def search(self, w):
if not self.ix.up_to_date():
self.initialize_trie(
) # if the index is not up to date, someone has added cards, so we reinitialize the trie
# If we searched for an alias, make it the exact hit
for alias, name in fetcher.card_aliases():
if w == card.canonicalize(alias):
return SearchResult(name, None, None, None)
normalized = list(WhooshConstants.normalized_analyzer(w))[0].text
# If we get matches by prefix, we return that
exact, prefix_whole_word, other_prefixed = self.find_matches_by_prefix(
normalized)
if exact or len(prefix_whole_word) > 0 or len(other_prefixed) > 0:
return SearchResult(exact, prefix_whole_word, other_prefixed, None)
# We try fuzzy and stemmed queries
query_normalized = fuzzy_term(normalized, self.DIST, "name_normalized")
query_stemmed = And([
Term('name_stemmed', q.text)
for q in WhooshConstants.stem_analyzer(w)
])
query_tokenized = And([
fuzzy_term(q.text, self.DIST, "name_tokenized")
for q in WhooshConstants.tokenized_analyzer(w)
])
query = Or([query_normalized, query_tokenized, query_stemmed])
with self.ix.searcher() as searcher:
fuzzy = [(r['name'], r.score)
for r in searcher.search(query, limit=40)]
return SearchResult(exact, prefix_whole_word, other_prefixed, fuzzy)
def update_fuzzy_matching():
format_id = get_format_id('Penny Dreadful', True)
if db().is_sqlite():
db().execute('DROP TABLE IF EXISTS fuzzy')
db().execute(
'CREATE VIRTUAL TABLE IF NOT EXISTS fuzzy USING spellfix1')
sql = """INSERT INTO fuzzy (word, rank)
SELECT LOWER(bq.name), bq.pd_legal
FROM ({base_query}) AS bq
""".format(base_query=base_query())
db().execute(sql)
sql = """INSERT INTO fuzzy (word, rank)
SELECT LOWER(f.name), SUM(CASE WHEN cl.format_id = {format_id} THEN 1 ELSE 0 END) > 0
FROM face AS f
INNER JOIN card AS c ON f.card_id = c.id
LEFT OUTER JOIN card_legality AS cl ON cl.card_id = c.id AND cl.format_id = {format_id}
WHERE LOWER(f.name) NOT IN (SELECT word FROM fuzzy)
GROUP BY f.id
""".format(format_id=format_id)
db().execute(sql)
aliases = fetcher.card_aliases()
for alias, name in aliases:
db().execute(
'INSERT INTO fuzzy (word, soundslike) VALUES (LOWER(?), ?)',
[name, alias])
def reindex() -> None:
writer = WhooshWriter()
cs = get_all_cards()
for alias, name in fetcher.card_aliases():
for c in cs:
if c.name == name:
c.names.append(alias)
writer.rewrite_index(cs)
def cards_from_query(query):
# Skip searching if the request is too short.
if len(query) <= 2:
return []
mode = 0
if query.startswith('$'):
mode = '$'
query = query[1:]
query = card.canonicalize(query)
# If we searched for an alias, change query so we can find the card in the results.
for alias, name in fetcher.card_aliases():
if query == card.canonicalize(alias):
query = card.canonicalize(name)
cards = search(query)
cards = [c for c in cards if c.layout != 'token' and c.type != 'Vanguard']
# First look for an exact match.
results = []
for c in cards:
c.mode = mode
if query == card.canonicalize(c.name):
results.append(c)
if len(results) > 0:
return results
for c in cards:
names = [card.canonicalize(name) for name in c.names]
if query in names:
results.append(c)
if len(results) > 0:
return results
# If not found, use cards that start with the query and a punctuation char.
for c in cards:
names = [card.canonicalize(name) for name in c.names]
for name in names:
if name.startswith(
'{query} '.format(query=query)) or name.startswith(
'{query},'.format(query=query)):
results.append(c)
if len(results) > 0:
return results
# If not found, use cards that start with the query.
for c in cards:
names = [card.canonicalize(name) for name in c.names]
for name in names:
if name.startswith(query):
results.append(c)
if len(results) > 0:
return results
# If we didn't find any of those then use all search results.
return cards