X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/409d130ddcc4b76def5fa8d61d2584725c61152a..093ec98e4a065acfc537ea22c08c337c115fe273:/sdk/python/tests/test_collections.py diff --git a/sdk/python/tests/test_collections.py b/sdk/python/tests/test_collections.py index 0e3d5e13f1..5cf4993b2f 100644 --- a/sdk/python/tests/test_collections.py +++ b/sdk/python/tests/test_collections.py @@ -1,20 +1,26 @@ -# usage example: +# Copyright (C) The Arvados Authors. All rights reserved. # -# ARVADOS_API_TOKEN=abc ARVADOS_API_HOST=arvados.local python -m unittest discover +# SPDX-License-Identifier: Apache-2.0 +from __future__ import absolute_import + +from builtins import object import arvados import copy import mock import os -import pprint +import random import re -import tempfile +import sys +import datetime +import ciso8601 +import time import unittest -import run_test_server +from . import run_test_server from arvados._ranges import Range, LocatorAndRange from arvados.collection import Collection, CollectionReader -import arvados_testutil as tutil +from . import arvados_testutil as tutil class TestResumableWriter(arvados.ResumableCollectionWriter): KEEP_BLOCK_SIZE = 1024 # PUT to Keep every 1K. @@ -30,7 +36,8 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, @classmethod def setUpClass(cls): super(ArvadosCollectionsTest, cls).setUpClass() - run_test_server.authorize_with('active') + # need admin privileges to make collections with unsigned blocks + run_test_server.authorize_with('admin') cls.api_client = arvados.api('v1') cls.keep_client = arvados.KeepClient(api_client=cls.api_client, local_store=cls.local_store) @@ -40,24 +47,28 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, self.assertEqual(cw.current_stream_name(), '.', 'current_stream_name() should be "." now') cw.set_current_file_name('foo.txt') - cw.write('foo') + cw.write(b'foo') self.assertEqual(cw.current_file_name(), 'foo.txt', 'current_file_name() should be foo.txt now') cw.start_new_file('bar.txt') - cw.write('bar') + cw.write(b'bar') cw.start_new_stream('baz') - cw.write('baz') + cw.write(b'baz') cw.set_current_file_name('baz.txt') self.assertEqual(cw.manifest_text(), ". 3858f62230ac3c915f300c664312c63f+6 0:3:foo.txt 3:3:bar.txt\n" + "./baz 73feffa4b7f6bb68e44cf984c85f6e88+3 0:3:baz.txt\n", "wrong manifest: got {}".format(cw.manifest_text())) - cw.finish() + cw.save_new() return cw.portable_data_hash() + def test_pdh_is_native_str(self): + pdh = self.write_foo_bar_baz() + self.assertEqual(type(''), type(pdh)) + def test_keep_local_store(self): - self.assertEqual(self.keep_client.put('foo'), 'acbd18db4cc2f85cedef654fccc4a4d8+3', 'wrong md5 hash from Keep.put') - self.assertEqual(self.keep_client.get('acbd18db4cc2f85cedef654fccc4a4d8+3'), 'foo', 'wrong data from Keep.get') + self.assertEqual(self.keep_client.put(b'foo'), 'acbd18db4cc2f85cedef654fccc4a4d8+3', 'wrong md5 hash from Keep.put') + self.assertEqual(self.keep_client.get('acbd18db4cc2f85cedef654fccc4a4d8+3'), b'foo', 'wrong data from Keep.get') def test_local_collection_writer(self): self.assertEqual(self.write_foo_bar_baz(), @@ -72,21 +83,23 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, for s in cr.all_streams(): for f in s.all_files(): got += [[f.size(), f.stream_name(), f.name(), f.read(2**26)]] - expected = [[3, '.', 'foo.txt', 'foo'], - [3, '.', 'bar.txt', 'bar'], - [3, './baz', 'baz.txt', 'baz']] + expected = [[3, '.', 'foo.txt', b'foo'], + [3, '.', 'bar.txt', b'bar'], + [3, './baz', 'baz.txt', b'baz']] self.assertEqual(got, expected) stream0 = cr.all_streams()[0] self.assertEqual(stream0.readfrom(0, 0), - '', + b'', 'reading zero bytes should have returned empty string') self.assertEqual(stream0.readfrom(0, 2**26), - 'foobar', + b'foobar', 'reading entire stream failed') self.assertEqual(stream0.readfrom(2**26, 0), - '', + b'', 'reading zero bytes should have returned empty string') + self.assertEqual(3, len(cr)) + self.assertTrue(cr) def _test_subset(self, collection, expected): cr = arvados.CollectionReader(collection, self.api_client) @@ -102,50 +115,50 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, def test_collection_manifest_subset(self): foobarbaz = self.write_foo_bar_baz() self._test_subset(foobarbaz, - [[3, '.', 'bar.txt', 'bar'], - [3, '.', 'foo.txt', 'foo'], - [3, './baz', 'baz.txt', 'baz']]) + [[3, '.', 'bar.txt', b'bar'], + [3, '.', 'foo.txt', b'foo'], + [3, './baz', 'baz.txt', b'baz']]) self._test_subset((". %s %s 0:3:foo.txt 3:3:bar.txt\n" % - (self.keep_client.put("foo"), - self.keep_client.put("bar"))), - [[3, '.', 'bar.txt', 'bar'], - [3, '.', 'foo.txt', 'foo']]) + (self.keep_client.put(b"foo"), + self.keep_client.put(b"bar"))), + [[3, '.', 'bar.txt', b'bar'], + [3, '.', 'foo.txt', b'foo']]) self._test_subset((". %s %s 0:2:fo.txt 2:4:obar.txt\n" % - (self.keep_client.put("foo"), - self.keep_client.put("bar"))), - [[2, '.', 'fo.txt', 'fo'], - [4, '.', 'obar.txt', 'obar']]) + (self.keep_client.put(b"foo"), + self.keep_client.put(b"bar"))), + [[2, '.', 'fo.txt', b'fo'], + [4, '.', 'obar.txt', b'obar']]) self._test_subset((". %s %s 0:2:fo.txt 2:0:zero.txt 2:2:ob.txt 4:2:ar.txt\n" % - (self.keep_client.put("foo"), - self.keep_client.put("bar"))), - [[2, '.', 'ar.txt', 'ar'], - [2, '.', 'fo.txt', 'fo'], - [2, '.', 'ob.txt', 'ob'], - [0, '.', 'zero.txt', '']]) + (self.keep_client.put(b"foo"), + self.keep_client.put(b"bar"))), + [[2, '.', 'ar.txt', b'ar'], + [2, '.', 'fo.txt', b'fo'], + [2, '.', 'ob.txt', b'ob'], + [0, '.', 'zero.txt', b'']]) def test_collection_empty_file(self): cw = arvados.CollectionWriter(self.api_client) cw.start_new_file('zero.txt') - cw.write('') + cw.write(b'') self.assertEqual(cw.manifest_text(), ". d41d8cd98f00b204e9800998ecf8427e+0 0:0:zero.txt\n") self.check_manifest_file_sizes(cw.manifest_text(), [0]) cw = arvados.CollectionWriter(self.api_client) cw.start_new_file('zero.txt') - cw.write('') + cw.write(b'') cw.start_new_file('one.txt') - cw.write('1') + cw.write(b'1') cw.start_new_stream('foo') cw.start_new_file('zero.txt') - cw.write('') + cw.write(b'') self.check_manifest_file_sizes(cw.manifest_text(), [0,1,0]) def test_no_implicit_normalize(self): cw = arvados.CollectionWriter(self.api_client) cw.start_new_file('b') - cw.write('b') + cw.write(b'b') cw.start_new_file('a') - cw.write('') + cw.write(b'') self.check_manifest_file_sizes(cw.manifest_text(), [1,0]) self.check_manifest_file_sizes( arvados.CollectionReader( @@ -307,18 +320,20 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, def __init__(self, content, num_retries=0): self.content = content - def get(self, locator, num_retries=0): + def get(self, locator, num_retries=0, prefetch=False): return self.content[locator] def test_stream_reader(self): - keepblocks = {'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa+10': 'abcdefghij', - 'bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb+15': 'klmnopqrstuvwxy', - 'cccccccccccccccccccccccccccccccc+5': 'z0123'} + keepblocks = { + 'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa+10': b'abcdefghij', + 'bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb+15': b'klmnopqrstuvwxy', + 'cccccccccccccccccccccccccccccccc+5': b'z0123', + } mk = self.MockKeep(keepblocks) sr = arvados.StreamReader([".", "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa+10", "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb+15", "cccccccccccccccccccccccccccccccc+5", "0:30:foo"], mk) - content = 'abcdefghijklmnopqrstuvwxyz0123456789' + content = b'abcdefghijklmnopqrstuvwxyz0123456789' self.assertEqual(sr.readfrom(0, 30), content[0:30]) self.assertEqual(sr.readfrom(2, 30), content[2:30]) @@ -332,7 +347,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, self.assertEqual(sr.readfrom(15, 5), content[15:20]) self.assertEqual(sr.readfrom(20, 5), content[20:25]) self.assertEqual(sr.readfrom(25, 5), content[25:30]) - self.assertEqual(sr.readfrom(30, 5), '') + self.assertEqual(sr.readfrom(30, 5), b'') def test_extract_file(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt @@ -421,7 +436,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, def test_write_multiple_files(self): cwriter = arvados.CollectionWriter(self.api_client) for letter in 'ABC': - with self.make_test_file(letter) as testfile: + with self.make_test_file(letter.encode()) as testfile: cwriter.write_file(testfile.name, letter) self.assertEqual( cwriter.manifest_text(), @@ -464,7 +479,7 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, with self.make_test_file() as testfile: cwriter.write_file(testfile.name, 'test') orig_mtime = os.fstat(testfile.fileno()).st_mtime - testfile.write('extra') + testfile.write(b'extra') testfile.flush() os.utime(testfile.name, (orig_mtime, orig_mtime)) self.assertRaises(arvados.errors.StaleWriterStateError, @@ -491,15 +506,6 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, self.assertRaises(arvados.errors.AssertionError, cwriter.write, "badtext") - def test_read_arbitrary_data_with_collection_reader(self): - # arv-get relies on this to do "arv-get {keep-locator} -". - self.write_foo_bar_baz() - self.assertEqual( - 'foobar', - arvados.CollectionReader( - '3858f62230ac3c915f300c664312c63f+6' - ).manifest_text()) - class CollectionTestMixin(tutil.ApiClientMock): API_COLLECTIONS = run_test_server.fixture('collections') @@ -519,8 +525,8 @@ class CollectionTestMixin(tutil.ApiClientMock): @tutil.skip_sleep class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): - def mock_get_collection(self, api_mock, code, body): - body = self.API_COLLECTIONS.get(body) + def mock_get_collection(self, api_mock, code, fixturename): + body = self.API_COLLECTIONS.get(fixturename) self._mock_api_call(api_mock.collections().get, code, body) def api_client_mock(self, status=200): @@ -555,34 +561,13 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): api_client=client) self.assertEqual(self.DEFAULT_MANIFEST, reader.manifest_text()) - def test_locator_init_fallback_to_keep(self): - # crunch-job needs this to read manifests that have only ever - # been written to Keep. - client = self.api_client_mock(200) - self.mock_get_collection(client, 404, None) - 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()) - - def test_uuid_init_no_fallback_to_keep(self): - # Do not look up a collection UUID in Keep. - client = self.api_client_mock(404) - with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): - with self.assertRaises(arvados.errors.ApiError): - 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_keep_responses(self.ALT_MANIFEST, 200): - self.assertEqual( - self.ALT_MANIFEST, - arvados.CollectionReader( - self.ALT_DATA_HASH + '+Affffffffffffffffffffffffffffffffffffffff@fedcba98', - api_client=client).manifest_text()) + def test_init_no_fallback_to_keep(self): + # Do not look up a collection UUID or PDH in Keep. + for key in [self.DEFAULT_UUID, self.DEFAULT_DATA_HASH]: + client = self.api_client_mock(404) + with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): + with self.assertRaises(arvados.errors.ApiError): + reader = arvados.CollectionReader(key, api_client=client) def test_init_num_retries_propagated(self): # More of an integration test... @@ -590,8 +575,8 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client, num_retries=3) with tutil.mock_keep_responses('foo', 500, 500, 200): - self.assertEqual('foo', - ''.join(f.read(9) for f in reader.all_files())) + self.assertEqual(b'foo', + b''.join(f.read(9) for f in reader.all_files())) def test_read_nonnormalized_manifest_with_collection_reader(self): # client should be able to use CollectionReader on a manifest without normalizing it @@ -620,21 +605,14 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): reader = arvados.CollectionReader('d41d8cd98f00b204e9800998ecf8427e+0', api_client=client) self.assertEqual('', reader.manifest_text()) + self.assertEqual(0, len(reader)) + self.assertFalse(reader) def test_api_response(self): client = self.api_client_mock() reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) self.assertEqual(self.DEFAULT_COLLECTION, reader.api_response()) - def test_api_response_with_collection_from_keep(self): - client = self.api_client_mock() - self.mock_get_collection(client, 404, 'foo') - with tutil.mock_keep_responses(self.DEFAULT_MANIFEST, 200): - reader = arvados.CollectionReader(self.DEFAULT_DATA_HASH, - api_client=client) - api_response = reader.api_response() - self.assertIsNone(api_response) - 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()) @@ -644,7 +622,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): def test_open_collection_file_one_argument(self): client = self.api_client_mock(200) reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) - cfile = reader.open('./foo') + cfile = reader.open('./foo', 'rb') self.check_open_file(cfile, '.', 'foo', 3) def test_open_deep_file(self): @@ -653,7 +631,7 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): self.mock_get_collection(client, 200, coll_name) reader = arvados.CollectionReader( self.API_COLLECTIONS[coll_name]['uuid'], api_client=client) - cfile = reader.open('./subdir2/subdir3/file2_in_subdir3.txt') + cfile = reader.open('./subdir2/subdir3/file2_in_subdir3.txt', 'rb') self.check_open_file(cfile, './subdir2/subdir3', 'file2_in_subdir3.txt', 32) @@ -678,7 +656,7 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): kwargs.setdefault('api_client', self.api_client_mock()) writer = arvados.CollectionWriter(**kwargs) writer.start_new_file('foo') - writer.write('foo') + writer.write(b'foo') return writer def test_write_whole_collection(self): @@ -737,7 +715,7 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): with writer.open('out') as out_file: self.assertEqual('.', writer.current_stream_name()) self.assertEqual('out', writer.current_file_name()) - out_file.write('test data') + out_file.write(b'test data') 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') @@ -762,9 +740,9 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): 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: - out_file.write('flush1') + out_file.write(b'flush1') out_file.flush() - out_file.write('flush2') + out_file.write(b'flush2') self.assertEqual(". {} {} 0:12:flush_test\n".format(data_loc1, data_loc2), writer.manifest_text()) @@ -773,9 +751,9 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): client = self.api_client_mock() writer = arvados.CollectionWriter(client) with writer.open('.', '1') as out_file: - out_file.write('1st') + out_file.write(b'1st') with writer.open('.', '2') as out_file: - out_file.write('2nd') + out_file.write(b'2nd') 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), @@ -788,9 +766,9 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): with self.mock_keep((data_loc1, 200), (data_loc2, 200)) as keep_mock: writer = arvados.CollectionWriter(client) with writer.open('file') as out_file: - out_file.write('file') + out_file.write(b'file') with writer.open('./dir', 'indir') as out_file: - out_file.write('indir') + out_file.write(b'indir') expected = ". {} 0:4:file\n./dir {} 0:5:indir\n".format( data_loc1, data_loc2) self.assertEqual(expected, writer.manifest_text()) @@ -802,6 +780,114 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): self.assertRaises(arvados.errors.AssertionError, writer.open, 'two') +class CollectionMethods(run_test_server.TestCaseWithServers): + + def test_keys_values_items_support_indexing(self): + c = Collection() + with c.open('foo', 'wb') as f: + f.write(b'foo') + with c.open('bar', 'wb') as f: + f.write(b'bar') + self.assertEqual(2, len(c.keys())) + if sys.version_info < (3, 0): + # keys() supports indexing only for python2 callers. + fn0 = c.keys()[0] + fn1 = c.keys()[1] + else: + fn0, fn1 = c.keys() + self.assertEqual(2, len(c.values())) + f0 = c.values()[0] + f1 = c.values()[1] + self.assertEqual(2, len(c.items())) + self.assertEqual(fn0, c.items()[0][0]) + self.assertEqual(fn1, c.items()[1][0]) + + def test_get_properties(self): + c = Collection() + self.assertEqual(c.get_properties(), {}) + c.save_new(properties={"foo":"bar"}) + self.assertEqual(c.get_properties(), {"foo":"bar"}) + + def test_get_trash_at(self): + c = Collection() + self.assertEqual(c.get_trash_at(), None) + c.save_new(trash_at=datetime.datetime(2111, 1, 1, 11, 11, 11, 111111)) + self.assertEqual(c.get_trash_at(), ciso8601.parse_datetime('2111-01-01T11:11:11.111111000Z')) + + +class CollectionOpenModes(run_test_server.TestCaseWithServers): + + def test_open_binary_modes(self): + c = Collection() + for mode in ['wb', 'wb+', 'ab', 'ab+']: + with c.open('foo', mode) as f: + f.write(b'foo') + + def test_open_invalid_modes(self): + c = Collection() + for mode in ['+r', 'aa', '++', 'r+b', 'beer', '', None]: + with self.assertRaises(Exception): + c.open('foo', mode) + + def test_open_text_modes(self): + c = Collection() + with c.open('foo', 'wb') as f: + f.write('foo') + for mode in ['r', 'rt', 'r+', 'rt+', 'w', 'wt', 'a', 'at']: + with c.open('foo', mode) as f: + if mode[0] == 'r' and '+' not in mode: + self.assertEqual('foo', f.read(3)) + else: + f.write('bar') + f.seek(0, os.SEEK_SET) + self.assertEqual('bar', f.read(3)) + + +class TextModes(run_test_server.TestCaseWithServers): + + def setUp(self): + arvados.config.KEEP_BLOCK_SIZE = 4 + if sys.version_info < (3, 0): + import unicodedata + self.sailboat = unicodedata.lookup('SAILBOAT') + self.snowman = unicodedata.lookup('SNOWMAN') + else: + self.sailboat = '\N{SAILBOAT}' + self.snowman = '\N{SNOWMAN}' + + def tearDown(self): + arvados.config.KEEP_BLOCK_SIZE = 2 ** 26 + + def test_read_sailboat_across_block_boundary(self): + c = Collection() + f = c.open('sailboats', 'wb') + data = self.sailboat.encode('utf-8') + f.write(data) + f.write(data[:1]) + f.write(data[1:]) + f.write(b'\n') + f.close() + self.assertRegex(c.portable_manifest_text(), r'\+4 .*\+3 ') + + f = c.open('sailboats', 'r') + string = f.readline() + self.assertEqual(string, self.sailboat+self.sailboat+'\n') + f.close() + + def test_write_snowman_across_block_boundary(self): + c = Collection() + f = c.open('snowmany', 'w') + data = self.snowman + f.write(data+data+'\n'+data+'\n') + f.close() + self.assertRegex(c.portable_manifest_text(), r'\+4 .*\+4 .*\+3 ') + + f = c.open('snowmany', 'r') + self.assertEqual(f.readline(), self.snowman+self.snowman+'\n') + self.assertEqual(f.readline(), self.snowman+'\n') + f.close() + + class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): def test_replication_desired_kept_on_load(self): @@ -822,6 +908,37 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): self.assertEqual(c1.manifest_text, c2.manifest_text) self.assertNotEqual(c1.replication_desired, c2.replication_desired) + def test_storage_classes_desired_kept_on_load(self): + m = '. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n' + c1 = Collection(m, storage_classes_desired=['archival']) + c1.save_new() + loc = c1.manifest_locator() + c2 = Collection(loc) + self.assertEqual(c1.manifest_text, c2.manifest_text) + self.assertEqual(c1.storage_classes_desired(), c2.storage_classes_desired()) + + def test_storage_classes_change_after_save(self): + m = '. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n' + c1 = Collection(m, storage_classes_desired=['archival']) + c1.save_new() + loc = c1.manifest_locator() + c2 = Collection(loc) + self.assertEqual(['archival'], c2.storage_classes_desired()) + c2.save(storage_classes=['highIO']) + self.assertEqual(['highIO'], c2.storage_classes_desired()) + c3 = Collection(loc) + self.assertEqual(c1.manifest_text, c3.manifest_text) + self.assertEqual(['highIO'], c3.storage_classes_desired()) + + def test_storage_classes_desired_not_loaded_if_provided(self): + m = '. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n' + c1 = Collection(m, storage_classes_desired=['archival']) + c1.save_new() + loc = c1.manifest_locator() + c2 = Collection(loc, storage_classes_desired=['default']) + self.assertEqual(c1.manifest_text, c2.manifest_text) + self.assertNotEqual(c1.storage_classes_desired(), c2.storage_classes_desired()) + def test_init_manifest(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt . 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt @@ -864,10 +981,49 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): self.assertIs(c.find("./nonexistant.txt"), None) self.assertIs(c.find("./nonexistantsubdir/nonexistant.txt"), None) + def test_escaped_paths_dont_get_unescaped_on_manifest(self): + # Dir & file names are literally '\056' (escaped form: \134056) + manifest = './\\134056\\040Test d41d8cd98f00b204e9800998ecf8427e+0 0:0:\\134056\n' + c = Collection(manifest) + self.assertEqual(c.portable_manifest_text(), manifest) + + def test_other_special_chars_on_file_token(self): + cases = [ + ('\\000', '\0'), + ('\\011', '\t'), + ('\\012', '\n'), + ('\\072', ':'), + ('\\134400', '\\400'), + ] + for encoded, decoded in cases: + manifest = '. d41d8cd98f00b204e9800998ecf8427e+0 0:0:some%sfile.txt\n' % encoded + c = Collection(manifest) + self.assertEqual(c.portable_manifest_text(), manifest) + self.assertIn('some%sfile.txt' % decoded, c.keys()) + + def test_escaped_paths_do_get_unescaped_on_listing(self): + # Dir & file names are literally '\056' (escaped form: \134056) + manifest = './\\134056\\040Test d41d8cd98f00b204e9800998ecf8427e+0 0:0:\\134056\n' + c = Collection(manifest) + self.assertIn('\\056 Test', c.keys()) + self.assertIn('\\056', c['\\056 Test'].keys()) + + def test_make_empty_dir_with_escaped_chars(self): + c = Collection() + c.mkdirs('./Empty\\056Dir') + self.assertEqual(c.portable_manifest_text(), + './Empty\\134056Dir d41d8cd98f00b204e9800998ecf8427e+0 0:0:\\056\n') + + def test_make_empty_dir_with_spaces(self): + c = Collection() + c.mkdirs('./foo bar/baz waz') + self.assertEqual(c.portable_manifest_text(), + './foo\\040bar/baz\\040waz d41d8cd98f00b204e9800998ecf8427e+0 0:0:\\056\n') + def test_remove_in_subdir(self): 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()) + self.assertEqual(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo d41d8cd98f00b204e9800998ecf8427e+0 0:0:\\056\n", c.portable_manifest_text()) def test_remove_empty_subdir(self): c = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt\n./foo 781e5e245d69b566979b86e28d23f2c7+10 0:10:count2.txt\n') @@ -929,11 +1085,15 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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"])]) + self.assertEqual(sorted(d), [ + ('add', './count1.txt', c1["count1.txt"]), + ('del', './count2.txt', c2["count2.txt"]), + ]) d = c1.diff(c2) - self.assertEqual(d, [('del', './count1.txt', c1["count1.txt"]), - ('add', './count2.txt', c2["count2.txt"])]) + self.assertEqual(sorted(d), [ + ('add', './count2.txt', c2["count2.txt"]), + ('del', './count1.txt', c1["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()) @@ -966,11 +1126,15 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('del', './count2.txt', c2["count2.txt"]), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) d = c1.diff(c2) - self.assertEqual(d, [('add', './count2.txt', c2["count2.txt"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('add', './count2.txt', c2["count2.txt"]), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) c1.apply(d) @@ -980,12 +1144,15 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('del', './foo', c2["foo"]), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) d = c1.diff(c2) - self.assertEqual(d, [('add', './foo', c2["foo"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) - + self.assertEqual(sorted(d), [ + ('add', './foo', c2["foo"]), + ('tok', './count1.txt', c2["count1.txt"], c1["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()) @@ -993,15 +1160,18 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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")), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('add', './foo/count2.txt', c1.find("foo/count2.txt")), + ('del', './foo/count3.txt', c2.find("foo/count3.txt")), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.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")), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('add', './foo/count3.txt', c2.find("foo/count3.txt")), + ('del', './foo/count2.txt', c1.find("foo/count2.txt")), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) c1.apply(d) @@ -1011,11 +1181,15 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('mod', './foo', c2["foo"], c1["foo"]), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) d = c1.diff(c2) - self.assertEqual(d, [('mod', './foo', c1["foo"], c2["foo"]), - ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + self.assertEqual(sorted(d), [ + ('mod', './foo', c1["foo"], c2["foo"]), + ('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"]), + ]) self.assertNotEqual(c1.portable_manifest_text(), c2.portable_manifest_text()) c1.apply(d) @@ -1025,10 +1199,12 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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") + self.assertEqual(sorted(d), [ + ('add', './count2.txt', c2["count2.txt"]), + ('del', './count1.txt', c1["count1.txt"]), + ]) + f = c1.open("count1.txt", "wb") + f.write(b"zzzzz") # c1 changed, so it should not be deleted. c1.apply(d) @@ -1039,25 +1215,31 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): 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") + f = c1.open("count1.txt", "wb") + f.write(b"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~$") + self.assertRegex( + c1.portable_manifest_text(), + r"\. 95ebc3c7b3b9f1d2c40fec14415d3cb8\+5 5348b82a029fd9e971a811ce1f71360b\+43 0:5:count1\.txt 5:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") 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") + self.assertEqual(sorted(d), [ + ('add', './count1.txt', c2["count1.txt"]), + ('del', './count2.txt', c1["count2.txt"]), + ]) + f = c1.open("count1.txt", "wb") + f.write(b"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~$") + self.assertRegex( + c1.portable_manifest_text(), + r"\. 95ebc3c7b3b9f1d2c40fec14415d3cb8\+5 5348b82a029fd9e971a811ce1f71360b\+43 0:5:count1\.txt 5:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") def test_conflict_del(self): c1 = Collection('. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt') @@ -1068,31 +1250,143 @@ class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): # 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~$") + self.assertRegex( + c1.portable_manifest_text(), + r"\. 5348b82a029fd9e971a811ce1f71360b\+43 0:10:count1\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") 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") + f = c1.open("foo.txt", "wb") 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() + c1.open("count1.txt", "wb").close() self.assertEqual(c1["count1.txt"].size(), 0) +class NewCollectionTestCaseWithServersAndTokens(run_test_server.TestCaseWithServers): + MAIN_SERVER = {} + KEEP_SERVER = {} + local_locator_re = r"[0-9a-f]{32}\+\d+\+A[a-f0-9]{40}@[a-f0-9]{8}" + remote_locator_re = r"[0-9a-f]{32}\+\d+\+R[a-z]{5}-[a-f0-9]{40}@[a-f0-9]{8}" + + def setUp(self): + self.keep_put = getattr(arvados.keep.KeepClient, 'put') + + @mock.patch('arvados.keep.KeepClient.put', autospec=True) + def test_storage_classes_desired(self, put_mock): + put_mock.side_effect = self.keep_put + c = Collection(storage_classes_desired=['default']) + with c.open("file.txt", 'wb') as f: + f.write('content') + c.save_new() + _, kwargs = put_mock.call_args + self.assertEqual(['default'], kwargs['classes']) + + @mock.patch('arvados.keep.KeepClient.put', autospec=True) + def test_repacked_block_submission_get_permission_token(self, mocked_put): + ''' + Make sure that those blocks that are committed after repacking small ones, + get their permission tokens assigned on the collection manifest. + ''' + def wrapped_keep_put(*args, **kwargs): + # Simulate slow put operations + time.sleep(1) + return self.keep_put(*args, **kwargs) + + mocked_put.side_effect = wrapped_keep_put + c = Collection() + # Write 70 files ~1MiB each so we force to produce 1 big block by repacking + # small ones before finishing the upload. + for i in range(70): + f = c.open("file_{}.txt".format(i), 'wb') + f.write(random.choice('abcdefghijklmnopqrstuvwxyz') * (2**20+i)) + f.close(flush=False) + # We should get 2 blocks with their tokens + self.assertEqual(len(re.findall(self.local_locator_re, c.manifest_text())), 2) + + @mock.patch('arvados.keep.KeepClient.refresh_signature') + def test_copy_remote_blocks_on_save_new(self, rs_mock): + remote_block_loc = "acbd18db4cc2f85cedef654fccc4a4d8+3+Remote-" + "a" * 40 + "@abcdef01" + local_block_loc = "acbd18db4cc2f85cedef654fccc4a4d8+3+A" + "b" * 40 + "@abcdef01" + rs_mock.return_value = local_block_loc + c = Collection(". " + remote_block_loc + " 0:3:foofile.txt\n") + self.assertEqual( + len(re.findall(self.remote_locator_re, c.manifest_text())), 1) + self.assertEqual( + len(re.findall(self.local_locator_re, c.manifest_text())), 0) + c.save_new() + rs_mock.assert_called() + self.assertEqual( + len(re.findall(self.remote_locator_re, c.manifest_text())), 0) + self.assertEqual( + len(re.findall(self.local_locator_re, c.manifest_text())), 1) + + @mock.patch('arvados.keep.KeepClient.refresh_signature') + def test_copy_remote_blocks_on_save(self, rs_mock): + remote_block_loc = "acbd18db4cc2f85cedef654fccc4a4d8+3+Remote-" + "a" * 40 + "@abcdef01" + local_block_loc = "acbd18db4cc2f85cedef654fccc4a4d8+3+A" + "b" * 40 + "@abcdef01" + rs_mock.return_value = local_block_loc + # Remote collection + remote_c = Collection(". " + remote_block_loc + " 0:3:foofile.txt\n") + self.assertEqual( + len(re.findall(self.remote_locator_re, remote_c.manifest_text())), 1) + # Local collection + local_c = Collection() + with local_c.open('barfile.txt', 'wb') as f: + f.write('bar') + local_c.save_new() + self.assertEqual( + len(re.findall(self.local_locator_re, local_c.manifest_text())), 1) + self.assertEqual( + len(re.findall(self.remote_locator_re, local_c.manifest_text())), 0) + # Copy remote file to local collection + local_c.copy('./foofile.txt', './copied/foofile.txt', remote_c) + self.assertEqual( + len(re.findall(self.local_locator_re, local_c.manifest_text())), 1) + self.assertEqual( + len(re.findall(self.remote_locator_re, local_c.manifest_text())), 1) + # Save local collection: remote block should be copied + local_c.save() + rs_mock.assert_called() + self.assertEqual( + len(re.findall(self.local_locator_re, local_c.manifest_text())), 2) + self.assertEqual( + len(re.findall(self.remote_locator_re, local_c.manifest_text())), 0) + + class NewCollectionTestCaseWithServers(run_test_server.TestCaseWithServers): + def test_preserve_version_on_save(self): + c = Collection() + c.save_new(preserve_version=True) + coll_record = arvados.api().collections().get(uuid=c.manifest_locator()).execute() + self.assertEqual(coll_record['version'], 1) + self.assertEqual(coll_record['preserve_version'], True) + with c.open("foo.txt", "wb") as foo: + foo.write(b"foo") + c.save(preserve_version=True) + coll_record = arvados.api().collections().get(uuid=c.manifest_locator()).execute() + self.assertEqual(coll_record['version'], 2) + self.assertEqual(coll_record['preserve_version'], True) + with c.open("bar.txt", "wb") as foo: + foo.write(b"bar") + c.save(preserve_version=False) + coll_record = arvados.api().collections().get(uuid=c.manifest_locator()).execute() + self.assertEqual(coll_record['version'], 3) + self.assertEqual(coll_record['preserve_version'], False) + def test_get_manifest_text_only_committed(self): c = Collection() - with c.open("count.txt", "w") as f: + with c.open("count.txt", "wb") as f: # One file committed - with c.open("foo.txt", "w") as foo: - foo.write("foo") + with c.open("foo.txt", "wb") as foo: + foo.write(b"foo") foo.flush() # Force block commit - f.write("0123456789") + f.write(b"0123456789") # Other file not committed. Block not written to keep yet. self.assertEqual( c._get_manifest_text(".", @@ -1111,21 +1405,79 @@ class NewCollectionTestCaseWithServers(run_test_server.TestCaseWithServers): def test_only_small_blocks_are_packed_together(self): c = Collection() - # Write a couple of small files, - f = c.open("count.txt", "w") - f.write("0123456789") + # Write a couple of small files, + f = c.open("count.txt", "wb") + f.write(b"0123456789") f.close(flush=False) - foo = c.open("foo.txt", "w") - foo.write("foo") + foo = c.open("foo.txt", "wb") + foo.write(b"foo") foo.close(flush=False) # Then, write a big file, it shouldn't be packed with the ones above - big = c.open("bigfile.txt", "w") - big.write("x" * 1024 * 1024 * 33) # 33 MB > KEEP_BLOCK_SIZE/2 + big = c.open("bigfile.txt", "wb") + big.write(b"x" * 1024 * 1024 * 33) # 33 MB > KEEP_BLOCK_SIZE/2 big.close(flush=False) self.assertEqual( c.manifest_text("."), '. 2d303c138c118af809f39319e5d507e9+34603008 a8430a058b8fbf408e1931b794dbd6fb+13 0:34603008:bigfile.txt 34603008:10:count.txt 34603018:3:foo.txt\n') + def test_flush_after_small_block_packing(self): + c = Collection() + # Write a couple of small files, + f = c.open("count.txt", "wb") + f.write(b"0123456789") + f.close(flush=False) + foo = c.open("foo.txt", "wb") + foo.write(b"foo") + foo.close(flush=False) + + self.assertEqual( + c.manifest_text(), + '. a8430a058b8fbf408e1931b794dbd6fb+13 0:10:count.txt 10:3:foo.txt\n') + + f = c.open("count.txt", "rb+") + f.close(flush=True) + + self.assertEqual( + c.manifest_text(), + '. a8430a058b8fbf408e1931b794dbd6fb+13 0:10:count.txt 10:3:foo.txt\n') + + def test_write_after_small_block_packing2(self): + c = Collection() + # Write a couple of small files, + f = c.open("count.txt", "wb") + f.write(b"0123456789") + f.close(flush=False) + foo = c.open("foo.txt", "wb") + foo.write(b"foo") + foo.close(flush=False) + + self.assertEqual( + c.manifest_text(), + '. a8430a058b8fbf408e1931b794dbd6fb+13 0:10:count.txt 10:3:foo.txt\n') + + f = c.open("count.txt", "rb+") + f.write(b"abc") + f.close(flush=False) + + self.assertEqual( + c.manifest_text(), + '. 900150983cd24fb0d6963f7d28e17f72+3 a8430a058b8fbf408e1931b794dbd6fb+13 0:3:count.txt 6:7:count.txt 13:3:foo.txt\n') + + + def test_small_block_packing_with_overwrite(self): + c = Collection() + c.open("b1", "wb").close() + c["b1"].writeto(0, b"b1", 0) + + c.open("b2", "wb").close() + c["b2"].writeto(0, b"b2", 0) + + c["b1"].writeto(0, b"1b", 0) + + self.assertEqual(c.manifest_text(), ". ed4f3f67c70b02b29c50ce1ea26666bd+4 0:2:b1 2:2:b2\n") + self.assertEqual(c["b1"].manifest_text(), ". ed4f3f67c70b02b29c50ce1ea26666bd+4 0:2:b1\n") + self.assertEqual(c["b2"].manifest_text(), ". ed4f3f67c70b02b29c50ce1ea26666bd+4 2:2:b2\n") + class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): MAIN_SERVER = {} @@ -1140,8 +1492,8 @@ class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): 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") + with c.open("count.txt", "wb") as f: + f.write(b"0123456789") self.assertEqual(c.portable_manifest_text(), ". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count.txt\n") @@ -1149,21 +1501,51 @@ class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): 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$",) + c.save(properties={'type' : 'Intermediate'}, + storage_classes=['archive'], + trash_at=datetime.datetime(2111, 1, 1, 11, 11, 11, 111111)) + + self.assertRegex( + c.manifest_text(), + r"^\. 781e5e245d69b566979b86e28d23f2c7\+10\+A[a-f0-9]{40}@[a-f0-9]{8} 0:10:count\.txt$",) + self.assertEqual(c.api_response()["storage_classes_desired"], ['archive']) + self.assertEqual(c.api_response()["properties"], {'type' : 'Intermediate'}) + self.assertEqual(c.api_response()["trash_at"], '2111-01-01T11:11:11.111111000Z') + 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$",) + c.save_new(properties={'type' : 'Intermediate'}, + storage_classes=['archive'], + trash_at=datetime.datetime(2111, 1, 1, 11, 11, 11, 111111)) + + self.assertRegex( + c.manifest_text(), + r"^\. 781e5e245d69b566979b86e28d23f2c7\+10\+A[a-f0-9]{40}@[a-f0-9]{8} 0:10:count\.txt$",) + self.assertEqual(c.api_response()["storage_classes_desired"], ['archive']) + self.assertEqual(c.api_response()["properties"], {'type' : 'Intermediate'}) + self.assertEqual(c.api_response()["trash_at"], '2111-01-01T11:11:11.111111000Z') + + def test_create_and_save_after_commiting(self): + c = self.create_count_txt() + c.save(properties={'type' : 'Intermediate'}, + storage_classes=['hot'], + trash_at=datetime.datetime(2111, 1, 1, 11, 11, 11, 111111)) + c.save(properties={'type' : 'Output'}, + storage_classes=['cold'], + trash_at=datetime.datetime(2222, 2, 2, 22, 22, 22, 222222)) + + self.assertEqual(c.api_response()["storage_classes_desired"], ['cold']) + self.assertEqual(c.api_response()["properties"], {'type' : 'Output'}) + self.assertEqual(c.api_response()["trash_at"], '2222-02-02T22:22:22.222222000Z') 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") + with c2.open("count.txt", "wb") as f: + f.write(b"abcdefg") diff = c1.diff(c2) @@ -1190,8 +1572,8 @@ class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): c1.save() c2 = arvados.collection.Collection(c1.manifest_locator()) - with c2.open("count.txt", "w") as f: - f.write("abcdefg") + with c2.open("count.txt", "wb") as f: + f.write(b"abcdefg") c2.save() @@ -1204,17 +1586,24 @@ class CollectionCreateUpdateTest(run_test_server.TestCaseWithServers): c1 = self.create_count_txt() c1.save() - with c1.open("count.txt", "w") as f: - f.write("XYZ") + with c1.open("count.txt", "wb") as f: + f.write(b"XYZ") c2 = arvados.collection.Collection(c1.manifest_locator()) - with c2.open("count.txt", "w") as f: - f.write("abcdefg") + with c2.open("count.txt", "wb") as f: + f.write(b"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~$") + self.assertRegex( + c1.manifest_text(), + r"\. e65075d550f9b5bf9992fa1d71a131be\+3\S* 7ac66c0f148de9519b8bd264312c4d64\+7\S* 0:3:count\.txt 3:7:count\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") + + def test_pdh_is_native_str(self): + c1 = self.create_count_txt() + pdh = c1.portable_data_hash() + self.assertEqual(type(''), type(pdh)) if __name__ == '__main__':