Korma 0.5.0-RC1 API documentation

Add a comment clause to a select query

Use a SQL aggregator function, aliasing the results, and optionally grouping by
a field:

(select users
  (aggregate (count :*) :cnt :status))

Aggregates available: count, sum, avg, min, max, first, last

Force a query to return a string of SQL when (exec) is called.

Add a belongs-to relationship for the given entity. It is assumed that the foreign key
is on the current entity with the format sub-ent-table_id: email.user_id = user.id.
Can optionally pass a map with a :fk key collection or use the helper `fk` function
to explicitly set the foreign key.

(belongs-to users email)
(belongs-to users email (fk :userId))

Create an entity representing a table in a database.

Set the database connection to be used for this entity.

Define an entity representing a table in the database, applying any modifications in
the body.

Creates a delete query, applies any modifying functions in the body and then
executes it. `ent` is either a string or an entity created by defentity.
Returns number of deleted rows as provided by the JDBC driver.

ex: (delete user
      (where {:id 7}))

Create an empty delete query. Ent can either be an entity defined by defentity,
or a string of the table name

Wrap around a set of queries to print to the console all SQL that would
be run and return dummy values instead of executing them.

Set the fields to be retrieved in all select queries for the
entity.

Execute a query map and return the results.

Execute a raw SQL string, supplying whether results should be returned. `sql`
can either be a string or a vector of the sql string and its params. You can
also optionally provide the connection to execute against as the first
parameter.

(exec-raw ["SELECT * FROM users WHERE age > ?" [5]] :results)

Set the fields to be selected in a query. Fields can either be a keyword
or a vector of two keywords [field alias]:

(fields query :name [:firstname :first])

Set the foreign key used for an entity relationship.

Add tables to the from clause.

Add a group-by clause to a select query

Add a has-many relation for the given entity. It is assumed that the foreign key
is on the sub-entity with the format table_id: user.id = email.user_id
Can optionally pass a map with a :fk key collection or use the helper `fk` function
to explicitly set the foreign key.

(has-many users email)
(has-many users email (fk :emailID))

Add a has-one relationship for the given entity. It is assumed that the foreign key
is on the sub-entity with the format table_id: user.id = address.user_id
Can optionally pass a map with a :fk key collection or use the helper `fk` function
to explicitly set the foreign key.

(has-one users address)
(has-one users address (fk :userId))

Add a having clause to the query, expressing the clause in clojure expressions
with keywords used to reference fields.
e.g. (having query (or (= :hits 1) (> :hits 5)))

Available predicates: and, or, =, not=, <, >, <=, >=, in, like, not, between

Having can also take a map at any point and will create a clause that compares
keys to values. The value can be a vector with one of the above predicate
functions describing how the key is related to the value:
  (having query {:name [like "chris"})

Having only works if you have an aggregation, using it without one will cause
an error.

Add a having clause to the query. Clause can be either a map or a string, and
will be AND'ed to the other clauses.

Creates an insert query, applies any modifying functions in the body
and then executes it. `ent` is either a string or an entity created by
defentity. The return value is the last inserted item, but its
representation is dependent on the database driver used
(e.g. postgresql returns the full row as a hash,
MySQL returns a {:generated_key <ID>} hash,
and MSSQL returns a {:generated_keys <ID>} hash).

ex: (insert user
      (values [{:name "chris"} {:name "john"}]))

Create an empty insert query. Ent can either be an entity defined by defentity,
or a string of the table name

Creates an intersect query, applies any modifying functions in the body and then
executes it.

ex: (intersect
      (queries (subselect user
                 (where {:id 7}))
               (subselect user-backup
                 (where {:id 8})))
      (order :name))

Create an empty intersect query.

Add a join clause to a select query, specifying an entity defined by defentity, or the table name to
join and the predicate to join on. If the entity relationship uses a join
table then two clauses will be added. Otherwise, only one clause
will be added.

(join query addresses)
(join query :right addresses)
(join query addresses (= :addres.users_id :users.id))
(join query :right addresses (= :address.users_id :users.id))

Set the left foreign key used for an entity relationship.

Add a limit clause to a select query.

Add a many-to-many relation for the given entity.  It is assumed that a join
table is used to implement the relationship and that the foreign keys are in
the join table with the format table_id:
  user.id = user_email.user_id
  user_email.email_id = email.id
Can optionally pass a map with :lfk and :rfk key collections or use the
helper `lfk`, `rfk` functions to explicitly set the join keys.

(many-to-many email :user_email)
(many-to-many email :user_email (lfk :user_id)
                                (rfk :email_id))

Add a modifer to the beginning of a query:

(select orders
  (modifier "DISTINCT"))

Add an offset clause to a select query.

Add an ORDER BY clause to a select, union, union-all, or intersect query.
field should be a keyword of the field name, dir is ASC by default.

(order query :created :asc)

Set the primary key used for an entity. :id by default.

Add a function representing a query that should be executed for each result
in a select. This is done lazily over the result set.

Add a function to be applied to records/values going into the database

Adds a group of queries to a union, union-all or intersect

Wrap around a set of queries to force them to return their query objects.

Embed a raw string of SQL in a query. This is used when Korma doesn't
provide some specific functionality you're looking for:

(select users
  (fields (raw "PERIOD(NOW(), NOW())")))

Set the right foreign key used for an entity relationship.

Creates a select query, applies any modifying functions in the body and then
executes it. `ent` is either a string or an entity created by defentity.

ex: (select user
      (fields :name :email)
      (where {:id 2}))

Create a select query with fields provided in Ent.  If fields are not provided,
create an empty select query. Ent can either be an entity defined by defentity,
or a string of the table name

Set the fields and values for an update query.

Wrap around a set of queries so that instead of executing, each will return a
string of the SQL that would be used.

Call an arbitrary SQL function by providing func as a symbol or keyword
and its params

Call an arbitrary SQL function by providing the name of the function
and its params

Create a subselect clause to be used in queries. This works exactly like
(select ...) execept it will wrap the query in ( .. ) and make sure it can be
used in any current query:

(select users
  (where {:id [in (subselect users2 (fields :id))]}))

Set the name of the table and an optional alias to be used for the entity.
By default the table is the name of entity's symbol.

Add a function to be applied to results coming from the database

Creates a union query, applies any modifying functions in the body and then
executes it.

ex: (union
      (queries (subselect user
                 (where {:id 7}))
               (subselect user-backup
                 (where {:id 7})))
      (order :name))

Create an empty union query.

Creates a union-all query, applies any modifying functions in the body and then
executes it.

ex: (union-all
      (queries (subselect user
                 (where {:id 7}))
               (subselect user-backup
                 (where {:id 7})))
      (order :name))

Create an empty union-all query.

Creates an update query, applies any modifying functions in the body and then
executes it. `ent` is either a string or an entity created by defentity.
Returns number of updated rows as provided by the JDBC driver.

ex: (update user
      (set-fields {:name "chris"})
      (where {:id 4}))

Create an empty update query. Ent can either be an entity defined by defentity,
or a string of the table name.

Add records to an insert clause. values can either be a vector of maps or a
single map.

(values query [{:name "john"} {:name "ed"}])

Add a where clause to the query, expressing the clause in clojure expressions
with keywords used to reference fields.
e.g. (where query (or (= :hits 1) (> :hits 5)))

Available predicates: and, or, =, not=, <, >, <=, >=, in, like, not, between

Where can also take a map at any point and will create a clause that compares keys
to values. The value can be a vector with one of the above predicate functions
describing how the key is related to the value:
  (where query {:name [like "chris"]})

Add a where clause to the query. Clause can be either a map or a string, and
will be AND'ed to the other clauses.

Add a related entity to the given select query. If the entity has a relationship
type of :belongs-to or :has-one, the requested fields will be returned directly in
the result map. If the entity is a :has-many, a second query will be executed lazily
and a key of the entity name will be assoc'd with a vector of the results.

(defentity email (entity-fields :email))
(defentity user (has-many email))
(select user
  (with email) => [{:name "chris" :email [{email: "c@c.com"}]} ...

With can also take a body that will further refine the relation:
(select user
   (with address
      (with state)
      (fields :address.city :state.state)
      (where {:address.zip x})))

Add a related entity. This behaves like `with`, except that, for has-many
relationships, it runs a single query to get relations of all fetched rows.
This is faster than regular `with` but it doesn't support many of the
additional options (order, limit, offset, group, having)