def test_submitted_to_insdc_status(self):
"""submitted_to_insdc_status works correctly"""
# False case
pd = PreprocessedData(1)
self.assertEqual(pd.submitted_to_insdc_status(), 'submitting')
# True case
pd = PreprocessedData(2)
self.assertEqual(pd.submitted_to_insdc_status(), 'not submitted')
def test_is_submitted_to_insdc(self):
"""is_submitted_to_insdc works correctly"""
# False case
pd = PreprocessedData(1)
self.assertTrue(pd.is_submitted_to_insdc())
# True case
pd = PreprocessedData(2)
self.assertFalse(pd.is_submitted_to_insdc())
def post(self, preprocessed_data_id):
user = self.current_user
# make sure user is admin and can therefore actually submit to EBI
if user.level != 'admin':
raise HTTPError(403, "User %s cannot submit to EBI!" %
user.id)
submission_type = self.get_argument('submission_type')
if submission_type not in ['ADD', 'MODIFY']:
raise HTTPError(403, "User: %s, %s is not a recognized submission "
"type" % (user.id, submission_type))
msg = ''
msg_level = 'success'
preprocessed_data = PreprocessedData(preprocessed_data_id)
state = preprocessed_data.submitted_to_insdc_status()
if state == 'submitting':
msg = "Cannot resubmit! Current state is: %s" % state
msg_level = 'danger'
elif state == 'success' and submission_type == "ADD":
msg = "Cannot resubmit! Current state is: %s, use MODIFY" % state
msg_level = 'danger'
else:
channel = user.id
job_id = submit(channel, submit_to_ebi, int(preprocessed_data_id),
submission_type)
self.render('compute_wait.html',
job_id=job_id, title='EBI Submission',
completion_redirect='/compute_complete/%s' % job_id)
return
self.display_template(preprocessed_data_id, msg, msg_level)
def post(self, preprocessed_data_id):
# make sure user is admin and can therefore actually submit to VAMPS
if self.current_user.level != 'admin':
raise HTTPError(403, "User %s cannot submit to VAMPS!" %
self.current_user.id)
msg = ''
msg_level = 'success'
preprocessed_data = PreprocessedData(preprocessed_data_id)
state = preprocessed_data.submitted_to_vamps_status()
demux = [path for _, path, ftype in preprocessed_data.get_filepaths()
if ftype == 'preprocessed_demux']
demux_length = len(demux)
if state in ('submitting', 'success'):
msg = "Cannot resubmit! Current state is: %s" % state
msg_level = 'danger'
elif demux_length != 1:
msg = "The study doesn't have demux files or have too many" % state
msg_level = 'danger'
else:
channel = self.current_user.id
job_id = submit(channel, submit_to_VAMPS,
int(preprocessed_data_id))
self.render('compute_wait.html',
job_id=job_id, title='VAMPS Submission',
completion_redirect='/compute_complete/%s' % job_id)
return
self.display_template(preprocessed_data_id, msg, msg_level)
def test_get_process_target_gene_cmd(self):
preprocessed_data = PreprocessedData(1)
params = ProcessedSortmernaParams(1)
obs_cmd, obs_output_dir = _get_process_target_gene_cmd(
preprocessed_data, params)
_, ref_dir = get_mountpoint('reference')[0]
_, preprocessed_dir = get_mountpoint('preprocessed_data')[0]
exp_cmd = ("pick_closed_reference_otus.py -i {}1_seqs.fna -r "
"{}GreenGenes_13_8_97_otus.fasta -o {} -p placeholder -t "
"{}GreenGenes_13_8_97_otu_taxonomy.txt".format(
preprocessed_dir, ref_dir, obs_output_dir, ref_dir))
obs_tokens = obs_cmd.split()[::-1]
exp_tokens = exp_cmd.split()[::-1]
self.assertEqual(len(obs_tokens), len(exp_tokens))
while obs_tokens:
o_t = obs_tokens.pop()
e_t = exp_tokens.pop()
if o_t == '-p':
# skip parameters file
obs_tokens.pop()
exp_tokens.pop()
else:
self.assertEqual(o_t, e_t)
def render(self, study):
avail_ppd = [(ppd_id, PreprocessedData(ppd_id))
for ppd_id in study.preprocessed_data()]
return self.render_string(
"study_description_templates/preprocessed_data_tab.html",
available_preprocessed_data=avail_ppd,
study_id=study.id)
def test_get_filepaths(self):
"""Correctly returns the filepaths to the preprocessed files"""
ppd = PreprocessedData(1)
obs = ppd.get_filepaths()
exp = [(join(self.db_test_ppd_dir,
'1_seqs.fna'), "preprocessed_sequences"),
(join(self.db_test_ppd_dir,
'1_seqs.qual'), "preprocessed_sequences_qual")]
self.assertEqual(obs, exp)
def submit_VAMPS(preprocessed_data_id):
"""Submit preprocessed data to VAMPS
Parameters
----------
preprocessed_data_id : int
The preprocesssed data id
"""
preprocessed_data = PreprocessedData(preprocessed_data_id)
study = Study(preprocessed_data.study)
sample_template = SampleTemplate(study.sample_template)
prep_template = PrepTemplate(preprocessed_data.prep_template)
status = preprocessed_data.submitted_to_vamps_status()
if status in ('submitting', 'success'):
raise ValueError("Cannot resubmit! Current status is: %s" % status)
preprocessed_data.update_vamps_status('submitting')
# Generating a tgz
targz_folder = mkdtemp(prefix=qiita_config.working_dir)
targz_fp = join(
targz_folder,
'%d_%d_%d.tgz' % (study.id, prep_template.id, preprocessed_data.id))
targz = taropen(targz_fp, mode='w:gz')
# adding sample/prep
samp_fp = join(targz_folder, 'sample_metadata.txt')
sample_template.to_file(samp_fp)
targz.add(samp_fp, arcname='sample_metadata.txt')
prep_fp = join(targz_folder, 'prep_metadata.txt')
prep_template.to_file(prep_fp)
targz.add(prep_fp, arcname='prep_metadata.txt')
# adding preprocessed data
for _, fp, fp_type in preprocessed_data.get_filepaths():
if fp_type == 'preprocessed_fasta':
targz.add(fp, arcname='preprocessed_fasta.fna')
targz.close()
# submitting
cmd = ("curl -F user=%s -F pass='******' -F uploadFile=@%s -F "
"press=UploadFile %s" %
(qiita_config.vamps_user, qiita_config.vamps_pass, targz_fp,
qiita_config.vamps_url))
obs, _, _ = system_call(cmd)
exp = ("<html>\n<head>\n<title>Process Uploaded File</title>\n</head>\n"
"<body>\n</body>\n</html>")
if obs != exp:
preprocessed_data.update_vamps_status('failure')
return False
else:
preprocessed_data.update_vamps_status('success')
return True
def test_link_filepaths_status_setter(self):
ppd = PreprocessedData(1)
self.assertEqual(ppd.link_filepaths_status, 'idle')
ppd._set_link_filepaths_status('linking')
self.assertEqual(ppd.link_filepaths_status, 'linking')
ppd._set_link_filepaths_status('unlinking')
self.assertEqual(ppd.link_filepaths_status, 'unlinking')
ppd._set_link_filepaths_status('failed: error')
self.assertEqual(ppd.link_filepaths_status, 'failed: error')
def test_get_filepaths(self):
"""Correctly returns the filepaths to the preprocessed files"""
ppd = PreprocessedData(1)
obs = ppd.get_filepaths()
exp = [(5, join(self.db_test_ppd_dir, '1_seqs.fna'),
"preprocessed_fasta"),
(6, join(self.db_test_ppd_dir, '1_seqs.qual'),
"preprocessed_fastq"),
(7, join(self.db_test_ppd_dir, '1_seqs.demux'),
"preprocessed_demux")]
self.assertEqual(obs, exp)
def test_processing_status(self):
"""processing_status works correctly"""
# Processed case
ppd = PreprocessedData(1)
self.assertEqual(ppd.processing_status, 'not_processed')
# not processed case
ppd = PreprocessedData.create(self.study, self.params_table,
self.params_id, self.filepaths,
data_type="18S")
self.assertEqual(ppd.processing_status, 'not_processed')
def test_update_preprocessed_data_from_cmd(self):
exp_ppd = PreprocessedData(Study(1).preprocessed_data()[0])
exp_fps = exp_ppd.get_filepaths()
# The original paths mush exist, but they're not included in the test
# so create them here
for _, fp, _ in exp_fps:
with open(fp, 'w') as f:
f.write("")
next_fp_id = get_count('qiita.filepath') + 1
exp_fps.append((next_fp_id,
join(self.db_ppd_dir,
"%s_split_library_log.txt" % exp_ppd.id), 'log'))
ppd = update_preprocessed_data_from_cmd(self.test_slo, 1)
# Check that the modified preprocessed data is the correct one
self.assertEqual(ppd.id, exp_ppd.id)
# Check that the filepaths returned are correct
# We need to sort the list returned from the db because the ordering
# on that list is based on db modification time, rather than id
obs_fps = sorted(ppd.get_filepaths())
self.assertEqual(obs_fps, exp_fps)
# Check that the checksums have been updated
sql = "SELECT checksum FROM qiita.filepath WHERE filepath_id=%s"
# Checksum of the fasta file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[0][0], ))[0]
self.assertEqual(obs_checksum, '3532748626')
# Checksum of the fastq file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[1][0], ))[0]
self.assertEqual(obs_checksum, '2958832064')
# Checksum of the demux file
# The checksum is generated dynamically, so the checksum changes
# We are going to test that the checksum is not the one that was
# before, which corresponds to an empty file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[2][0], ))[0]
self.assertTrue(isinstance(obs_checksum, str))
self.assertNotEqual(obs_checksum, '852952723')
self.assertTrue(len(obs_checksum) > 0)
# Checksum of the log file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[3][0], ))[0]
self.assertEqual(obs_checksum, '626839734')
def test_update_preprocessed_data_from_cmd_ppd(self):
exp_ppd = PreprocessedData(2)
next_fp_id = get_count('qiita.filepath') + 1
exp_fps = []
path_builder = partial(join, self.db_ppd_dir)
suffix_types = [("seqs.fna", "preprocessed_fasta"),
("seqs.fastq", "preprocessed_fastq"),
("seqs.demux", "preprocessed_demux"),
("split_library_log.txt", "log")]
for id_, vals in enumerate(suffix_types, start=next_fp_id):
suffix, fp_type = vals
exp_fps.append(
(id_, path_builder("%s_%s" % (exp_ppd.id, suffix)), fp_type))
ppd = update_preprocessed_data_from_cmd(self.test_slo, 1, 2)
# Check that the modified preprocessed data is the correct one
self.assertEqual(ppd.id, exp_ppd.id)
# Check that the filepaths returned are correct
# We need to sort the list returned from the db because the ordering
# on that list is based on db modification time, rather than id
obs_fps = sorted(ppd.get_filepaths())
self.assertEqual(obs_fps, exp_fps)
# Check that the checksums have been updated
sql = "SELECT checksum FROM qiita.filepath WHERE filepath_id=%s"
# Checksum of the fasta file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[0][0], ))[0]
self.assertEqual(obs_checksum, '3532748626')
# Checksum of the fastq file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[1][0], ))[0]
self.assertEqual(obs_checksum, '2958832064')
# Checksum of the demux file
# The checksum is generated dynamically, so the checksum changes
# We are going to test that the checksum is not the one that was
# before, which corresponds to an empty file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[2][0], ))[0]
self.assertTrue(isinstance(obs_checksum, str))
self.assertNotEqual(obs_checksum, '852952723')
self.assertTrue(len(obs_checksum) > 0)
# Checksum of the log file
obs_checksum = self.conn_handler.execute_fetchone(
sql, (obs_fps[3][0], ))[0]
self.assertEqual(obs_checksum, '626839734')
def test_vamps_status(self):
ppd = PreprocessedData(1)
# verifying current value
self.assertEqual(ppd.submitted_to_vamps_status(), 'not submitted')
# changing value and then verifying new value
ppd.update_vamps_status('failed')
self.assertEqual(ppd.submitted_to_vamps_status(), 'failed')
# checking failure
with self.assertRaises(ValueError):
ppd.update_vamps_status('not a valid status')
def setUp(self):
self.preprocessed_data = PreprocessedData(1)
self.params_table = "processed_params_uclust"
self.params_id = 1
fd, self.biom_fp = mkstemp(suffix='_table.biom')
close(fd)
self.filepaths = [(self.biom_fp, 6)]
self.date = datetime(2014, 5, 29, 12, 24, 51)
self.db_test_pd_dir = join(get_db_files_base_dir(), 'processed_data')
with open(self.biom_fp, "w") as f:
f.write("\n")
self._clean_up_files = []
def processor(preprocessed_data_id, param_id, param_constructor):
"""Dispatch the processor work"""
preprocessed_data = PreprocessedData(preprocessed_data_id)
params = param_constructor(param_id)
sp = StudyProcessor()
try:
process_out = sp(preprocessed_data, params)
except Exception as e:
error_msg = ''.join(format_exception_only(e, exc_info()))
preprocessed_data.processing_status = "failed: %s" % error_msg
process_out = None
return process_out
def _get_template_variables(self, preprocessed_data_id, callback):
"""Generates all the variables needed to render the template
Parameters
----------
preprocessed_data_id : int
The preprocessed data identifier
callback : function
The callback function to call with the results once the processing
is done
Raises
------
HTTPError
If the preprocessed data does not have a log file
"""
# Get the objects and check user privileges
ppd = PreprocessedData(preprocessed_data_id)
study = Study(ppd.study)
check_access(self.current_user, study, raise_error=True)
# Get the return address
back_button_path = self.get_argument(
'back_button_path',
'/study/description/%d?top_tab=preprocessed_data_tab&sub_tab=%s'
% (study.id, preprocessed_data_id))
# Get all the filepaths attached to the preprocessed data
files_tuples = ppd.get_filepaths()
# Group the files by filepath type
files = defaultdict(list)
for _, fp, fpt in files_tuples:
files[fpt].append(fp)
try:
log_path = files['log'][0]
except KeyError:
raise HTTPError(500, "Log file not found in preprocessed data %s"
% preprocessed_data_id)
with open(log_path, 'U') as f:
contents = f.read()
contents = contents.replace('\n', '<br/>')
contents = contents.replace('\t', ' ')
title = 'Preprocessed Data: %d' % preprocessed_data_id
callback((title, contents, back_button_path))
def test_update_insdc_status(self):
"""Able to update insdc status"""
pd = PreprocessedData(1)
self.assertEqual(pd.submitted_to_insdc_status(), 'submitting')
pd.update_insdc_status('failed')
self.assertEqual(pd.submitted_to_insdc_status(), 'failed')
pd.update_insdc_status('success', 'foo', 'bar')
self.assertEqual(pd.submitted_to_insdc_status(), 'success')
self.assertEqual(pd.ebi_study_accession, 'foo')
self.assertEqual(pd.ebi_submission_accession, 'bar')
with self.assertRaises(ValueError):
pd.update_insdc_status('not valid state')
with self.assertRaises(ValueError):
pd.update_insdc_status('success', 'only one accession')
def display_template(self, preprocessed_data_id, msg, msg_level):
"""Simple function to avoid duplication of code"""
preprocessed_data_id = int(preprocessed_data_id)
try:
preprocessed_data = PreprocessedData(preprocessed_data_id)
except QiitaDBUnknownIDError:
raise HTTPError(404, "PreprocessedData %d does not exist!" %
preprocessed_data_id)
else:
user = self.current_user
if user.level != 'admin':
raise HTTPError(403, "No permissions of admin, "
"get/EBISubmitHandler: %s!" % user.id)
prep_template = PrepTemplate(preprocessed_data.prep_template)
sample_template = SampleTemplate(preprocessed_data.study)
study = Study(preprocessed_data.study)
stats = [('Number of samples', len(prep_template)),
('Number of metadata headers',
len(sample_template.metadata_headers()))]
demux = [path for _, path, ftype in preprocessed_data.get_filepaths()
if ftype == 'preprocessed_demux']
demux_length = len(demux)
if not demux_length:
msg = ("Study does not appear to have demultiplexed "
"sequences associated")
msg_level = 'danger'
elif demux_length > 1:
msg = ("Study appears to have multiple demultiplexed files!")
msg_level = 'danger'
elif demux_length == 1:
demux_file = demux[0]
demux_file_stats = demux_stats(demux_file)
stats.append(('Number of sequences', demux_file_stats.n))
msg_level = 'success'
self.render('ebi_submission.html',
study_title=study.title, stats=stats, message=msg,
study_id=study.id, level=msg_level,
preprocessed_data_id=preprocessed_data_id,
investigation_type=prep_template.investigation_type)
def submit_EBI(preprocessed_data_id, action, send, fastq_dir_fp=None):
"""Submit a preprocessed data to EBI
Parameters
----------
preprocessed_data_id : int
The preprocesssed data id
action : %s
The action to perform with this data
send : bool
True to actually send the files
fastq_dir_fp : str, optional
The fastq filepath
"""
preprocessed_data = PreprocessedData(preprocessed_data_id)
preprocessed_data_id_str = str(preprocessed_data_id)
study = Study(preprocessed_data.study)
sample_template = SampleTemplate(study.sample_template)
prep_template = PrepTemplate(preprocessed_data.prep_template)
investigation_type = None
new_investigation_type = None
status = preprocessed_data.submitted_to_insdc_status()
if status in ('submitting', 'success'):
raise ValueError("Cannot resubmit! Current status is: %s" % status)
if send:
# If we intend actually to send the files, then change the status in
# the database
preprocessed_data.update_insdc_status('submitting')
# we need to figure out whether the investigation type is a known one
# or if we have to submit a "new_investigation_type" to EBI
current_type = prep_template.investigation_type
ena_ontology = Ontology(convert_to_id('ENA', 'ontology'))
if current_type in ena_ontology.terms:
investigation_type = current_type
elif current_type in ena_ontology.user_defined_terms:
investigation_type = 'Other'
new_investigation_type = current_type
else:
# This should never happen
raise ValueError("Unrecognized investigation type: '%s'. This term "
"is neither one of the official terms nor one of the "
"user-defined terms in the ENA ontology")
if fastq_dir_fp is not None:
# If the user specifies a FASTQ directory, use it
# Set demux_samples to None so that MetadataTemplate.to_file will put
# all samples in the template files
demux_samples = None
else:
# If the user does not specify a FASTQ directory, create one and
# re-serialize the per-sample FASTQs from the demux file
fastq_dir_fp = mkdtemp(prefix=qiita_config.working_dir)
demux = [
path for _, path, ftype in preprocessed_data.get_filepaths()
if ftype == 'preprocessed_demux'
][0]
# Keep track of which files were actually in the demux file so that we
# can write those rows to the prep and samples templates
demux_samples = set()
with open_file(demux) as demux_fh:
for samp, iterator in to_per_sample_ascii(demux_fh,
list(sample_template)):
demux_samples.add(samp)
sample_fp = join(fastq_dir_fp, "%s.fastq.gz" % samp)
with gzopen(sample_fp, 'w') as fh:
for record in iterator:
fh.write(record)
output_dir = fastq_dir_fp + '_submission'
samp_fp = join(fastq_dir_fp, 'sample_metadata.txt')
prep_fp = join(fastq_dir_fp, 'prep_metadata.txt')
sample_template.to_file(samp_fp, demux_samples)
prep_template.to_file(prep_fp, demux_samples)
# Get specific output directory and set filepaths
get_output_fp = partial(join, output_dir)
study_fp = get_output_fp('study.xml')
sample_fp = get_output_fp('sample.xml')
experiment_fp = get_output_fp('experiment.xml')
run_fp = get_output_fp('run.xml')
submission_fp = get_output_fp('submission.xml')
if not isdir(output_dir):
makedirs(output_dir)
else:
raise IOError('The output folder already exists: %s' % output_dir)
with open(samp_fp, 'U') as st, open(prep_fp, 'U') as pt:
submission = EBISubmission.from_templates_and_per_sample_fastqs(
preprocessed_data_id_str,
study.title,
study.info['study_abstract'],
investigation_type,
st,
pt,
fastq_dir_fp,
new_investigation_type=new_investigation_type,
pmids=study.pmids)
submission.write_all_xml_files(study_fp, sample_fp, experiment_fp, run_fp,
submission_fp, action)
if send:
submission.send_sequences()
study_accession, submission_accession = submission.send_xml()
if study_accession is None or submission_accession is None:
preprocessed_data.update_insdc_status('failed')
raise ComputeError("EBI Submission failed!")
else:
preprocessed_data.update_insdc_status('success', study_accession,
submission_accession)
else:
study_accession, submission_accession = None, None
return study_accession, submission_accession
def test_processed_data(self):
"""Correctly returns the processed data id"""
ppd = PreprocessedData(1)
self.assertEqual(ppd.processed_data, [1])
def test_prep_template(self):
"""Correctly returns the prep template"""
ppd = PreprocessedData(1)
self.assertEqual(ppd.prep_template, 1)
def test_ebi_ebi_study_accession(self):
"""Correctly returns the ebi_study_accession"""
ppd = PreprocessedData(1)
self.assertEqual(ppd.ebi_study_accession, 'EBI123456-BB')
def test_study(self):
"""Correctly returns the study"""
ppd = PreprocessedData(1)
self.assertEqual(ppd.study, 1)
def test_raw_data(self):
"""Correctly returns the raw data"""
ppd = PreprocessedData(1)
self.assertEqual(ppd.raw_data, 1)
def test_link_filepaths_status(self):
ppd = PreprocessedData(1)
self.assertEqual(ppd.link_filepaths_status, 'idle')
def test_link_filepaths_status_setter_error(self):
ppd = PreprocessedData(1)
with self.assertRaises(ValueError):
ppd._set_link_filepaths_status('not a valid status')
