1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
|
import contextlib
import random
import unittest
import ruffus
import ruffus.drmaa_wrapper
import os
import shutil
import glob
import tempfile
try:
import gevent
HAVE_GEVENT = True
except ImportError:
HAVE_GEVENT = False
try:
import drmaa
HAVE_DRMAA = True
DRMAA_SESSION = drmaa.Session()
DRMAA_SESSION.initialize()
except ImportError:
HAVE_DRMAA = False
ROOT = os.path.abspath(os.path.dirname(__file__))
TESTS_TEMPDIR = os.path.abspath(os.path.join(os.path.dirname(__file__), "tmp"))
# number of cores used for testing parallelism
NUM_CORES = 4
@contextlib.contextmanager
def temp_cd(target):
curdir = os.path.curdir
os.chdir(target)
yield
os.chdir(curdir)
def save_pid(outfile):
with open(outfile + ".pid", "w") as outf:
outf.write("{}".format(os.getpid()))
def create_files(outfile):
with open(outfile, "w") as outf:
outf.write("\n".join(map(
str,
random.sample(range(0, 1000), 100))) + "\n")
save_pid(outfile)
def compute_mean(infile, outfile):
"""compute mean"""
with open(infile, "r") as inf:
n = [float(x.strip()) for x in inf.readlines()]
with open(outfile, "w") as outf:
outf.write("{}\n".format(sum(n) / len(n)))
save_pid(outfile)
def run_job(infile, outfile, **kwargs):
if not kwargs.get("run_locally", False):
kwargs["drmaa_session"] = DRMAA_SESSION
stdout, stderr = ruffus.drmaa_wrapper.run_job(
cmd_str="hostname > {}".format(os.path.abspath(outfile)),
verbose=1,
local_echo=False,
**kwargs)
def run_local_job1(*args):
return run_job(*args, run_locally=True)
def run_local_job2(*args):
return run_job(*args, run_locally=True)
def run_remote_job1(*args):
return run_job(*args, run_locally=False)
def run_remote_job2(*args):
return run_job(*args, run_locally=False)
def combine_means(infiles, outfile):
with open(outfile, "w") as outf:
for infile in infiles:
with open(infile) as inf:
outf.write(inf.read())
save_pid(outfile)
class BaseTest(unittest.TestCase):
expected_output_files = ["sample_{:02}.mean".format(x) for x in range(10)] +\
["sample_{:02}.txt".format(x) for x in range(10)]
def setUp(self):
try:
os.makedirs(TESTS_TEMPDIR)
except OSError:
pass
self.work_dir = tempfile.mkdtemp(suffix="",
prefix="ruffus_tmp_{}_".format(
self.id()),
dir=TESTS_TEMPDIR)
def tearDown(self):
# shutil.rmtree(self.work_dir)
pass
def check_files(self, present=[], absent=[]):
for fn in present:
path = os.path.join(self.work_dir, fn)
self.assertTrue(os.path.exists(path),
"file {} does not exist".format(path))
for fn in absent:
path = os.path.join(self.work_dir, fn)
self.assertFalse(os.path.exists(path),
"file {} does exist but not expected".format(path))
def build_pipeline(self, pipeline_name, **kwargs):
# fudge: clear all previous pipelines
ruffus.Pipeline.clear_all()
pipeline = ruffus.Pipeline(pipeline_name)
task_create_files = pipeline.originate(
task_func=create_files,
output=["sample_{:02}.txt".format(x) for x in range(10)])
task_compute_mean = pipeline.transform(
task_func=compute_mean,
input=task_create_files,
filter=ruffus.suffix(".txt"),
output=".mean")
task_combine_means = pipeline.merge(
task_func=combine_means,
input=task_compute_mean,
output="means.txt")
task_run_local_job1 = pipeline.transform(
task_func=run_local_job1,
input=task_create_files,
filter=ruffus.suffix(".txt"),
output=".local1")
# test jobs_limit with local running
task_run_local_job2 = pipeline.transform(
task_func=run_local_job2,
input=task_create_files,
filter=ruffus.suffix(".txt"),
output=".local2").jobs_limit(NUM_CORES // 2)
# multiprocessing and DRMAA do not work at the moment likely
# cause is the shared session object.
if not HAVE_DRMAA or (kwargs.get("multiprocess", 1) > 1):
return
task_run_remote_job1 = pipeline.transform(
task_func=run_remote_job1,
input=task_create_files,
filter=ruffus.suffix(".txt"),
output=".remote1")
# test jobs_limit with remote running
task_run_remote_job2 = pipeline.transform(
task_func=run_remote_job2,
input=task_create_files,
filter=ruffus.suffix(".txt"),
output=".remote2").jobs_limit(NUM_CORES // 2)
def run_pipeline(self, **kwargs):
pipeline = self.build_pipeline(self.id(), **kwargs)
with temp_cd(self.work_dir):
ruffus.pipeline_run(pipeline=pipeline, verbose=5, **kwargs)
self.check_files(self.expected_output_files)
def read_pids(self):
pids = []
for fn in glob.glob(os.path.join(self.work_dir, "*.pid")):
with open(fn) as inf:
pids.append(int(inf.readline().strip()))
return pids
class TestExecutionEngines(BaseTest):
def test_pipeline_runs_with_multiprocessing(self):
self.run_pipeline(multiprocess=NUM_CORES)
pids = self.read_pids()
self.assertEqual(len(set(pids)), NUM_CORES)
def test_pipeline_runs_with_multithreading(self):
self.run_pipeline(multithread=NUM_CORES)
pids = self.read_pids()
self.assertEqual(len(set(pids)), 1)
self.assertEqual(pids[0], os.getpid())
@unittest.skipIf(not HAVE_GEVENT, "no gevent installed")
def test_pipeline_runs_with_gevent_manager(self):
self.run_pipeline(multithread=NUM_CORES, pool_manager="gevent")
pids = self.read_pids()
self.assertEqual(len(set(pids)), 1)
self.assertEqual(pids[0], os.getpid())
def test_pipeline_fails_with_unknown_manager(self):
self.assertRaises(ValueError,
self.run_pipeline,
multithread=NUM_CORES,
pool_manager="mystical_manager")
if __name__ == "__main__":
unittest.main()
|
