hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-02-16 15:47:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

reinstate word column in icu word table

Browse Source 
Postgresql is very bad at creating statistics for jsonb
columns. The result is that the query planer tends to
use JIT for queries with a where over 'info' even when
there is an index.
This commit is contained in:
Sarah Hoffmann
2021-07-25 15:08:11 +02:00
parent 324b1b5575
commit 1db098c05d
6 changed files with 53 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
test/bdd/steps/steps_db_ops.py
View File

@@ -266,22 +266,6 @@ def check_location_postcode(context):
db_row.assert_row(row, ('country', 'postcode'))
@then("word contains(?P<exclude> not)?")
def check_word_table(context, exclude):
""" Check the contents of the word table. Each row represents a table row
and all data must match. Data not present in the expected table, may
be arbitry. The rows are identified via all given columns.
"""
with context.db.cursor(cursor_factory=psycopg2.extras.DictCursor) as cur:
for row in context.table:
wheres = ' AND '.join(["{} = %s".format(h) for h in row.headings])
cur.execute("SELECT * from word WHERE " + wheres, list(row.cells))
if exclude:
assert cur.rowcount == 0, "Row still in word table: %s" % '/'.join(values)
else:
assert cur.rowcount > 0, "Row not in word table: %s" % '/'.join(values)
@then("there are(?P<exclude> no)? word tokens for postcodes (?P<postcodes>.*)")
def check_word_table_for_postcodes(context, exclude, postcodes):
""" Check that the tokenizer produces postcode tokens for the given
@@ -297,8 +281,7 @@ def check_word_table_for_postcodes(context, exclude, postcodes):
with context.db.cursor(cursor_factory=psycopg2.extras.DictCursor) as cur:
if nctx.tokenizer == 'legacy_icu':
cur.execute("""SELECT info->>'postcode' FROM word
WHERE type = 'P' and info->>'postcode' = any(%s)""",
cur.execute("SELECT word FROM word WHERE type = 'P' and word = any(%s)",
(plist,))
else:
cur.execute("""SELECT word FROM word WHERE word = any(%s)

24
test/python/mock_icu_word_table.py
View File

@@ -12,16 +12,16 @@ class MockIcuWordTable:
cur.execute("""CREATE TABLE word (word_id INTEGER,
word_token text NOT NULL,
type text NOT NULL,
word text,
info jsonb)""")
conn.commit()
def add_special(self, word_token, word, cls, typ, oper):
with self.conn.cursor() as cur:
cur.execute("""INSERT INTO word (word_token, type, info)
VALUES (%s, 'S',
json_build_object('word', %s,
'class', %s,
cur.execute("""INSERT INTO word (word_token, type, word, info)
VALUES (%s, 'S', %s,
json_build_object('class', %s,
'type', %s,
'op', %s))
""", (word_token, word, cls, typ, oper))
@@ -30,16 +30,16 @@ class MockIcuWordTable:
def add_country(self, country_code, word_token):
with self.conn.cursor() as cur:
cur.execute("""INSERT INTO word (word_token, type, info)
VALUES(%s, 'C', json_build_object('cc', %s))""",
cur.execute("""INSERT INTO word (word_token, type, word)
VALUES(%s, 'C', %s)""",
(word_token, country_code))
self.conn.commit()
def add_postcode(self, word_token, postcode):
with self.conn.cursor() as cur:
cur.execute("""INSERT INTO word (word_token, type, info)
VALUES (%s, 'P', json_build_object('postcode', %s))
cur.execute("""INSERT INTO word (word_token, type, word)
VALUES (%s, 'P', %s)
""", (word_token, postcode))
self.conn.commit()
@@ -56,8 +56,8 @@ class MockIcuWordTable:
def get_special(self):
with self.conn.cursor() as cur:
cur.execute("SELECT word_token, info FROM word WHERE type = 'S'")
result = set(((row[0], row[1]['word'], row[1]['class'],
cur.execute("SELECT word_token, info, word FROM word WHERE type = 'S'")
result = set(((row[0], row[2], row[1]['class'],
row[1]['type'], row[1]['op']) for row in cur))
assert len(result) == cur.rowcount, "Word table has duplicates."
return result
@@ -65,7 +65,7 @@ class MockIcuWordTable:
def get_country(self):
with self.conn.cursor() as cur:
cur.execute("SELECT info->>'cc', word_token FROM word WHERE type = 'C'")
cur.execute("SELECT word, word_token FROM word WHERE type = 'C'")
result = set((tuple(row) for row in cur))
assert len(result) == cur.rowcount, "Word table has duplicates."
return result
@@ -73,7 +73,7 @@ class MockIcuWordTable:
def get_postcodes(self):
with self.conn.cursor() as cur:
cur.execute("SELECT info->>'postcode' FROM word WHERE type = 'P'")
cur.execute("SELECT word FROM word WHERE type = 'P'")
return set((row[0] for row in cur))