def from_txt(cls, txt, session=None, **kw):
session = session or DBSession
lines = nfilter(txt.split('\n'))
m = LANGUAGE_LINE_PATTERN.match(lines[0])
assert m
kw['id'] = m.group('name')
kw['name'] = ' '.join(s.capitalize() for s in kw['id'].split('_'))
for cname in ['wals', 'ethnologue', 'glottolog']:
if m.group(cname[0]):
kw['classification_' + cname] = m.group(cname[0])
kw.update(parse_metadata(lines[1]))
doculect = cls(**kw)
if doculect.classification_ethnologue:
doculect.ethnologue_family = doculect.classification_ethnologue.split(
',')[0]
if doculect.classification_glottolog:
doculect.glottolog_family = doculect.classification_glottolog.split(
',')[0]
doculect.wordlist = Contribution(id=kw['id'],
language=doculect,
name=doculect.id)
parameters = {p.id: p for p in session.query(Parameter)}
for line in lines[2:]:
if '\t' in line:
wid, words, comment = parse_word(line)
# if int(wid) not in MEANINGS_ALL:
# # drop non-core meanings
# continue
vsid = '%s-%s' % (doculect.id, wid)
vs = Synset(id=vsid,
description=comment,
language=doculect,
contribution=doculect.wordlist,
parameter=parameters[wid])
for i, word in enumerate(words):
id_ = '%s-%s' % (vsid, i + 1)
word, loan = word
word = Word(id=id_, name=word, valueset=vs, loan=loan)
return doculect
def import_values(values, lang, features, codes, contributors,
sources): # pragma: no cover
c = Contribution(
id=lang['ID'],
name='Dataset for {0}'.format(lang['Name']),
)
for i, cid in enumerate(lang['Coders'], start=1):
DBSession.add(
ContributionContributor(
contribution=c,
contributor_pk=contributors[cid],
ord=i,
))
l = GrambankLanguage(
id=lang['ID'],
name=lang['Name'],
macroarea=lang['Macroarea'],
latitude=lang['Latitude'],
longitude=lang['Longitude'],
)
for value in values:
vs = ValueSet(
id=value['ID'],
parameter_pk=features[value['Parameter_ID']],
language=l,
contribution=c,
)
Value(id=value['ID'],
valueset=vs,
name=value['Value'],
description=value['Comment'],
domainelement_pk=codes[value['Code_ID']
or '{}-NA'.format(value['Parameter_ID'])])
if value['Source']:
for ref in value['Source']:
sid, pages = Sources.parse(ref)
ValueSetReference(valueset=vs,
source_pk=sources[sid],
description=pages)
DBSession.add(c)
