table=whoosh.fields.ID(stored=True),
row_id=whoosh.fields.ID(stored=True),
language=whoosh.fields.STORED,
+ iso3166=whoosh.fields.STORED,
display_name=whoosh.fields.STORED, # non-lowercased name
forme_name=whoosh.fields.ID,
)
row_id=unicode(row.id),
forme_name=u'XXX')
- def add(name, language, score):
+ def add(name, language, iso3166, score):
writer.add_document(name=name.lower(), display_name=name,
language=language,
+ iso3166=iso3166,
**row_key)
speller_entries.append((name.lower(), score))
row_key['forme_name'] = row.forme_name
name = row.name
- add(name, None, 1)
+ add(name, None, u'us', 1)
# Pokemon also get other languages
for foreign_name in getattr(row, 'foreign_names', []):
# no point and it makes spell results confusing
continue
- add(moonspeak, foreign_name.language.name, 3)
+ add(moonspeak, foreign_name.language.name,
+ foreign_name.language.iso3166,
+ 3)
# Add Roomaji too
if foreign_name.language.name == 'Japanese':
roomaji = romanize(foreign_name.name)
- add(roomaji, u'Roomaji', 8)
+ add(roomaji, u'Roomaji', u'jp', 8)
writer.commit()
rx_is_number = re.compile('^\d+$')
LookupResult = namedtuple('LookupResult',
- ['object', 'name', 'language', 'exact'])
+ ['object', 'name', 'language', 'iso3166', 'exact'])
def _parse_table_name(name):
"""Takes a singular table name, table name, or table object and returns the
results.append(LookupResult(object=obj,
name=record['display_name'],
language=record['language'],
+ iso3166=record['iso3166'],
exact=exact))
return results
def lookup(input, valid_types=[], session=None, indices=None, exact_only=False):
"""Attempts to find some sort of object, given a database session and name.
- Returns a list of named (object, name, language, exact) tuples. `object`
- is a database object, `name` is the name under which the object was found,
- `language` is the name of the language in which the name was found, and
- `exact` is True iff this was an exact match.
+ Returns a list of named (object, name, language, iso3166, exact) tuples.
+ `object` is a database object, `name` is the name under which the object
+ was found, `language` and `iso3166` are the name and country code of the
+ language in which the name was found, and `exact` is True iff this was an
+ exact match.
This function currently ONLY does fuzzy matching if there are no exact
matches.