def add_pronouns(ctx):
"""Populates `PronounStem` and `Pronoun`."""
session = ctx.session
gender_group = ENUM['gender_group']
gender = ENUM['gender']
case = ENUM['case']
number = ENUM['number']
seen_stems = {} # (stem, genders_id) -> id
for row in util.read_csv(ctx.config['PRONOUNS']):
stem = row['stem']
genders_id = gender_group[row['stem_genders']]
if (stem, genders_id) not in seen_stems:
pronoun_stem = PronounStem(name=stem, genders_id=genders_id)
session.add(pronoun_stem)
session.flush()
seen_stems[(stem, genders_id)] = pronoun_stem.id
util.tick(stem)
stem_id = seen_stems[(stem, genders_id)]
session.add(
Nominal(stem_id=stem_id,
name=row['form'],
gender_id=gender[row['form_gender']],
case_id=case[row['case']],
number_id=number[row['number']]))
session.flush()
session.commit()
session.close()
def add_nominal_endings(ctx):
"""Populates `NominalEnding`."""
session = ctx.session
gender = ENUM['gender']
case = ENUM['case']
number = ENUM['number']
for row in util.read_csv(ctx.config['COMPOUNDED_NOMINAL_ENDINGS']):
ending = NominalEnding(name=row['ending'],
stem_type=row['stem_type'],
gender_id=gender[row['form_gender']],
case_id=None,
number_id=None,
compounded=True)
session.add(ending)
seen = set()
for row in util.read_csv(ctx.config['INFLECTED_NOMINAL_ENDINGS']):
ending = NominalEnding(name=row['ending'],
stem_type=row['stem_type'],
gender_id=gender[row['form_gender']],
case_id=case[row['case']],
number_id=number[row['number']],
compounded=False)
session.add(ending)
if row['stem_type'] not in seen:
util.tick(row['stem_type'])
seen.add(row['stem_type'])
session.commit()
session.close()
def add_verbs(ctx, root_map):
"""Add inflected verbs to the database."""
session = ctx.session
vclass = ENUM['vclass']
person = ENUM['person']
number = ENUM['number']
mode = ENUM['mode']
voice = ENUM['voice']
skipped = set()
i = 0
for row in util.read_csv(ctx.config['VERBS']):
root = row['root']
hom = row['hom']
try:
root_id = root_map[(root, hom)]
except KeyError:
skipped.add((root, hom))
continue
data = {
'name': row['form'],
'root_id': root_id,
'vclass_id': vclass[row['class']] if row['class'] else None,
'person_id': person[row['person']],
'number_id': number[row['number']],
'mode_id': mode[row['mode']],
'voice_id': voice[row['voice']]
}
session.add(Verb(**data))
i += 1
if i % 1000 == 0:
util.tick(row['form'])
session.commit()
session.commit()
session.close()
print('Skipped', len(skipped), 'roots.')
def add_irregular_adjectives(ctx):
"""Add regular irregular adjectives to the database."""
session = ctx.session
gender = ENUM['gender']
case = ENUM['case']
number = ENUM['number']
with open(ctx.config['IRREGULAR_ADJECTIVES']) as f:
for adj in yaml.load_all(f):
stem = AdjectiveStem(name=adj['name'])
session.add(stem)
session.flush()
# Mark the stem as irregular
complete = adj['complete']
irreg = StemIrregularity(stem=stem, fully_described=complete)
session.add(irreg)
session.flush()
util.tick(stem.name)
for form in adj['forms']:
name = form['name']
gender_id = gender[form['gender']]
case_id = case[form['case']]
number_id = number[form['number']]
result = Adjective(stem=stem,
name=name,
gender_id=gender_id,
case_id=case_id,
number_id=number_id)
session.add(result)
session.commit()
session.close()
def add_participle_stems(ctx):
"""Populates `ParticipleStem`."""
session = ctx.session
mode = ENUM['mode']
voice = ENUM['voice']
i = 0
for row in util.read_csv(ctx.config['PARTICIPLE_STEMS']):
data = {
'name': row['stem'].split("#")[0],
'mode_id': mode[row['mode']],
'voice_id': voice[row['voice']]
}
session.add(ParticipleStem(**data))
i += 1
if i % 100 == 0:
util.tick(row['stem'])
session.commit()
session.commit()
session.close()
def add_enums(ctx):
"""Add enumerated data to the database. Among others, this includes:
- persons
- numbers
- modes
- voices
- genders
- cases
and any other data with small, known limits.
"""
session = ctx.session
type_to_class = {
'case': Case,
'class': VClass,
'gender': Gender,
'gender_group': GenderGroup,
'modification': Modification,
'mode': Mode,
'number': Number,
'person': Person,
'sandhi_rule_type': SandhiType,
'voice': Voice,
}
# First pass: ordinary enums
for row in util.read_csv(ctx.config['ENUMS']):
if row['enum_type'] == 'gender_group':
continue
cls = type_to_class.get(row['enum_type'], None)
# TODO: always non-None?
if cls is None:
continue
enum_abbr = cls.__tablename__
if enum_abbr not in ENUM:
util.tick(cls.__name__)
ENUM.setdefault(enum_abbr, {})
abbreviation = row['abbreviation']
e = cls(name=row['human_readable_value'], abbr=abbreviation)
session.add(e)
session.flush()
ENUM[enum_abbr][abbreviation] = e.id
session.commit()
# Second pass: gender groups
for row in util.read_csv(ctx.config['ENUMS']):
if row['enum_type'] != 'gender_group':
continue
cls = type_to_class.get(row['enum_type'], None)
enum_abbr = cls.__tablename__
if enum_abbr not in ENUM:
util.tick(cls.__name__)
ENUM.setdefault(enum_abbr, {})
abbreviation = row['abbreviation']
e = cls(name=row['human_readable_value'], abbr=abbreviation)
session.add(e)
session.flush()
if set(abbreviation).issubset('mfn'):
e.members = [ENUM['gender'][x] for x in abbreviation]
ENUM[enum_abbr][abbreviation] = e.id
session.commit()
session.close()