def test_db(tmpdir, dataset, mocker, capsys):
db = Database(str(tmpdir.join('lexibank.sqlite')))
db.load(dataset)
db.create(exists_ok=True)
with pytest.raises(ValueError):
db.create()
db.create(force=True)
db.load(dataset)
db.load_glottolog_data(dataset.glottolog)
db.load_concepticon_data(mocker.Mock(conceptsets={}))
for sql in db.sql:
db.fetchall(sql)
with db.connection() as conn:
db.fetchall('select * from dataset', conn=conn, verbose=True)
out, _ = capsys.readouterr()
assert 'select' in out
db.create(force=True)
db.load(dataset)
cldf_ds = dataset.cldf_reader()
cols = cldf_ds['FormTable'].tableSchema.columns
cols.append(Column(name='custom'))
cldf_ds.write_metadata()
db.load(dataset)
cols.pop()
cols.append(Column(name='custom', datatype='integer'))
cldf_ds.write_metadata()
with pytest.raises(ValueError):
db.load(dataset)
cols.pop()
cldf_ds.write_metadata()
db.load(dataset)
def test_db_multiple_datasets(tmpdir, dataset, dataset_cldf, capsys):
db = Database(str(tmpdir.join('lexibank.sqlite')))
db.load(dataset)
db.load(dataset_cldf, verbose=True)
with db.connection() as conn:
res = db.fetchall('select `id`, `name` from LanguageTable', conn=conn)
assert len(res) == 3
assert ('1', 'Lang CLDF') in [(r[0], r[1]) for r in res]
res = db.fetchall('select `id`, `value` from FormTable', conn=conn)
assert ('1', 'abc') in [(r[0], r[1]) for r in res]
def requirements(args):
if args.cfg.datasets:
print(
'-e git+https://github.com/clld/glottolog.git@{0}#egg=pyglottolog'.
format(git_hash(args.cfg.datasets[0].glottolog.repos)))
print(
'-e git+https://github.com/clld/concepticon-data.git@{0}#egg=pyconcepticon'
.format(git_hash(args.cfg.datasets[0].concepticon.repos)))
if pylexibank.__version__.endswith('dev0'):
print(
'-e git+https://github.com/lexibank/pylexibank.git@{0}#egg=pylexibank'
.format(git_hash(Path(pylexibank.__file__).parent.parent.parent)))
db = Database(args.db)
db.create(exists_ok=True)
for r in db.fetchall('select id, version from dataset'):
print(
'-e git+https://github.com/lexibank/{0}.git@{1}#egg=lexibank_{0}'.
format(*r))
def test_db_multiple_datasets_error(tmpdir, dataset, dataset_factory):
import shutil
from clldutils.jsonlib import load, dump
db = Database(str(tmpdir.join('lexibank.sqlite')))
assert not db.fname.exists()
db.load(dataset)
ds_dir = dataset.dir.parent / 'dbtest'
shutil.copytree(str(dataset.dir), str(ds_dir))
# Now modify the CLDF data:
md = load(ds_dir / 'cldf' / 'cldf-metadata.json')
for t in md['tables']:
if t['url'] == 'parameters.csv':
for col in t['tableSchema']['columns']:
if col['name'] == 'Chinese':
col['name'] = 'chinese'
col['datatype'] = 'integer'
dump(md, ds_dir / 'cldf' / 'cldf-metadata.json')
ds2 = dataset_factory('dbtest.td')
with pytest.raises(ValueError) as e:
db.load(ds2)
assert 'redefined' in str(e)
def _unload(ds, **kw):
db = Database(kw['db'])
db.create(exists_ok=True)
db.unload(ds)
def _load(ds, **kw):
db = Database(kw['db'])
db.create(exists_ok=True)
db.load(ds)
db.load_concepticon_data(ds.concepticon)
db.load_glottolog_data(ds.glottolog)
def ls(args):
"""
lexibank ls [COLS]+
column specification:
- license
- lexemes
- macroareas
"""
db = Database(args.db)
db.create(exists_ok=True)
in_db = {
r[0]: r[1]
for r in db.fetchall('select id, version from dataset')
}
# FIXME: how to smartly choose columns?
table = Table('ID', 'Title')
cols = OrderedDict([(col, {}) for col in args.args if col in [
'version',
'location',
'changes',
'license',
'all_lexemes',
'lexemes',
'concepts',
'languages',
'families',
'varieties',
'macroareas',
]])
tl = 40
if cols:
tl = 25
table.columns.extend(col.capitalize() for col in cols)
for col, sql in [
('languages', 'glottocodes_by_dataset'),
('concepts', 'conceptsets_by_dataset'),
('lexemes', 'mapped_lexemes_by_dataset'),
('all_lexemes', 'lexemes_by_dataset'),
('macroareas', 'macroareas_by_dataset'),
('families', 'families_by_dataset'),
]:
if col in cols:
cols[col] = {r[0]: r[1] for r in db.fetchall(sql)}
for ds in args.cfg.datasets:
row = [
colored(ds.id, 'green' if ds.id in in_db else 'red'),
truncate_with_ellipsis(ds.metadata.title or '', width=tl),
]
for col in cols:
if col == 'version':
row.append(git_hash(ds.dir))
elif col == 'location':
row.append(colored(str(ds.dir), 'green'))
elif col == 'changes':
row.append(ds.git_repo.is_dirty())
elif col == 'license':
lic = licenses.find(ds.metadata.license or '')
row.append(lic.id if lic else ds.metadata.license)
elif col in [
'languages', 'concepts', 'lexemes', 'all_lexemes',
'families'
]:
row.append(float(cols[col].get(ds.id, 0)))
elif col == 'macroareas':
row.append(', '.join(
sorted((cols[col].get(ds.id) or '').split(','))))
else:
row.append('')
table.append(row)
totals = ['zztotal', len(args.cfg.datasets)]
for i, col in enumerate(cols):
if col in ['lexemes', 'all_lexemes']:
totals.append(sum([r[i + 2] for r in table]))
elif col == 'languages':
totals.append(
float(
db.fetchone(
"SELECT count(distinct glottocode) FROM languagetable")
[0]))
elif col == 'concepts':
totals.append(
float(
db.fetchone(
"SELECT count(distinct concepticon_id) FROM parametertable"
)[0]))
elif col == 'families':
totals.append(
float(
db.fetchone(
"SELECT count(distinct family) FROM languagetable")
[0]))
else:
totals.append('')
table.append(totals)
print(
table.render(tablefmt='simple',
sortkey=lambda r: r[0],
condensed=False,
floatfmt=',.0f'))
def db(args):
db = str(Database(args.db).fname)
args.log.info('connecting to {0}'.format(colored(db, 'green')))
check_call(['sqlite3', db])
def get_db(args):
db = Database(args.db)
db.create(exists_ok=True)
return db
def run(args):
db = str(Database(args.db).fname)
args.log.info('connecting to {0}'.format(termcolor.colored(db, 'green')))
subprocess.check_call(['sqlite3', db])
def db(tmp_path):
return Database(tmp_path / 'lexibank.sqlite')