X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/69594694568da1abc2bd8e46134da81ec1f2a7c8..ebb2559b3a09636ff687316bbe512e0e8a86b168:/sdk/python/tests/test_collections.py diff --git a/sdk/python/tests/test_collections.py b/sdk/python/tests/test_collections.py index fb29881223..0e3d5e13f1 100644 --- a/sdk/python/tests/test_collections.py +++ b/sdk/python/tests/test_collections.py @@ -3,18 +3,17 @@ # ARVADOS_API_TOKEN=abc ARVADOS_API_HOST=arvados.local python -m unittest discover import arvados -import bz2 import copy -import hashlib import mock import os import pprint import re -import subprocess 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 class TestResumableWriter(arvados.ResumableCollectionWriter): @@ -124,25 +123,6 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, [2, '.', 'ob.txt', 'ob'], [0, '.', 'zero.txt', '']]) - def _test_readline(self, what_in, what_out): - cw = arvados.CollectionWriter(self.api_client) - cw.start_new_file('test.txt') - cw.write(what_in) - test1 = cw.finish() - cr = arvados.CollectionReader(test1, self.api_client) - got = [] - for x in list(cr.all_files())[0].readlines(): - got += [x] - self.assertEqual(got, - what_out, - "readlines did not split lines correctly: %s" % got) - - def test_collection_readline(self): - self._test_readline("\na\nbcd\n\nefg\nz", - ["\n", "a\n", "bcd\n", "\n", "efg\n", "z"]) - self._test_readline("ab\ncd\n", - ["ab\n", "cd\n"]) - def test_collection_empty_file(self): cw = arvados.CollectionWriter(self.api_client) cw.start_new_file('zero.txt') @@ -179,53 +159,6 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, got_sizes += [f.size()] self.assertEqual(got_sizes, expect_sizes, "got wrong file sizes %s, expected %s" % (got_sizes, expect_sizes)) - def test_collection_bz2_decompression(self): - n_lines_in = 2**18 - data_in = "abc\n" - for x in xrange(0, 18): - data_in += data_in - compressed_data_in = bz2.compress(data_in) - cw = arvados.CollectionWriter(self.api_client) - cw.start_new_file('test.bz2') - cw.write(compressed_data_in) - bz2_manifest = cw.manifest_text() - - cr = arvados.CollectionReader(bz2_manifest, self.api_client) - - got = 0 - for x in list(cr.all_files())[0].readlines(): - self.assertEqual(x, "abc\n", "decompression returned wrong data: %s" % x) - got += 1 - self.assertEqual(got, - n_lines_in, - "decompression returned %d lines instead of %d" % (got, n_lines_in)) - - def test_collection_gzip_decompression(self): - n_lines_in = 2**18 - data_in = "abc\n" - for x in xrange(0, 18): - data_in += data_in - p = subprocess.Popen(["gzip", "-1cn"], - stdout=subprocess.PIPE, - stdin=subprocess.PIPE, - stderr=subprocess.PIPE, - shell=False, close_fds=True) - compressed_data_in, stderrdata = p.communicate(data_in) - - cw = arvados.CollectionWriter(self.api_client) - cw.start_new_file('test.gz') - cw.write(compressed_data_in) - gzip_manifest = cw.manifest_text() - - cr = arvados.CollectionReader(gzip_manifest, self.api_client) - got = 0 - for x in list(cr.all_files())[0].readlines(): - self.assertEqual(x, "abc\n", "decompression returned wrong data: %s" % x) - got += 1 - self.assertEqual(got, - n_lines_in, - "decompression returned %d lines instead of %d" % (got, n_lines_in)) - def test_normalized_collection(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt . 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt @@ -278,170 +211,97 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, self.assertEqual(arvados.CollectionReader(m8, self.api_client).manifest_text(normalize=True), m8) def test_locators_and_ranges(self): - blocks2 = [['a', 10, 0], - ['b', 10, 10], - ['c', 10, 20], - ['d', 10, 30], - ['e', 10, 40], - ['f', 10, 50]] - - self.assertEqual(arvados.locators_and_ranges(blocks2, 2, 2), [['a', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 12, 2), [['b', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 22, 2), [['c', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 32, 2), [['d', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 42, 2), [['e', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 52, 2), [['f', 10, 2, 2]]) + blocks2 = [Range('a', 0, 10), + Range('b', 10, 10), + Range('c', 20, 10), + Range('d', 30, 10), + Range('e', 40, 10), + Range('f', 50, 10)] + + self.assertEqual(arvados.locators_and_ranges(blocks2, 2, 2), [LocatorAndRange('a', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 12, 2), [LocatorAndRange('b', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 22, 2), [LocatorAndRange('c', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 32, 2), [LocatorAndRange('d', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 42, 2), [LocatorAndRange('e', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 52, 2), [LocatorAndRange('f', 10, 2, 2)]) self.assertEqual(arvados.locators_and_ranges(blocks2, 62, 2), []) self.assertEqual(arvados.locators_and_ranges(blocks2, -2, 2), []) - self.assertEqual(arvados.locators_and_ranges(blocks2, 0, 2), [['a', 10, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 10, 2), [['b', 10, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 20, 2), [['c', 10, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 30, 2), [['d', 10, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 40, 2), [['e', 10, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 50, 2), [['f', 10, 0, 2]]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 0, 2), [LocatorAndRange('a', 10, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 10, 2), [LocatorAndRange('b', 10, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 20, 2), [LocatorAndRange('c', 10, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 30, 2), [LocatorAndRange('d', 10, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 40, 2), [LocatorAndRange('e', 10, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 50, 2), [LocatorAndRange('f', 10, 0, 2)]) self.assertEqual(arvados.locators_and_ranges(blocks2, 60, 2), []) self.assertEqual(arvados.locators_and_ranges(blocks2, -2, 2), []) - self.assertEqual(arvados.locators_and_ranges(blocks2, 9, 2), [['a', 10, 9, 1], ['b', 10, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 19, 2), [['b', 10, 9, 1], ['c', 10, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 29, 2), [['c', 10, 9, 1], ['d', 10, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 39, 2), [['d', 10, 9, 1], ['e', 10, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 49, 2), [['e', 10, 9, 1], ['f', 10, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks2, 59, 2), [['f', 10, 9, 1]]) - - - blocks3 = [['a', 10, 0], - ['b', 10, 10], - ['c', 10, 20], - ['d', 10, 30], - ['e', 10, 40], - ['f', 10, 50], - ['g', 10, 60]] - - self.assertEqual(arvados.locators_and_ranges(blocks3, 2, 2), [['a', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 12, 2), [['b', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 22, 2), [['c', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 32, 2), [['d', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 42, 2), [['e', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 52, 2), [['f', 10, 2, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks3, 62, 2), [['g', 10, 2, 2]]) - - - blocks = [['a', 10, 0], - ['b', 15, 10], - ['c', 5, 25]] + self.assertEqual(arvados.locators_and_ranges(blocks2, 9, 2), [LocatorAndRange('a', 10, 9, 1), LocatorAndRange('b', 10, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 19, 2), [LocatorAndRange('b', 10, 9, 1), LocatorAndRange('c', 10, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 29, 2), [LocatorAndRange('c', 10, 9, 1), LocatorAndRange('d', 10, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 39, 2), [LocatorAndRange('d', 10, 9, 1), LocatorAndRange('e', 10, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 49, 2), [LocatorAndRange('e', 10, 9, 1), LocatorAndRange('f', 10, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks2, 59, 2), [LocatorAndRange('f', 10, 9, 1)]) + + + blocks3 = [Range('a', 0, 10), + Range('b', 10, 10), + Range('c', 20, 10), + Range('d', 30, 10), + Range('e', 40, 10), + Range('f', 50, 10), + Range('g', 60, 10)] + + self.assertEqual(arvados.locators_and_ranges(blocks3, 2, 2), [LocatorAndRange('a', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 12, 2), [LocatorAndRange('b', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 22, 2), [LocatorAndRange('c', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 32, 2), [LocatorAndRange('d', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 42, 2), [LocatorAndRange('e', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 52, 2), [LocatorAndRange('f', 10, 2, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks3, 62, 2), [LocatorAndRange('g', 10, 2, 2)]) + + + blocks = [Range('a', 0, 10), + Range('b', 10, 15), + Range('c', 25, 5)] self.assertEqual(arvados.locators_and_ranges(blocks, 1, 0), []) - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 5), [['a', 10, 0, 5]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 3, 5), [['a', 10, 3, 5]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 10), [['a', 10, 0, 10]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 11), [['a', 10, 0, 10], - ['b', 15, 0, 1]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 1, 11), [['a', 10, 1, 9], - ['b', 15, 0, 2]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 25), [['a', 10, 0, 10], - ['b', 15, 0, 15]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 30), [['a', 10, 0, 10], - ['b', 15, 0, 15], - ['c', 5, 0, 5]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 1, 30), [['a', 10, 1, 9], - ['b', 15, 0, 15], - ['c', 5, 0, 5]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 0, 31), [['a', 10, 0, 10], - ['b', 15, 0, 15], - ['c', 5, 0, 5]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 15, 5), [['b', 15, 5, 5]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 8, 17), [['a', 10, 8, 2], - ['b', 15, 0, 15]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 8, 20), [['a', 10, 8, 2], - ['b', 15, 0, 15], - ['c', 5, 0, 3]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 26, 2), [['c', 5, 1, 2]]) - - self.assertEqual(arvados.locators_and_ranges(blocks, 9, 15), [['a', 10, 9, 1], - ['b', 15, 0, 14]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 10, 15), [['b', 15, 0, 15]]) - self.assertEqual(arvados.locators_and_ranges(blocks, 11, 15), [['b', 15, 1, 14], - ['c', 5, 0, 1]]) - - class MockStreamReader(object): - def __init__(self, content): - self.content = content - self.num_retries = 0 - - def readfrom(self, start, size, num_retries=0): - return self.content[start:start+size] - - def test_file_stream(self): - content = 'abcdefghijklmnopqrstuvwxyz0123456789' - msr = self.MockStreamReader(content) - segments = [[0, 10, 0], - [10, 15, 10], - [25, 5, 25]] - - sfr = arvados.StreamFileReader(msr, segments, "test") - - self.assertEqual(sfr.name(), "test") - self.assertEqual(sfr.size(), 30) - - self.assertEqual(sfr.readfrom(0, 30), content[0:30]) - self.assertEqual(sfr.readfrom(2, 30), content[2:30]) - - self.assertEqual(sfr.readfrom(2, 8), content[2:10]) - self.assertEqual(sfr.readfrom(0, 10), content[0:10]) - - self.assertEqual(sfr.tell(), 0) - self.assertEqual(sfr.read(5), content[0:5]) - self.assertEqual(sfr.tell(), 5) - self.assertEqual(sfr.read(5), content[5:10]) - self.assertEqual(sfr.tell(), 10) - self.assertEqual(sfr.read(5), content[10:15]) - self.assertEqual(sfr.tell(), 15) - self.assertEqual(sfr.read(5), content[15:20]) - self.assertEqual(sfr.tell(), 20) - self.assertEqual(sfr.read(5), content[20:25]) - self.assertEqual(sfr.tell(), 25) - self.assertEqual(sfr.read(5), content[25:30]) - self.assertEqual(sfr.tell(), 30) - self.assertEqual(sfr.read(5), '') - self.assertEqual(sfr.tell(), 30) - - segments = [[26, 10, 0], - [0, 15, 10], - [15, 5, 25]] - - sfr = arvados.StreamFileReader(msr, segments, "test") - - self.assertEqual(sfr.size(), 30) - - self.assertEqual(sfr.readfrom(0, 30), content[26:36] + content[0:20]) - self.assertEqual(sfr.readfrom(2, 30), content[28:36] + content[0:20]) - - self.assertEqual(sfr.readfrom(2, 8), content[28:36]) - self.assertEqual(sfr.readfrom(0, 10), content[26:36]) - - self.assertEqual(sfr.tell(), 0) - self.assertEqual(sfr.read(5), content[26:31]) - self.assertEqual(sfr.tell(), 5) - self.assertEqual(sfr.read(5), content[31:36]) - self.assertEqual(sfr.tell(), 10) - self.assertEqual(sfr.read(5), content[0:5]) - self.assertEqual(sfr.tell(), 15) - self.assertEqual(sfr.read(5), content[5:10]) - self.assertEqual(sfr.tell(), 20) - self.assertEqual(sfr.read(5), content[10:15]) - self.assertEqual(sfr.tell(), 25) - self.assertEqual(sfr.read(5), content[15:20]) - self.assertEqual(sfr.tell(), 30) - self.assertEqual(sfr.read(5), '') - self.assertEqual(sfr.tell(), 30) - + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 5), [LocatorAndRange('a', 10, 0, 5)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 3, 5), [LocatorAndRange('a', 10, 3, 5)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 10), [LocatorAndRange('a', 10, 0, 10)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 11), [LocatorAndRange('a', 10, 0, 10), + LocatorAndRange('b', 15, 0, 1)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 1, 11), [LocatorAndRange('a', 10, 1, 9), + LocatorAndRange('b', 15, 0, 2)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 25), [LocatorAndRange('a', 10, 0, 10), + LocatorAndRange('b', 15, 0, 15)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 30), [LocatorAndRange('a', 10, 0, 10), + LocatorAndRange('b', 15, 0, 15), + LocatorAndRange('c', 5, 0, 5)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 1, 30), [LocatorAndRange('a', 10, 1, 9), + LocatorAndRange('b', 15, 0, 15), + LocatorAndRange('c', 5, 0, 5)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 0, 31), [LocatorAndRange('a', 10, 0, 10), + LocatorAndRange('b', 15, 0, 15), + LocatorAndRange('c', 5, 0, 5)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 15, 5), [LocatorAndRange('b', 15, 5, 5)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 8, 17), [LocatorAndRange('a', 10, 8, 2), + LocatorAndRange('b', 15, 0, 15)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 8, 20), [LocatorAndRange('a', 10, 8, 2), + LocatorAndRange('b', 15, 0, 15), + LocatorAndRange('c', 5, 0, 3)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 26, 2), [LocatorAndRange('c', 5, 1, 2)]) + + self.assertEqual(arvados.locators_and_ranges(blocks, 9, 15), [LocatorAndRange('a', 10, 9, 1), + LocatorAndRange('b', 15, 0, 14)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 10, 15), [LocatorAndRange('b', 15, 0, 15)]) + self.assertEqual(arvados.locators_and_ranges(blocks, 11, 15), [LocatorAndRange('b', 15, 1, 14), + LocatorAndRange('c', 5, 0, 1)]) class MockKeep(object): def __init__(self, content, num_retries=0): @@ -474,30 +334,6 @@ class ArvadosCollectionsTest(run_test_server.TestCaseWithServers, self.assertEqual(sr.readfrom(25, 5), content[25:30]) self.assertEqual(sr.readfrom(30, 5), '') - def test_file_reader(self): - keepblocks = {'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa+10': 'abcdefghij', - 'bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb+15': 'klmnopqrstuvwxy', - 'cccccccccccccccccccccccccccccccc+5': 'z0123'} - mk = self.MockKeep(keepblocks) - - sr = arvados.StreamReader([".", "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa+10", "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb+15", "cccccccccccccccccccccccccccccccc+5", "0:10:foo", "15:10:foo"], mk) - - content = 'abcdefghijpqrstuvwxy' - - f = sr.files()["foo"] - - # f.read() calls will be aligned on block boundaries (as a - # result of ticket #3663). - - f.seek(0) - self.assertEqual(f.read(20), content[0:10]) - - f.seek(0) - self.assertEqual(f.read(6), content[0:6]) - self.assertEqual(f.read(6), content[6:10]) - self.assertEqual(f.read(6), content[10:16]) - self.assertEqual(f.read(6), content[16:20]) - def test_extract_file(self): m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt . 085c37f02916da1cad16f93c54d899b7+41 0:41:md6sum.txt @@ -558,7 +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) @@ -596,7 +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): @@ -665,17 +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'] @@ -685,20 +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 @@ -708,9 +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): @@ -729,14 +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()) @@ -746,7 +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()) @@ -754,17 +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( @@ -776,7 +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())) @@ -801,16 +614,69 @@ class CollectionReaderTestCase(unittest.TestCase, CollectionTestMixin): [[f.size(), f.stream_name(), f.name()] for f in reader.all_streams()[0].all_files()]) + def test_read_empty_collection(self): + client = self.api_client_mock(200) + self.mock_get_collection(client, 200, 'empty') + reader = arvados.CollectionReader('d41d8cd98f00b204e9800998ecf8427e+0', + api_client=client) + self.assertEqual('', reader.manifest_text()) + + 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()) + self.assertEqual(file_name, coll_file.name) + self.assertEqual(file_size, coll_file.size()) + + 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') + 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) + 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') + self.check_open_file(cfile, './subdir2/subdir3', 'file2_in_subdir3.txt', + 32) + + def test_open_nonexistent_stream(self): + client = self.api_client_mock(200) + reader = arvados.CollectionReader(self.DEFAULT_UUID, api_client=client) + 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(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 @@ -826,6 +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, @@ -839,6 +731,491 @@ class CollectionWriterTestCase(unittest.TestCase, CollectionTestMixin): writer.flush_data() self.assertEqual(self.DEFAULT_MANIFEST, writer.manifest_text()) + def test_one_open(self): + client = self.api_client_mock() + writer = arvados.CollectionWriter(client) + 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') + 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: + self.assertEqual(". {} 0:9:out\n".format(data_loc), + writer.manifest_text()) + + def test_open_writelines(self): + client = self.api_client_mock() + writer = arvados.CollectionWriter(client) + with writer.open('six') as out_file: + out_file.writelines(['12', '34', '56']) + 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 = 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: + out_file.write('flush1') + out_file.flush() + out_file.write('flush2') + self.assertEqual(". {} {} 0:12:flush_test\n".format(data_loc1, + data_loc2), + writer.manifest_text()) + + def test_two_opens_same_stream(self): + client = self.api_client_mock() + writer = arvados.CollectionWriter(client) + with writer.open('.', '1') as out_file: + out_file.write('1st') + with writer.open('.', '2') as out_file: + out_file.write('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), + writer.manifest_text()) + + def test_two_opens_two_streams(self): + client = self.api_client_mock() + 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: + out_file.write('file') + with writer.open('./dir', 'indir') as out_file: + out_file.write('indir') + expected = ". {} 0:4:file\n./dir {} 0:5:indir\n".format( + data_loc1, data_loc2) + self.assertEqual(expected, writer.manifest_text()) + + def test_dup_open_fails(self): + client = self.api_client_mock() + writer = arvados.CollectionWriter(client) + file1 = writer.open('one') + self.assertRaises(arvados.errors.AssertionError, writer.open, 'two') + + +class NewCollectionTestCase(unittest.TestCase, CollectionTestMixin): + + def test_replication_desired_kept_on_load(self): + m = '. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n' + c1 = Collection(m, replication_desired=1) + c1.save_new() + loc = c1.manifest_locator() + c2 = Collection(loc) + self.assertEqual(c1.manifest_text, c2.manifest_text) + self.assertEqual(c1.replication_desired, c2.replication_desired) + + def test_replication_desired_not_loaded_if_provided(self): + m = '. 781e5e245d69b566979b86e28d23f2c7+10 0:10:count1.txt 0:10:count2.txt\n' + c1 = Collection(m, replication_desired=1) + c1.save_new() + loc = c1.manifest_locator() + c2 = Collection(loc, replication_desired=2) + self.assertEqual(c1.manifest_text, c2.manifest_text) + self.assertNotEqual(c1.replication_desired, c2.replication_desired) + + 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(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_with_collision(self): + m1 = """. 5348b82a029fd9e971a811ce1f71360b+43 0:43:md5sum.txt +./md5sum.txt 085c37f02916da1cad16f93c54d899b7+41 0:41:md5sum.txt +""" + 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): + 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("") + self.assertIs(c.find("./nonexistant.txt"), None) + self.assertIs(c.find("./nonexistantsubdir/nonexistant.txt"), None) + + 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()) + + def test_remove_empty_subdir(self): + 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): + 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): + 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): + 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, [('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + d = c1.diff(c2) + self.assertEqual(d, [('tok', './count1.txt', c2["count1.txt"], c1["count1.txt"])]) + + 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"]), + ('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.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"]), + ('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.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")), + ('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.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"]), + ('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.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 NewCollectionTestCaseWithServers(run_test_server.TestCaseWithServers): + def test_get_manifest_text_only_committed(self): + c = Collection() + with c.open("count.txt", "w") as f: + # One file committed + with c.open("foo.txt", "w") as foo: + foo.write("foo") + foo.flush() # Force block commit + f.write("0123456789") + # Other file not committed. Block not written to keep yet. + self.assertEqual( + c._get_manifest_text(".", + strip=False, + normalize=False, + only_committed=True), + '. acbd18db4cc2f85cedef654fccc4a4d8+3 0:0:count.txt 0:3:foo.txt\n') + # And now with the file closed... + f.flush() # Force block commit + self.assertEqual( + c._get_manifest_text(".", + strip=False, + normalize=False, + only_committed=True), + ". 781e5e245d69b566979b86e28d23f2c7+10 acbd18db4cc2f85cedef654fccc4a4d8+3 0:10:count.txt 10:3:foo.txt\n") + + 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") + f.close(flush=False) + foo = c.open("foo.txt", "w") + foo.write("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.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') + + +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$",) + + def test_create_diff_apply(self): + c1 = self.create_count_txt() + c1.save() + + c2 = Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + + diff = c1.diff(c2) + + self.assertEqual(diff[0], (arvados.collection.MOD, u'./count.txt', c1["count.txt"], c2["count.txt"])) + + c1.apply(diff) + self.assertEqual(c1.portable_data_hash(), c2.portable_data_hash()) + + def test_diff_apply_with_token(self): + baseline = CollectionReader(". 781e5e245d69b566979b86e28d23f2c7+10+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:10:count.txt\n") + c = Collection(". 781e5e245d69b566979b86e28d23f2c7+10 0:10:count.txt\n") + other = CollectionReader(". 7ac66c0f148de9519b8bd264312c4d64+7+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:7:count.txt\n") + + diff = baseline.diff(other) + self.assertEqual(diff, [('mod', u'./count.txt', c["count.txt"], other["count.txt"])]) + + c.apply(diff) + + self.assertEqual(c.manifest_text(), ". 7ac66c0f148de9519b8bd264312c4d64+7+A715fd31f8111894f717eb1003c1b0216799dd9ec@54f5dd1a 0:7:count.txt\n") + + + def test_create_and_update(self): + c1 = self.create_count_txt() + c1.save() + + c2 = arvados.collection.Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + + c2.save() + + self.assertNotEqual(c1.portable_data_hash(), c2.portable_data_hash()) + c1.update() + self.assertEqual(c1.portable_data_hash(), c2.portable_data_hash()) + + + def test_create_and_update_with_conflict(self): + c1 = self.create_count_txt() + c1.save() + + with c1.open("count.txt", "w") as f: + f.write("XYZ") + + c2 = arvados.collection.Collection(c1.manifest_locator()) + with c2.open("count.txt", "w") as f: + f.write("abcdefg") + + c2.save() + + c1.update() + self.assertRegexpMatches(c1.manifest_text(), r"\. e65075d550f9b5bf9992fa1d71a131be\+3\S* 7ac66c0f148de9519b8bd264312c4d64\+7\S* 0:3:count\.txt 3:7:count\.txt~\d\d\d\d\d\d\d\d-\d\d\d\d\d\d~conflict~$") + if __name__ == '__main__': unittest.main()