arvfile import SYNC_EXPLICIT class TestResumableWriter(arvados.ResumableCollectionWriter): KEEP_BLOCK_SIZE = 1024 # PUT to Keep every 1K. @@ -396,7 +394,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, def test_write_directory_tree_with_zero_recursion(self): cwriter = arvados.CollectionWriter(self.api_client) content = 'd1/d2/f3d1/f2f1' - blockhash = hashlib.md5(content).hexdigest() + '+' + str(len(content)) + blockhash = tutil.str_keep_locator(content) cwriter.write_directory_tree( self.build_directory_tree(['f1', 'd1/f2', 'd1/d2/f3']), max_manifest_depth=0) @@ -434,7 +432,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, with self.make_test_file() as testfile: cwriter.write_file(testfile.name, 'test') resumed = TestResumableWriter.from_state(cwriter.current_state()) - self.assertEquals(cwriter.manifest_text(), resumed.manifest_text(), + self.assertEqual(cwriter.manifest_text(), resumed.manifest_text(), "resumed CollectionWriter had different manifest") def test_resume_fails_when_missing_dependency(self): @@ -503,17 +501,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, ).manifest_text()) -class CollectionTestMixin(object): - PROXY_RESPONSE = { - 'items_available': 1, - 'items': [{ - 'uuid': 'zzzzz-bi6l4-mockproxy012345', - 'owner_uuid': 'zzzzz-tpzed-mockowner012345', - 'service_host': tutil.TEST_HOST, - 'service_port': 65535, - 'service_ssl_flag': True, - 'service_type': 'proxy', - }]} +class CollectionTestMixin(tutil.ApiClientMock): API_COLLECTIONS = run_test_server.fixture('collections') DEFAULT_COLLECTION = API_COLLECTIONS['foo_file'] DEFAULT_DATA_HASH = DEFAULT_COLLECTION['portable_data_hash'] @@ -523,20 +511,9 @@ class CollectionTestMixin(object): ALT_DATA_HASH = ALT_COLLECTION['portable_data_hash'] ALT_MANIFEST = ALT_COLLECTION['manifest_text'] - def _mock_api_call(self, mock_method, code, body): - mock_method = mock_method().execute - if code == 200: - mock_method.return_value = body - else: - mock_method.side_effect = arvados.errors.ApiError( - tutil.fake_httplib2_response(code), "{}") - - def mock_keep_services(self, api_mock, code, body): - self._mock_api_call(api_mock.keep_services().accessible, code, body) - - def api_client_mock(self, code=200): - client = mock.MagicMock(name='api_client') - self.mock_keep_services(client, code, self.PROXY_RESPONSE) + def api_client_mock(self, status=200): + client = super(CollectionTestMixin, self).api_client_mock() + self.mock_keep_services(client, status=status, service_type='proxy', count=1) return client @@ -546,9 +523,9 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): body = self.API_COLLECTIONS.get(body) self._mock_api_call(api_mock.collections().get, code, body) - def api_client_mock(self, code=200): - client = super(CollectionReaderTestCase, self).api_client_mock(code) - self.mock_get_collection(client, code, 'foo_file') + def api_client_mock(self, status=200): + client = super(CollectionReaderTestCase, self).api_client_mock() + self.mock_get_collection(client, status, 'foo_file') return client def test_init_no_default_retries(self): @@ -567,14 +544,13 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def test_uuid_init_failure_raises_api_error(self): client = self.api_client_mock(500) - reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) with self.assertRaises(arvados.errors.ApiError): - reader.manifest_text() + reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) def test_locator_init(self): client = self.api_client_mock(200) # Ensure Keep will not return anything if asked. - with tutil.mock_get_responses(None, 404): + with tutil.mock_keep_responses(None, 404): reader = arvados.CollectionReader(self.DEFAULT_DATA_HASH, api_client=client) self.assertEqual(self.DEFAULT_MANIFEST, reader.manifest_text()) @@ -584,7 +560,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): # been written to Keep. client = self.api_client_mock(200) self.mock_get_collection(client, 404, None) - with tutil.mock_get_responses(self.DEFAULT_MANIFEST, 200): + with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): reader = arvados.CollectionReader(self.DEFAULT_DATA_HASH, api_client=client) self.assertEqual(self.DEFAULT_MANIFEST, reader.manifest_text()) @@ -592,17 +568,16 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def test_uuid_init_no_fallback_to_keep(self): # Do not look up a collection UUID in Keep. client = self.api_client_mock(404) - reader = arvados.CollectionReader(self.DEFAULT_UUID, - api_client=client) - with tutil.mock_get_responses(self.DEFAULT_MANIFEST, 200): + with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): with self.assertRaises(arvados.errors.ApiError): - reader.manifest_text() + reader = arvados.CollectionReader(self.DEFAULT_UUID, + api_client=client) def test_try_keep_first_if_permission_hint(self): # To verify that CollectionReader tries Keep first here, we # mock API server to return the wrong data. client = self.api_client_mock(200) - with tutil.mock_get_responses(self.ALT_MANIFEST, 200): + with tutil.mock_keep_responses(self.ALT_MANIFEST, 200): self.assertEqual( self.ALT_MANIFEST, arvados.CollectionReader( @@ -614,7 +589,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): client = self.api_client_mock(200) reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client, num_retries=3) - with tutil.mock_get_responses('foo', 500, 500, 200): + with tutil.mock_keep_responses('foo', 500, 500, 200): self.assertEqual('foo', ''.join(f.read(9) for f in reader.all_files())) @@ -654,7 +629,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def test_api_response_with_collection_from_keep(self): client = self.api_client_mock() self.mock_get_collection(client, 404, 'foo') - with tutil.mock_get_responses(self.DEFAULT_MANIFEST, 200): + with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): reader = arvados.CollectionReader(self.DEFAULT_DATA_HASH, api_client=client) api_response = reader.api_response() @@ -663,7 +638,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def check_open_file(self, coll_file, stream_name, file_name, file_size): self.assertFalse(coll_file.closed, "returned file is not open") self.assertEqual(stream_name, coll_file.stream_name()) - self.assertEqual(file_name, coll_file.name()) + self.assertEqual(file_name, coll_file.name) self.assertEqual(file_size, coll_file.size()) def test_open_collection_file_one_argument(self): @@ -672,12 +647,6 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): cfile = reader.open('./foo') self.check_open_file(cfile, '.', 'foo', 3) - def test_open_collection_file_two_arguments(self): - client = self.api_client_mock(200) - reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) - cfile = reader.open('.', 'foo') - self.check_open_file(cfile, '.', 'foo', 3) - def test_open_deep_file(self): coll_name = 'collection_with_files_in_subdir' client = self.api_client_mock(200) @@ -691,23 +660,23 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def test_open_nonexistent_stream(self): client = self.api_client_mock(200) reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) - self.assertRaises(ValueError, reader.open, './nonexistent', 'foo') + self.assertRaises(IOError, reader.open, './nonexistent/foo') def test_open_nonexistent_file(self): client = self.api_client_mock(200) reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) - self.assertRaises(ValueError, reader.open, '.', 'nonexistent') + self.assertRaises(IOError, reader.open, 'nonexistent') @tutil.skip_sleep class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): def mock_keep(self, body, *codes, **headers): headers.setdefault('x-keep-replicas-stored', 2) - return tutil.mock_put_responses(body, *codes, **headers) + return tutil.mock_keep_responses(body, *codes, **headers) def foo_writer(self, **kwargs): - api_client = self.api_client_mock() - writer = arvados.CollectionWriter(api_client, **kwargs) + kwargs.setdefault('api_client', self.api_client_mock()) + writer = arvados.CollectionWriter(**kwargs) writer.start_new_file('foo') writer.write('foo') return writer @@ -723,6 +692,32 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): with self.assertRaises(arvados.errors.KeepWriteError): writer.finish() + def test_write_insufficient_replicas_via_proxy(self): + writer = self.foo_writer(replication=3) + with self.mock_keep(None, 200, **{'x-keep-replicas-stored': 2}): + with self.assertRaises(arvados.errors.KeepWriteError): + writer.manifest_text() + + def test_write_insufficient_replicas_via_disks(self): + client = mock.MagicMock(name='api_client') + with self.mock_keep( + None, 200, 200, + **{'x-keep-replicas-stored': 1}) as keepmock: + self.mock_keep_services(client, status=200, service_type='disk', count=2) + writer = self.foo_writer(api_client=client, replication=3) + with self.assertRaises(arvados.errors.KeepWriteError): + writer.manifest_text() + + def test_write_three_replicas(self): + client = mock.MagicMock(name='api_client') + with self.mock_keep( + "", 500, 500, 500, 200, 200, 200, + **{'x-keep-replicas-stored': 1}) as keepmock: + self.mock_keep_services(client, status=200, service_type='disk', count=6) + writer = self.foo_writer(api_client=client, replication=3) + writer.manifest_text() + self.assertEqual(6, keepmock.call_count) + def test_write_whole_collection_through_retries(self): writer = self.foo_writer(num_retries=2) with self.mock_keep(self.DEFAULT_DATA_HASH, @@ -743,7 +738,7 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): self.assertEqual('.', writer.current_stream_name()) self.assertEqual('out', writer.current_file_name()) out_file.write('test data') - data_loc = hashlib.md5('test data').hexdigest() + '+9' + data_loc = tutil.str_keep_locator('test data') self.assertTrue(out_file.closed, "writer file not closed after context") self.assertRaises(ValueError, out_file.write, 'extra text') with self.mock_keep(data_loc, 200) as keep_mock: @@ -755,15 +750,15 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): writer = arvados.CollectionWriter(client) with writer.open('six') as out_file: out_file.writelines(['12', '34', '56']) - data_loc = hashlib.md5('123456').hexdigest() + '+6' + data_loc = tutil.str_keep_locator('123456') with self.mock_keep(data_loc, 200) as keep_mock: self.assertEqual(". {} 0:6:six\n".format(data_loc), writer.manifest_text()) def test_open_flush(self): client = self.api_client_mock() - data_loc1 = hashlib.md5('flush1').hexdigest() + '+6' - data_loc2 = hashlib.md5('flush2').hexdigest() + '+6' + data_loc1 = tutil.str_keep_locator('flush1') + data_loc2 = tutil.str_keep_locator('flush2') with self.mock_keep((data_loc1, 200), (data_loc2, 200)) as keep_mock: writer = arvados.CollectionWriter(client) with writer.open('flush_test') as out_file: @@ -781,15 +776,15 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): out_file.write('1st') with writer.open('.', '2') as out_file: out_file.write('2nd') - data_loc = hashlib.md5('1st2nd').hexdigest() + '+6' + data_loc = tutil.str_keep_locator('1st2nd') with self.mock_keep(data_loc, 200) as keep_mock: self.assertEqual(". {} 0:3:1 3:3:2\n".format(data_loc), writer.manifest_text()) def test_two_opens_two_streams(self): client = self.api_client_mock() - data_loc1 = hashlib.md5('file').hexdigest() + '+4' - data_loc2 = hashlib.md5('indir').hexdigest() + '+5' + data_loc1 = tutil.str_keep_locator('file') + data_loc2 = tutil.str_keep_locator('indir') with self.mock_keep((data_loc1, 200), (data_loc2, 200)) as keep_mock: writer = arvados.CollectionWriter(client) with writer.open('file') as out_file: @@ -808,91 +803,347 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): - def test_import_manifest(self): + + def test_init_manifest(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt . 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt . 8b22da26f9f433dea0a10e5ec66d73ba+43 0:43:md5sum.txt """ - self.assertEqual(". 5348b82a029fd9e971a811ce1f71360b+43 085c37f02916da1cad16f93c54d899b7+41 8b22da26f9f433dea0a10e5ec66d73ba+43 0:127:md5sum.txt\n", arvados.export_manifest(arvados.import_manifest(m1))) + self.assertEqual(m1, CollectionReader(m1).manifest_text(normalize=False)) + self.assertEqual(". 5348b82a029fd9e971a811ce1f71360b+43 085c37f02916da1cad16f93c54d899b7+41 8b22da26f9f433dea0a10e5ec66d73ba+43 0:127:md5sum.txt\n", CollectionReader(m1).manifest_text(normalize=True)) - def test_init_manifest(self): + def test_init_manifest_with_collision(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt -. 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt -. 8b22da26f9f433dea0a10e5ec66d73ba+43 0:43:md5sum.txt +./md5sum.txt 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt """ - self.assertEqual(". 5348b82a029fd9e971a811ce1f71360b+43 085c37f02916da1cad16f93c54d899b7+41 8b22da26f9f433dea0a10e5ec66d73ba+43 0:127:md5sum.txt\n", arvados.export_manifest(arvados.Collection(m1))) + with self.assertRaises(arvados.errors.ArgumentError): + self.assertEqual(m1, CollectionReader(m1)) + def test_init_manifest_with_error(self): + m1 = """. 0:43:md5sum.txt""" + with self.assertRaises(arvados.errors.ArgumentError): + self.assertEqual(m1, CollectionReader(m1)) def test_remove(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n', sync=SYNC_EXPLICIT) as c: - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", export_manifest(c)) - self.assertIn("count1.txt", c) - c.remove("count1.txt") - self.assertNotIn("count1.txt", c) - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", export_manifest(c)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n') + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", c.portable_manifest_text()) + self.assertIn("count1.txt", c) + c.remove("count1.txt") + self.assertNotIn("count1.txt", c) + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", c.portable_manifest_text()) + with self.assertRaises(arvados.errors.ArgumentError): + c.remove("") + + def test_find(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n') + self.assertIs(c.find("."), c) + self.assertIs(c.find("./count1.txt"), c["count1.txt"]) + self.assertIs(c.find("count1.txt"), c["count1.txt"]) + with self.assertRaises(IOError): + c.find("/.") + with self.assertRaises(arvados.errors.ArgumentError): + c.find("") def test_remove_in_subdir(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n', sync=SYNC_EXPLICIT) as c: - c.remove("foo/count2.txt") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + c.remove("foo/count2.txt") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", c.portable_manifest_text()) def test_remove_empty_subdir(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n', sync=SYNC_EXPLICIT) as c: - c.remove("foo/count2.txt") - c.remove("foo") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + c.remove("foo/count2.txt") + c.remove("foo") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", c.portable_manifest_text()) def test_remove_nonempty_subdir(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n', sync=SYNC_EXPLICIT) as c: - with self.assertRaises(IOError): - c.remove("foo") - c.remove("foo", rm_r=True) - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c)) - - def test_copy_to_dir1(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n', sync=SYNC_EXPLICIT) as c: - c.copy("count1.txt", "foo/count2.txt") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", export_manifest(c)) - - def test_copy_to_dir2(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n', sync=SYNC_EXPLICIT) as c: - c.copy("count1.txt", "foo") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c)) - - def test_copy_to_dir2(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n', sync=SYNC_EXPLICIT) as c: - c.copy("count1.txt", "foo/") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + with self.assertRaises(IOError): + c.remove("foo") + c.remove("foo", recursive=True) + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", c.portable_manifest_text()) + + def test_copy_to_file_in_dir(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c.copy("count1.txt", "foo/count2.txt") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", c.portable_manifest_text()) def test_copy_file(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n', sync=SYNC_EXPLICIT) as c: - c.copy("count1.txt", "count2.txt") - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", export_manifest(c)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c.copy("count1.txt", "count2.txt") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", c.portable_manifest_text()) + + def test_copy_to_existing_dir(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + c.copy("count1.txt", "foo") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", c.portable_manifest_text()) + + def test_copy_to_new_dir(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c.copy("count1.txt", "foo/") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", c.portable_manifest_text()) + + def test_rename_file(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c.rename("count1.txt", "count2.txt") + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", c.manifest_text()) + + def test_move_file_to_dir(self): + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c.mkdirs("foo") + c.rename("count1.txt", "foo/count2.txt") + self.assertEqual("./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", c.manifest_text()) + + def test_move_file_to_other(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection() + c2.rename("count1.txt", "count2.txt", source_collection=c1) + self.assertEqual("", c1.manifest_text()) + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", c2.manifest_text()) def test_clone(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n', sync=SYNC_EXPLICIT) as c: - cl = c.clone() - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", export_manifest(cl)) + c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + cl = c.clone() + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n", cl.portable_manifest_text()) + + def test_diff_del_add(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count2.txt\n') + d = c2.diff(c1) + self.assertEqual(d, [('del', './count2.txt', c2["count2.txt"]), + ('add', './count1.txt', c1["count1.txt"])]) + d = c1.diff(c2) + self.assertEqual(d, [('del', './count1.txt', c1["count1.txt"]), + ('add', './count2.txt', c2["count2.txt"])]) + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_same(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + d = c2.diff(c1) + self.assertEqual(d, []) + d = c1.diff(c2) + self.assertEqual(d, []) + + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_mod(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt\n') + d = c2.diff(c1) + self.assertEqual(d, [('mod', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + d = c1.diff(c2) + self.assertEqual(d, [('mod', './count1.txt', c1["count1.txt"], c2["count1.txt"])]) + + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_add(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt 10:20:count2.txt\n') + d = c2.diff(c1) + self.assertEqual(d, [('del', './count2.txt', c2["count2.txt"])]) + d = c1.diff(c2) + self.assertEqual(d, [('add', './count2.txt', c2["count2.txt"])]) + + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_add_in_subcollection(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 5348b82a029fd9e971a811ce1f71360b+43 0:10:count2.txt\n') + d = c2.diff(c1) + self.assertEqual(d, [('del', './foo', c2["foo"])]) + d = c1.diff(c2) + self.assertEqual(d, [('add', './foo', c2["foo"])]) + + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_del_add_in_subcollection(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 5348b82a029fd9e971a811ce1f71360b+43 0:10:count2.txt\n') + c2 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 5348b82a029fd9e971a811ce1f71360b+43 0:3:count3.txt\n') + + d = c2.diff(c1) + self.assertEqual(d, [('del', './foo/count3.txt', c2.find("foo/count3.txt")), + ('add', './foo/count2.txt', c1.find("foo/count2.txt"))]) + d = c1.diff(c2) + self.assertEqual(d, [('del', './foo/count2.txt', c1.find("foo/count2.txt")), + ('add', './foo/count3.txt', c2.find("foo/count3.txt"))]) + + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_diff_mod_in_subcollection(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 5348b82a029fd9e971a811ce1f71360b+43 0:10:count2.txt\n') + c2 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:3:foo\n') + d = c2.diff(c1) + self.assertEqual(d, [('mod', './foo', c2["foo"], c1["foo"])]) + d = c1.diff(c2) + self.assertEqual(d, [('mod', './foo', c1["foo"], c2["foo"])]) + + self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) + + def test_conflict_keep_local_change(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count2.txt\n') + d = c1.diff(c2) + self.assertEqual(d, [('del', './count1.txt', c1["count1.txt"]), + ('add', './count2.txt', c2["count2.txt"])]) + f = c1.open("count1.txt", "w") + f.write("zzzzz") + + # c1 changed, so it should not be deleted. + c1.apply(d) + self.assertEqual(c1.portable_manifest_text(), ". 95ebc3c7b3b9f1d2c40fec14415d3cb8+5 5348b82a029fd9e971a811ce1f71360b+43 0:5:count1.txt 5:10:count2.txt\n") + + def test_conflict_mod(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt') + d = c1.diff(c2) + self.assertEqual(d, [('mod', './count1.txt', c1["count1.txt"], c2["count1.txt"])]) + f = c1.open("count1.txt", "w") + f.write("zzzzz") + + # c1 changed, so c2 mod will go to a conflict file + c1.apply(d) + self.assertRegexpMatches(c1.portable_manifest_text(), r"\. 95ebc3c7b3b9f1d2c40fec14415d3cb8\+5 5348b82a029fd9e971a811ce1f71360b\+43 0:5:count1\.txt 5:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") + + def test_conflict_add(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt\n') + d = c1.diff(c2) + self.assertEqual(d, [('del', './count2.txt', c1["count2.txt"]), + ('add', './count1.txt', c2["count1.txt"])]) + f = c1.open("count1.txt", "w") + f.write("zzzzz") + + # c1 added count1.txt, so c2 add will go to a conflict file + c1.apply(d) + self.assertRegexpMatches(c1.portable_manifest_text(), r"\. 95ebc3c7b3b9f1d2c40fec14415d3cb8\+5 5348b82a029fd9e971a811ce1f71360b\+43 0:5:count1\.txt 5:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") + + def test_conflict_del(self): + c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt') + c2 = Collection('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt') + d = c1.diff(c2) + self.assertEqual(d, [('mod', './count1.txt', c1["count1.txt"], c2["count1.txt"])]) + c1.remove("count1.txt") + + # c1 deleted, so c2 mod will go to a conflict file + c1.apply(d) + self.assertRegexpMatches(c1.portable_manifest_text(), r"\. 5348b82a029fd9e971a811ce1f71360b\+43 0:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") + + def test_notify(self): + c1 = Collection() + events = [] + c1.subscribe(lambda event, collection, name, item: events.append((event, collection, name, item))) + f = c1.open("foo.txt", "w") + self.assertEqual(events[0], (arvados.collection.ADD, c1, "foo.txt", f.arvadosfile)) + + def test_open_w(self): + c1 = Collection(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n") + self.assertEqual(c1["count1.txt"].size(), 10) + c1.open("count1.txt", "w").close() + self.assertEqual(c1["count1.txt"].size(), 0) + + +class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): + MAIN_SERVER = {} + KEEP_SERVER = {} + + def create_count_txt(self): + # Create an empty collection, save it to the API server, then write a + # file, but don't save it. + + c = Collection() + c.save_new("CollectionCreateUpdateTest", ensure_unique_name=True) + self.assertEqual(c.portable_data_hash(), "d41d8cd98f00b204e9800998ecf8427e+0") + self.assertEqual(c.api_response()["portable_data_hash"], "d41d8cd98f00b204e9800998ecf8427e+0" ) + + with c.open("count.txt", "w") as f: + f.write("0123456789") + + self.assertEqual(c.portable_manifest_text(), ". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count.txt\n") + + return c + + def test_create_and_save(self): + c = self.create_count_txt() + c.save() + self.assertRegexpMatches(c.manifest_text(), r"^\. 781e5e245d69b566979b86e28d23f2c7\+10\+A[a-f0-9]{40}@[a-f0-9]{8} 0:10:count\.txt$",) + + def test_create_and_save_new(self): + c = self.create_count_txt() + c.save_new() + self.assertRegexpMatches(c.manifest_text(), r"^\. 781e5e245d69b566979b86e28d23f2c7\+10\+A[a-f0-9]{40}@[a-f0-9]{8} 0:10:count\.txt$",) + + def test_create_diff_apply(self): + c1 = self.create_count_txt() + c1.save() + + c2 = Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + + diff = c1.diff(c2) + + self.assertEqual(diff[0], (arvados.collection.MOD, u'./count.txt', c1["count.txt"], c2["count.txt"])) + + c1.apply(diff) + self.assertEqual(c1.portable_data_hash(), c2.portable_data_hash()) + + def test_diff_apply_with_token(self): + baseline = CollectionReader(". 781e5e245d69b566979b86e28d23f2c7+10+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:10:count.txt\n") + c = Collection(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count.txt\n") + other = CollectionReader(". 7ac66c0f148de9519b8bd264312c4d64+7+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:7:count.txt\n") + + diff = baseline.diff(other) + self.assertEqual(diff, [('mod', u'./count.txt', c["count.txt"], other["count.txt"])]) + + c.apply(diff) + + self.assertEqual(c.manifest_text(), ". 7ac66c0f148de9519b8bd264312c4d64+7+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:7:count.txt\n") + + + def test_create_and_update(self): + c1 = self.create_count_txt() + c1.save() + + c2 = arvados.collection.Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + + c2.save() + + self.assertNotEqual(c1.portable_data_hash(), c2.portable_data_hash()) + c1.update() + self.assertEqual(c1.portable_data_hash(), c2.portable_data_hash()) + - def test_merge1(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt', sync=SYNC_EXPLICIT) as c1: - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt') as c2: - c1.merge(c2) - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n", export_manifest(c1)) + def test_create_and_update_with_conflict(self): + c1 = self.create_count_txt() + c1.save() - def test_merge2(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt', sync=SYNC_EXPLICIT) as c1: - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt') as c2: - c1.merge(c2) - self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n", export_manifest(c1)) + with c1.open("count.txt", "w") as f: + f.write("XYZ") - def test_merge3(self): - with arvados.import_manifest('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt', sync=SYNC_EXPLICIT) as c1: - with arvados.import_manifest('. 5348b82a029fd9e971a811ce1f71360b+43 0:10:count1.txt') as c2: - c1.merge(c2) - self.assertTrue(re.match(r". 781e5e245d69b566979b86e28d23f2c7\+10 5348b82a029fd9e971a811ce1f71360b\+43 0:10:count1.txt 10:10:count1.txt~conflict-\d\d\d\d-\d\d-\d\d_\d\d:\d\d:\d\d~\n", export_manifest(c1))) + c2 = arvados.collection.Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + c2.save() + c1.update() + self.assertRegexpMatches(c1.manifest_text(), r"\. e65075d550f9b5bf9992fa1d71a131be\+3\S* 7ac66c0f148de9519b8bd264312c4d64\+7\S* 0:3:count\.txt 3:7:count\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") if __name__ == '__main__':