How to count rows with SELECT COUNT(*) with SQLAlchemy?

I'd like to know if it's possible to generate a SELECT COUNT(*) FROM TABLE statement in SQLAlchemy without explicitly asking for it with execute(). If I use:

session.query(table).count()

then it generates something like:

SELECT count(*) AS count_1 FROM
    (SELECT table.col1 as col1, table.col2 as col2, ... from table)

which is significantly slower in MySQL with InnoDB. I am looking for a solution that doesn't require the table to have a known primary key, as suggested in Get the number of rows in table using SQLAlchemy.

Solution 1:

I managed to render the following SELECT with SQLAlchemy on both layers.

SELECT count(*) AS count_1
FROM "table"

Usage from the SQL Expression layer

from sqlalchemy import select, func, Integer, Table, Column, MetaData

metadata = MetaData()

table = Table("table", metadata,
              Column('primary_key', Integer),
              Column('other_column', Integer)  # just to illustrate
             )   

print select([func.count()]).select_from(table)

Usage from the ORM layer

You just subclass Query (you have probably anyway) and provide a specialized count() method, like this one.

from sqlalchemy.sql.expression import func

class BaseQuery(Query):
    def count_star(self):
        count_query = (self.statement.with_only_columns([func.count()])
                       .order_by(None))
        return self.session.execute(count_query).scalar()

Please note that order_by(None) resets the ordering of the query, which is irrelevant to the counting.

Using this method you can have a count(*) on any ORM Query, that will honor all the filter andjoin conditions already specified.

Solution 2:

Query for just a single known column:

session.query(MyTable.col1).count()

Solution 3:

I needed to do a count of a very complex query with many joins. I was using the joins as filters, so I only wanted to know the count of the actual objects. count() was insufficient, but I found the answer in the docs here:

http://docs.sqlalchemy.org/en/latest/orm/tutorial.html

The code would look something like this (to count user objects):

from sqlalchemy import func

session.query(func.count(User.id)).scalar()

Solution 4:

Addition to the Usage from the ORM layer in the accepted answer: count(*) can be done for ORM using the query.with_entities(func.count()), like this:

session.query(MyModel).with_entities(func.count()).scalar()

It can also be used in more complex cases, when we have joins and filters - the important thing here is to place with_entities after joins, otherwise SQLAlchemy could raise the Don't know how to join error.

For example:

we have User model (id, name) and Song model (id, title, genre)
we have user-song data - the UserSong model (user_id, song_id, is_liked) where user_id + song_id is a primary key)

We want to get a number of user's liked rock songs:

SELECT count(*) 
  FROM user_song
  JOIN song ON user_song.song_id = song.id 
 WHERE user_song.user_id = %(user_id)
   AND user_song.is_liked IS 1
   AND song.genre = 'rock'

This query can be generated in a following way:

user_id = 1

query = session.query(UserSong)
query = query.join(Song, Song.id == UserSong.song_id)
query = query.filter(
    and_(
        UserSong.user_id == user_id, 
        UserSong.is_liked.is_(True),
        Song.genre == 'rock'
    )
)
# Note: important to place `with_entities` after the join
query = query.with_entities(func.count())
liked_count = query.scalar()

Complete example is here.

Solution 5:

If you are using the SQL Expression Style approach there is another way to construct the count statement if you already have your table object.

Preparations to get the table object. There are also different ways.

import sqlalchemy

database_engine = sqlalchemy.create_engine("connection string")

# Populate existing database via reflection into sqlalchemy objects
database_metadata = sqlalchemy.MetaData()
database_metadata.reflect(bind=database_engine)

table_object = database_metadata.tables.get("table_name") # This is just for illustration how to get the table_object

Issuing the count query on the table_object

query = table_object.count()
# This will produce something like, where id is a primary key column in "table_name" automatically selected by sqlalchemy
# 'SELECT count(table_name.id) AS tbl_row_count FROM table_name'

count_result = database_engine.scalar(query)