def test_new_nodes01(self):
source_name = 'some_source'
# add some nodes and names
args = ['add_nodes', self.dbname, data_path('new_nodes_ok_nosource.yml'),
'--source-name', source_name]
main(args)
args = ['extract_nodes', self.dbname, source_name,
'-o', self.outfile]
main(args)
def test05(self):
args = ['update_taxids', self.infile, self.db,
'-o', self.outfile,
'--unknown-action', 'drop']
main(args)
expected = [
('tax_id', 'tax_name', 'comment'),
('1280', '', 'ok'),
('1287', 'Staphylococcus staphylolyticus', 'merged with 1287'),
]
self.assertEqual(expected, self.get_rows(self.outfile))
def test06(self):
args = [
'update_taxids', self.infile, self.db, '-o', self.outfile,
'--unknown-action', 'ignore', '--unknowns', self.unknowns
]
main(args)
unknowns = [
('tax_id', 'tax_name', 'comment'),
('', 'who knows?', 'blank'),
('foo', 'unknown', 'completely unknown'),
]
self.assertEqual(unknowns, self.get_rows(self.unknowns))
def test06(self):
args = ['update_taxids', self.infile, self.db,
'-o', self.outfile,
'--unknown-action', 'ignore',
'--unknowns', self.unknowns]
main(args)
unknowns = [
('tax_id', 'tax_name', 'comment'),
('', 'who knows?', 'blank'),
('foo', 'unknown', 'completely unknown'),
]
self.assertEqual(unknowns, self.get_rows(self.unknowns))
def test04(self):
args = ['update_taxids', self.infile, self.db,
'-o', self.outfile,
'--unknown-action', 'ignore']
main(args)
expected = [
('tax_id', 'tax_name', 'comment'),
('1280', '', 'ok'),
('1287', 'Staphylococcus staphylolyticus', 'merged with 1287'),
('', 'who knows?', 'blank'),
('foo', 'unknown', 'completely unknown'),
]
self.assertEqual(expected, self.get_rows(self.outfile))
def test_lineage_table(self):
outfile = os.path.join(self.outdir, 'taxonomy.txt')
args = ['lineage_table', self.taxtable, self.seq_info,
'--taxonomy-table', outfile]
main(args)
with open(outfile) as f:
output = list(csv.reader(f, delimiter='\t'))
self.assertEqual(
[row[0] for row in output],
[row[0] for row in self.info[1:]]
)
for expected, actual in zip(self.info[1:], output):
self.assertTrue(actual[1].endswith(actual[-1]))
def test_new_nodes02(self):
# fails without --source-name
args = [
'add_nodes', self.dbname,
data_path('new_nodes_ok_nosource.yml')
]
self.assertNonZeroExitStatus(main(args))
def test_new_nodes03(self):
args = [
'add_nodes', self.dbname,
data_path('new_nodes_ok_nosource.yml'), '--source-name',
'some_source'
]
self.assertZeroExitStatus(main(args))
def setUp(self):
self.outdir = self.mkoutdir()
self.taxtable = os.path.join(self.outdir, 'taxtable.csv')
self.seq_info = os.path.join(self.outdir, 'seq_info.csv')
self.info = [
('seqname', 'tax_id', 'species', 'mothur'),
('s1', '1280', 'Staphylococcus aureus', 's__Staphylococcus_aureus;'),
('s2', '246432', 'Staphylococcus equorum', 's__Staphylococcus_equorum;'),
('s3', '29383', 'Staphylococcus equorum', 's__Staphylococcus_equorum;'),
('s4', '1279', '', 'g__Staphylococcus;')
]
with open(self.seq_info, 'w') as f:
csv.writer(f).writerows(self.info)
main(['taxtable', config.ncbi_master_db, '-i',
self.seq_info, '-o', self.taxtable])
def test_csv_table(self):
outfile = os.path.join(self.outdir, 'lineages.csv')
args = ['lineage_table', self.taxtable, self.seq_info,
'--csv-table', outfile]
main(args)
with open(outfile) as f:
output = list(csv.DictReader(f))
self.assertEqual(
[row['seqname'] for row in output],
[row[0] for row in self.info[1:]]
)
self.assertEqual(
[row['species'] for row in output],
[row[-2] for row in self.info[1:]]
)
def wrap_cmd(self, cmd):
cmd = cmd % self
try:
retval = main(cmd.split())
except SystemExit as err:
return None, err
else:
return retval, None
def test_seqinfo(self):
args = [
'add_to_taxtable',
data_path('minimal_taxonomy.csv'),
data_path('minimal_add_taxonomy.csv'), '-o',
os.path.join(self.mkoutdir(), 'taxonomy.csv')
]
self.assertIsNone(main(args))
def wrap_cmd(self, cmd):
cmd = cmd % self
try:
retval = main(cmd.split())
except SystemExit as err:
return None, err
else:
return retval, None
def test_seqinfo(self):
outdir = self.mkoutdir()
args = [
'taxtable', config.ncbi_master_db, '--seq-info',
data_path('simple_seqinfo.csv'), '-o',
os.path.join(outdir, 'taxonomy.csv')
]
self.assertIsNone(main(args))
def test_new_nodes05(self):
args = ['add_nodes', self.dbname, data_path('staph_species_group2.yml')]
self.assertZeroExitStatus(main(args))
tax = Taxonomy(sqlalchemy.create_engine('sqlite:///' + self.dbname))
with tax.engine.connect() as con:
result = con.execute(
'select * from nodes where parent_id = ?', ('stapha_sg',))
keys = list(result.keys())
nodes = [dict(list(zip(keys, row))) for row in result.fetchall()]
self.assertEqual(len(nodes), 5)
self.assertEqual([row['source_id'] for row in nodes], [2] * len(nodes))
def test_seqinfo(self):
outdir = self.mkoutdir()
args = ['taxtable', config.ncbi_master_db,
'--seq-info', data_path('simple_seqinfo.csv'),
'-o', os.path.join(outdir, 'taxonomy.csv')]
self.assertIsNone(main(args))
def test_new_nodes01(self):
args = ['add_nodes', self.dbname, data_path('new_nodes_ok.yml')]
self.assertZeroExitStatus(main(args))
def test_new_nodes02(self):
# fails without --source-name
args = ['add_nodes', self.dbname, data_path('new_nodes_ok_nosource.yml')]
self.assertNonZeroExitStatus(main(args))
def test_new_nodes01(self):
args = ['add_nodes', self.dbname, data_path('new_nodes_ok.yml')]
self.assertZeroExitStatus(main(args))
def test_new_nodes03(self):
args = ['add_nodes', self.dbname, data_path('new_nodes_ok_nosource.yml'),
'--source-name', 'some_source']
self.assertZeroExitStatus(main(args))
def main(self, arguments):
taxit.main(['update_taxids'] + [str(a) for a in arguments])
#!/usr/bin/env python
# This file is part of taxtastic.
#
# taxtastic is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# taxtastic is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with taxtastic. If not, see <http://www.gnu.org/licenses/>.
import sys
from taxtastic.scripts.taxit import main
if __name__ == '__main__':
sys.exit(main(sys.argv[1:]))
def test_seqinfo(self):
args = ['add_to_taxtable',
data_path('minimal_taxonomy.csv'),
data_path('minimal_add_taxonomy.csv'),
'-o', os.path.join(self.mkoutdir(), 'taxonomy.csv')]
self.assertIsNone(main(args))
