S-SQL Reference Manual

This is the reference manual for the S-SQL component of the postmodern library.

S-SQL provides a lispy syntax for SQL queries, and knows how to convert various lisp types to their textual SQL representation. It takes care to do as much of the work as possible at compile-time, so that at runtime a string concatenation is all that is needed to produce the final SQL query.

Interface

macro sql (form)

→ string

Convert the given form (a list starting with a keyword) to an SQL query string at compile time, according to the rules described here. For example:

(sql (:select '* :from 'country :where (:= 'a 1)))

 "(SELECT * FROM country WHERE (a = 1))"

but

(sql '(:select '* :from 'country :where (:= 'a 1)))

would throw an error. For the later case you need to use sql-compile.

function sql-compile (form)

→ string

This is the run-time variant of the sql macro. It converts the given list to an SQL query, with the same rules except that symbols in this list do not have to be quoted to be interpreted as identifiers. For example:

(sql-compile '(:select '* :from 'country :where (:= 'a 1)))

\"(SELECT * FROM country WHERE (a = 1))\"

but

(sql (:select '* :from 'country :where (:= 'a 1)))

would throw an error. For the later case you need to use sql.

function sql-template (form)

In cases where you do need to build the query at run time, yet you do not want to re-compile it all the time, this function can be used to compile it once and store the result. It takes an S-SQL form, which may contain \[ placeholder symbols, and returns a function that takes one argument for every \]. When called, this returned function produces an SQL string in which the placeholders have been replaced by the values of the arguments.

function enable-s-sql-syntax (&optional (char #\Q))

Modifies the current readtable to add a #Q syntax that is read as (sql …). The character to use can be overridden by passing an argument.

function sql-escape-string (string)

→ string

Escapes a string for inclusion in a PostgreSQL query. A quoted symbol will generate an error. Example:

(sql-escape-string \"Puss in 'Boots'\")

\"E'Puss in ''Boots'''\"

method sql-escape (value)

→ string

A generalisation of sql-escape-string looks at the type of the value passed, and properly writes it out it for inclusion in an SQL query. Symbols will be converted to SQL names. Examples:

(sql-escape "tr'-x")

"E'tr''-x'"

(sql-escape (/ 1 13))

"0.0769230769230769230769230769230769230"

(sql-escape #("Baden-Wurttemberg" "Bavaria" "Berlin" "Brandenburg"))

"ARRAY[E'Baden-Wurttemberg', E'Bavaria', E'Berlin', E'Brandenburg']"

variable *downcase-symbols*

When converting symbols to strings, whether to downcase the symbols is set here. The default is to downcase symbols.

variable *standard-sql-strings*

Used to configure whether S-SQL will use standard SQL strings (just replace #\' with ''), or backslash-style escaping. Setting this to NIL is always safe, but when the server is configured to allow standard strings (compile-time parameter 'standard_conforming_strings' is 'on', which will become the default in future versions of PostgreSQL), the noise in queries can be reduced by setting this to T.

variable *postgres-reserved-words* hashtable

A set of all Postgresql's reserved words, for automatic escaping. Probably not a good idea to use these words as identifiers anyway. '("all" "analyse" "analyze" "and" "any" "array" "as" "asc" "asymmetric" "authorization" "between" "binary" "both" "case" "cast" "check" "collate" "column" "concurrently" "constraint" "create" "cross" "current-catalog" "current-date" "current-role" "current-schema" "current-time" "current-timestamp" "current-user" "default" "deferrable" "desc" "distinct" "do" "else" "end" "except" "false" "fetch" "filter" "for" "foreign" "freeze" "from" "full" "grant" "group" "having" "ilike" "in" "initially" "inner" "intersect" "into" "is" "isnull" "join" "lateral" "leading" "left" "like" "limit" "localtime" "localtimestamp" "natural" "new" "not" "notnull" "nowait" "null" "off" "offset" "old" "on" "only" "or" "order" "outer" "overlaps" "placing" "primary" "references" "returning" "right" "select" "session-user" "Share" "similar" "some" "symmetric" "table" "then" "to" "trailing" "true" "union" "unique" "user" "using" "variadic" "verbose" "when" "where" "window" "with"))

variable *escape-sql-names-p*

Determines whether double quotes are added around column, table, and ** function names in queries. Valid values:

  • T, in which case every name is escaped,
  • NIL, in which case no name is escape,
  • :auto, which causes only reserved words to be escaped, or.
  • :literal which is the same as :auto except it has added consequence in to-sql-name (see below).

The default value is :auto.

Be careful when binding this with let and such ― since a lot of SQL compilation tends to happen at compile-time, the result might not be what you expect. Mixed case sensitivity is not currently well supported. Postgresql itself will downcase unquoted identifiers. This will be revisited in the future if requested.

function sql-type-name (type)

→ string

Transform a lisp type into a string containing something SQL understands. Default is to just use the type symbol's name.

function to-sql-name (name &optional (escape-p *escape-sql-names-p*)(ignore-reserved-words nil)

→ string

Convert a symbol or string into a name that can be a sql table, column, or operation name. Add quotes when escape-p is true, or escape-p is :auto and the name contains reserved words. Quoted or delimited identifiers can be used by passing :literal as the value of escape-p. If escape-p is :literal, and the name is a string then the string is still escaped but the symbol or string is not downcased, regardless of the setting for *downcase-symbols* and the hyphen and forward slash characters are not replaced with underscores.

Ignore-reserved-words is only used internally for column names which are allowed to be reserved words, but it is not recommended.

function from-sql-name (string)

→ keyword

Convert a string that represents an SQL identifier to a keyword by uppercasing it and converting the underscores to dashes.

macro register-sql-operators (arity &rest names)

Define simple SQL operators. Arity is one of :unary (like 'not'), :unary-postfix (the operator comes after the operand), :n-ary (like '\+': the operator falls away when there is only one operand), :2+-ary (like '=', which is meaningless for one operand), or :n-or-unary (like '-', where the operator is kept in the unary case). After the arity may follow any number of operators, either just a keyword, in which case the downcased symbol name is used as the SQL operator, or a two-element list containing a keyword and a name string.

SQL Types

S-SQL knows the SQL equivalents to a number of Lisp types, and defines some extra types that can be used to denote other SQL types. The following table (yes, I know this table is duplicated on other pages) shows the correspondence:

Lisp type SQL type Description
integer smallint -32,768 to +32,768 2-byte storage
integer integer -2147483648 to +2147483647 integer, 4-byte storage
integer bigint -9223372036854775808 to 9223372036854775807 8-byte storage
(numeric X Y) numeric(X, Y) see discussion below
float, real real float, 6 decimal digit precision 4-byte storage
double-float double-precision float, 15 decimal digit precision 8-byte storage
string, text text variable length string, no limit specified
string char(X) char(length), blank-padded string, fixed storage length
string varchar(X) varchar(length), non-blank-padded string, variable storage
boolean boolean boolean, 'true'/'false', 1 byte
bytea bytea binary string which allows non-printable octets
date date date range: 4713 BC to 5874897 AD
interval interval time intervals
array array See Array-Notes

Numeric and decimal are variable storage size numbers with user specified precision. Up to 131072 digits before the decimal point; up to 16383 digits after the decimal point.

The syntax is numeric(precision, scale). Numeric columns with a specified scale will coerce input values to that scale. For more detail, see https://www.postgresql.org/docs/current/datatype-numeric.html

type db-null

This is a type of which only the keyword :null is a member. It is used to represent NULL values from the database.

SQL Syntax

An S-SQL form is converted to a query through the following rules:

  • Lists starting with a keyword are operators. They are expanded as described below if they are known, otherwise they are expanded in the standard way: operator(arguments, …)
  • Quoted symbols or keywords are interpreted as names of columns or tables, and converted to strings with to-sql-name.
  • Anything else is evaluated and the resulting Lisp value is converted to its textual SQL representation (or an error is raised when there is no rule for converting objects of this type). Self-quoting atoms may be converted to strings at compile-time.

sql-op :select (&rest args)

Creates a select query. The arguments are split on the keywords found among them. The group of arguments immediately after :select is interpreted as the expressions that should be selected. After this, an optional :distinct may follow, which will cause the query to only select distinct rows, or alternatively :distinct-on followed by a group of row names. Next comes the optional keyword :from, followed by at least one table name and then any number of join statements.

Join statements start with one of :join, :left-join, :right-join, :inner-join, :outer-join, :cross-join (or those with -lateral, e.g :join-lateral, :left-join-lateral, :right-join-lateral, :inner-join-lateral, :outer-join-lateral). S-sql will accept :join, but best usage is to explicitly use :inner-join instead.

Then comes a table name or subquery,

Then there is an optional :with-ordinality or :with-ordinality-as alisa

Then the keyword :on or :using, if applicable, and then a form. A join can be preceded by :natural (leaving off the :on clause) to use a natural join.

After the joins an optional :where followed by a single form may occur.

Finally :group-by and :having can optionally be specified. The first takes any number of arguments, and the second only one.

A few examples:

(query (:select 'item :distinct
        :from 'item-table
        :where (:= 'col1 "Albania")))

(query (:select (:+ 'field-1 100) 'field-5
        :from (:as 'my-table 'x)
        :left-join 'your-table
        :on (:= 'x.field-2 'your-table.field-1)
        :where (:not-null 'a.field-3)))

(query (:order-by
         (:select 'regions.name
            (:count 'regions.name)
          :from 'countries 'regions
          :where (:= 'regions.id 'countries.region-id)
          :group-by 'regions.name)
        'regions.name))

(query (:select (:count 'c.id) 'r.name
        :from (:as 'countries 'c)
        :inner-join (:as 'regions 'r)
        :on (:= 'c.region-id 'r.id)
        :group-by 'r.name
        :having (:< (:count 'c.id) 10)))

(query (:select 'i.* 'p.*
        :from (:as 'individual 'i)
        :inner-join (:as 'publisher 'p)
        :using ('individualid)
        :left-join-lateral (:as 'anothertable 'a)
        :on (:= 'a.identifier 'i.individualid)
        :where (:= 'a.something \"something\")))

(query (:select 't1.id 'a.elem 'a.nr
        :from (:as 't12 't1)
        :left-join (:unnest (:string-to-array 't1.elements ","))
        :with-ordinality-as (:a 'elem 'nr)
        :on 't))

Other examples can be found in s-sql/tests/tests.lisp

Joins

Allowable join keywords are:

  • :left-join
  • :right-join
  • :inner-join
  • :outer-join
  • :cross-join
  • :join-lateral
  • :left-join-lateral (left join with an additional sql keyword LATERAL)
  • :right-join-lateral (right join with an additional sql keyword LATERAL)
  • :inner-join-lateral (inner join with an additional sql keyword LATERAL)
  • :outer-join-lateral (outer join with an additional sql keyword LATERAL)
  • :cross-join-lateral (cross join with an additional sql keyword LATERAL)

The lateral joins will not be discussed separately.

Cross Join/ Cross Join Lateral

From the postgresql documentation: "For every possible combination of rows from T1 and T2 (i.e., a Cartesian product), the joined table will contain a row consisting of all columns in T1 followed by all columns in T2. If the tables have N and M rows respectively, the joined table will have N * M rows."

(query (:select '* from 'employee :cross-join 'compensation))

Inner Join / Inner Join Lateral

An inner join looks at two tables and creates a new result consisting of the selected elements in the rows from the two tables that match the specified conditions. You can simplistically think of it as the intersection of the two sets. In reality, it is creating a new set consisting of certain elements of the intersecting rows. An inner join is the default and need not be specified.

A sample of standard sql using two inner joins to collect information from three tables could look like this:

(SELECT foo, bar, baz
 FROM (SELECT foo FROM x WHERE some-condition-here) AS tmp1
 INNER JOIN (SELECT bar FROM x WHERE some-condition-here) AS tmp2
 ON (tmp1.id = tmp2.id)
 INNER JOIN (SELECT baz FROM x WHERE some-condition-here) AS tmp3
 ON (tmp2.id = tmp3.id))

The same query could be expressed in s-sql as:

(query (:select 'foo 'bar 'baz
                    :from (:as
                           (:select 'foo
                                    :from 'x
                                    :where 'x) 'tmp1)
                    :inner-join (:as
                                 (:select 'bar
                                          :from 'x
                                          :where 'x) 'tmp2)
                    :on (:= 'tmp1.id 'tmp2.id)
                    :inner-join (:as
                                 (:select 'baz
                                          :from 'x
                                          :where 'x)
                                 'tmp3)
                    :on (:= 'tmp2.id 'tmp3.id)))

The pre-ansi shorthand example, using a countries and regions tables would look like this:

(query (:select 'countries.name
                :from 'countries 'regions
                :where (:and (:= 'countries.region-id 'regions.id)
                             (:= 'regions.name "North America"))))

The full portable ansi version, using inner join would look like this.

(query (:select 'tmp1.name :from (:as (:select 'name 'region-id
                                               :from 'countries)
                                 'tmp1)
                :inner-join (:as (:select 'id
                                          :from 'regions
                                          :where (:= 'name "North America"))
                                 'tmp2)
                :on (:= 'tmp1.region-id 'tmp2.id)))

Outer Join / Outer Join Lateral

An outer join not only generates an inner join, it also joins the rows from one table that matches the conditions and adds null values for the joined columns from the second table (which obviously did not match the condition.) Under Postgresql, a "left join", "right join" or "full join" all imply an outer join.

A left join (or left outer join) looks at two tables, keeps the matched rows from both and the unmatched rows from the left table and drops the unmatched rows from the right table. A right outer join keeps the matched rows, the unmatched rows from the right table and drops the unmatched rows from the left table. A full outer join includes the rows that match from each table individually, with null values for the missing matching columns.

Left Join / Left Join Lateral / Right Join / Right Join Lateral

Example: Here we assume two tables. A countries table and a many-to-many linking table named countries-topics. (There is an implicit third table named topics.) We are looking for records from the countries table which do not have a match in the countries-topics table. In other words, where do we have a note, but not matched it to a topic?

(query (:order-by (:select 'countries.id 'countries.name
                               :distinct :from 'countries
      :left-join 'countries-topics
      :on (:= 'countries.id 'countries-topics.country-id)
      :where (:is-null 'countries-topics.country-id))
                      'countries.id))

Here is a somewhat contrived example using a countries and regions table. We want to get the names of all the regions and also return the country names in one specified region. Assume that we only want the names of the countries in Central America, which happens to have a region-id of 3.

(query (:select 'tmp2.name 'tmp1.name
         :from (:as (:select 'id 'name
                     :from 'regions)
                'tmp2)
         :left-join (:as (:select 'name 'region-id
                          :from 'countries
                          :where (:= 'region-id 3))
                     'tmp1)
         :on (:= 'tmp1.region-id 'tmp2.id)))

Defined Operators

The following operators are defined:

sql-op :+, :*, :%, :&, :|, :||, :and, :or, :=, :/, :!=, :<, :>, :<=, :>=, :^, :union, :union-all, :call, :intersect, :intersect-all, :except, :except-all (&rest args)

These are expanded as infix operators. When meaningful, they allow more than two arguments. :- can also be used as a unary operator to negate a value. Note that the arguments to :union, :union-all, :intersect, and :except should be queries (:select forms).

Note that you'll have to escape pipe characters to enter them as keywords. S-SQL handles the empty keyword symbol (written :||) specially, and treats it like :\|\|, so that it can be written without escapes. With :\|, this doesn't work.

We also have registered operators for Postgis, hstore and cube operations. See the following registration examples from the source code which might be useful if you need operators for other extensions:

  ;; PostGIS operators
(register-sql-operators :2+-ary :&& :&< :|&<\|| :&> :<< :|<<\|| :>> :|@| :|\|&>|
                        :|\|>>| :~= :|@>| :|@<|)

;; hstore operators
(register-sql-operators :2+-ary :-> :=> :? :?& :?\| :|<@| :|#=| :unary :%% :%#)

;; cube operators
(register-sql-operators :2+-ary :&& :|@>| :|<@| :-> :~> :<-> :<#> :<=>)

sql-op :or

(query (:select 'countries.name
                :from 'countries 'regions
                :where (:and
                        (:or (:= 'regions.name "North America")
                             (:= 'regions.name "Central America"))
                        (:= 'regions.id 'countries.region-id))))

or using parameterized queries

(query (:select 'countries.name
                :from 'countries 'regions
                :where (:and
                        (:or (:= 'regions.name '$1)
                             (:= 'regions.name '$2))
                        (:= 'regions.id 'countries.region-id)))
  "North America" "Central America")

sql-op :call

:Call calls a Postgresql procedure. Remember that Postgresql procedures are not called within a select statement. The procedure name must be quoted or be a variable with a value of a quoted procedure name. Assuming a procedure named set_x_procedure1 exists, an example of calling it with parameters would be:

(query (:call 'set_x_procedure1 1 13))

(let ((a 1) (b 2) (p 'my-proc))
  (query (:call p a b 3)))

For more discussion, see Calling Postgresql Stored Functions and Procedures.

sql-op :intersect

Intersect produces a result contain rows that appear on all the sub-selects.

(query (:intersect (:select 'countries.name
                            :from 'countries
                            :where (:< 'latitude 16.44))
                   (:select 'countries.name
                            :from 'countries 'regions
                            :where (:and (:= 'region-id 'regions.id)
                                         (:= 'regions.name "Caribbean")))))

sql-op :union, :union-all

The union operation generally eliminates what it thinks are duplicate rows. The union-all operation preserves duplicate rows. The examples below use the union-all operator, but the syntax would be the same with union.

(query (:select 'id 'name
                :from (:as (:union-all
                            (:select 'id 'name :from 'countries
                                     :where (:<= 'name "B" ))
                            (:select 'id 'name :from 'countries
                                     :where (:>= 'name "V" )))
                           'a)))

(query (:select 'a.id 'a.name 'a.region
                :from (:as (:union-all
                            (:select 'countries.id 'countries.name
                                     (:as 'regions.name 'region)
                                     :from 'countries 'regions
                                     :where (:and
                                             (:<= 'countries.name "B" )
                                             (:= 'regions.id 'countries.region-id )))
                            (:select 'countries.id 'countries.name
                                     (:as 'regions.name 'region)
                                     :from 'countries 'regions
                                     :where (:and
                                             (:>= 'countries.name "V" )
                                             (:= 'regions.id 'countries.region-id ))))
                           'a)
                :group-by 'a.id 'a.region 'a.name))

sql-op :except, :except-all

:except removes all matches. :except-all is slightly different. If the first select statement has two rows that match a single row in the second select statement, only one is removed.

(query (:except (:select 'id 'name
                         :from 'countries
                         :where (:like 'name "%New%"))
                (:select 'id 'name
                         :from 'countries
                         :where (:like 'name "%Zealand%"))))

(query (:except-all (:select '* :from 'clients) (:select '* :from 'vips)))

sql-op :~, :not (arg)

Unary operators for bitwise and logical negation.

(query (:order-by (:select 'recommendedby
                           (:count '*)
                           :from 'cd.members
                           :where (:not (:is-null 'recommendedby))
                           :group-by 'recommendedby)
                  'recommendedby))

sql-op :any, :any*

Any needs to be considered as a special case. Quoting Marijn Haverbeke here,"Postgres has both a function-call-style any and an infix any, and S-SQL's syntax doesn't allow them to be distinguished." As a result, postmodern has a regular :any sql-op and a :any* sql-op, which expand slightly differently.

  (sql (:select 'name :from 'employee :where (:= 'id (:any '$1))))
"(SELECT name FROM employee WHERE (id = ANY $1))"

  (sql (:select 'name :from 'employee :where (:= 'id (:any* '$1))))
"(SELECT name FROM employee WHERE (id = ANY($1)))"

In general, the any qualifier in an sql statement looks at a subquery and does a comparison against that subquery. Sticking with our countries table, we have latitude, longitude data for every country (I'm not sure whether my table pulled the capital cities or the geographic center) and some designated a region for each country, so we have a region-id that matches the primary key 'id' in a regions table.

Out of curiosity, let's determine which countries in "North America" have a longitude less than any country in "South America". The standard sql could look like this:

(query "select countries.name
        from countries,regions
        where regions.id=region_id
              and regions.name='North America'
              and longitude > any(select longitude
                                         from countries, regions
                                         where region_id = regions.id
                                         and regions.name='South America')")
(("Bermuda") ("Greenland"))

This can be re-phrased in s-sql as

(query
 (:select 'countries.name
          :from 'countries 'regions
          :where (:and (:= 'regions.id 'region-id)
                       (:= 'regions.name "North America")
                       (:> 'longitude
                           (:any
                            (:select 'longitude
                                     :from 'countries 'regions
                                     :where (:and (:= 'regions.id 'region-id)
                                                  (:= 'regions.name "South America"))))))))
(("Bermuda") ("Greenland"))

Subselects work fine in both regular sql and s-sql. If you have already calculated your subselect and put it in a variable, for versions of Postmodern before 1.33.7 that variable needs to be a vector and whether you should use the :any sql-op or the :any* sql-op depends on your phrasing. (Notice that the second variation has an asterisk). (If you try to use a list, you will trigger an error message that you cannot convert that into an sql literal.) For versions of Postmodern 1.33.7 or higher, you can use a list.

The SQL keyword ANY can be used in a parameterized sql statement if you provide it with a vector (for Postmodern versions pre 1.33.7) or (for Postmodern versions 1.33.7 or higher) either a vector or a list. The following two toy examples work in RAW sql.

(query "select name from countries where id=any($1)"
       (vector 21 22))
(("Iceland") ("US"))

(let ((toy-query (vector 21 22)))
     (query "select name from countries where id=any($1)"
       toy-query))
(("Iceland") ("US"))

Now using s-sql and keeping with the toy example, notice that using :any does not work, but using :any* does work.

(let ((toy-query (vector 21 22)))
  (query (:select 'name
                  :from 'countries
                  :where (:= 'id (:any '$1)))
         toy-query))
; Evaluation aborted on #<CL-POSTGRES-ERROR:SYNTAX-ERROR-OR-ACCESS-VIOLATION {10030AF6A1}>.

(let ((toy-query (vector 21 22)))
  (query (:select 'name
                  :from 'countries
                  :where (:= 'id (:any* '$1)))
         toy-query))
(("Iceland") ("US"))

;; For Postmodern versions 1.33.7 or higher, you can also use a list
(let ((toy-query (list 21 22)))
  (query (:select 'name
                  :from 'countries
                  :where (:= 'id (:any* '$1)))
         toy-query))
(("Iceland") ("US"))

Going back to our earlier example, remember that I said that for Postmodern versions before 1.33.7 unless you use a subselect, you need to provide a vector to :any or :any*. A standard query returns a list, not a vector. So you would need to coerce the variable into a vector before you pass it to :any*. See below as an example, noting the use of passing a :column parameter to the postmodern query function so that we just get back a list of the longitude, not a list of lists of longitude. For Postmodern versions 1.33.7 or higher, you do not have to coerce it to a vector, but you would still need to pass the :column parameter in the first select.

(let ((South-America
       (coerce
         (query (:select 'longitude
                         :from 'countries 'regions
                         :where (:and (:= 'regions.id 'region-id)
                                      (:= 'regions.name "South America")))
                 :column))
        'vector))
  (query (:select 'countries.name
                  :from 'countries 'regions
                  :where (:and (:= 'regions.id 'region-id)
                               (:= 'regions.name "North America")
                               (:> 'longitude
                                   (:any* South-America))))))
(("Bermuda") ("Greenland"))

sql-op :function (name (&rest arg-types) return-type stability body)

Create a stored procedure. The argument and return types are interpreted as type names and not evaluated. Stability should be one of :immutable, :stable, or :volatile (see the PostgreSQL documentation). For example, a function that gets foobars by id:

(:function 'get-foobar (integer) foobar :stable (:select '* :from 'foobar :where (:= 'id '$1)))

sql-op :~, :~*, :!~, :!~* (string pattern)

Regular expression matching operators. The exclamation mark means 'does not match', the asterisk makes the match case-insensitive.

(query (:select (:regexp_match "foobarbequebaz" "bar.*que")) :single)

#("barbeque")

(query (:select (:regexp_match "foobarbequebaz" "bar.~que")) :single)

:NULL

(query (:select (:~ "foobarbequebaz" "bar.*que") ) :single)

t

(query (:select (:!~ "foobarbequebaz" "bar.*que") ) :single)

nil

(query (:select (:~ "foobarbequebaz" "barque") ) :single)

nil

(query (:select (:~ "foobarbequebaz" "barbeque") ) :single)

t

(query (:select (:~ "foobarBequebaz" "barbeque") ) :single)

nil

(query (:select (:~* "foobarBequebaz" "barbeque") ) :single)

t

(query (:select 'id 'text :from 'text-search :where (:~ 'text "sushi")))

sql-op :like, :ilike (string pattern)

Simple SQL string matching operators (:ilike is case-insensitive).

(query (:select 'id 'name
                :from 'countries
                :where (:like 'name "%New%")))

sql-op :@@

Fast Text Search match operator.

sql-op :desc (column)

Used to invert the meaning of an operator in an :order-by clause.

(query (:order-by
        (:select 'location 'time 'report
                 :distinct-on 'location
                 :from 'weather-reports)
        'location  (:desc 'time)))

sql-op :nulls-first, :nulls-last (column)

Used to determine where :null values appear in an :order-by clause.

sql-op :as (form name &rest fields)

Also known in some explanations as "alias". This assigns a name to a column or table in a :select form. When fields are given, they are added after the name, in parentheses. For example, (:as 'table1 't1 'foo 'bar) becomes table1 AS t1(foo, bar). When you need to specify types for the fields, you can do something like (:as 'table2 't2 ('foo integer)). Note that names are quoted, types are not (when using sql-compile or sql-template, you can leave out the quotes entirely).

(query (:select (:as 'countries.name 'country)
                (:as 'regions.name 'region)
                :from 'countries 'regions
                       :where (:and (:= 'regions.id 'region-id)
                                     (:= 'regions.name "Central America")))
        :alists)

(((:COUNTRY . "Belize") (:REGION . "Central America")) ((:COUNTRY . "Costa Rica")
(:REGION . "Central America")) ((:COUNTRY . "El Salvador")
(:REGION . "Central America")) ((:COUNTRY . "Guatemala")
(:REGION . "Central America")) ((:COUNTRY . "Panama") (:REGION . "Central America"))
((:COUNTRY . "Nicaragua") (:REGION . "Central America")))

The following uses aliases for both columns and tables in the from and inner-join clauses:

(query (:order-by
        (:select (:as 'recs.firstname 'firstname)
                 (:as 'recs.surname 'surname)
                 :distinct
                 :from (:as 'cd.members 'mems)
                 :inner-join (:as 'cd.members 'recs)
                 :on (:= 'recs.memid 'mems.recommendedby))
        'surname 'firstname))

Note: Postmodern does not allow you to create an unescaped string alias. In other words, you cannot generate this:

"select sum(slots as "Total Slots" from cd.bookings"

without using :raw

sql-op :cast (query)

The CAST operator. Takes a query as an argument, and returns the result explicitly cast by postgresql to a specific type. Unlike :type, :cast can pass the type as a variable.

(query (:select (:cast (:as "20" 'integer)))
        :single)
20

(let ((type 'text))
   (query (:select (:cast (:as "20" type)))
     :single))
"20"

(let ((type 'integer))
   (query (:select (:cast (:as "20" type)))
     :single))
20

(query (:union (:select (:as 1 'real))
               (:select (:cast (:as "2.2" 'real)))))
((1.0) (2.2))

sql-op :type (query)

Is similar to cast but uses the postgresql :: formating. Unlike cast it will not accept a variable as the type.

E.g.

(sql (:select (:as (:- (:type (:now) 'date) 'x) 'some-date) :from (:as (:generate-series 1 10) 'x)))

"(SELECT (now()::DATE - x) AS some_date FROM generate_series(1, 10) AS x)"

sql-op :type (form type)

Add a type declaration to a value, as in in "4.3::real". The second argument is not evaluated normally, but put through sql-type-name to get a type identifier.

(query (:select (:type 1.0 int)))

sql-op :create-composite-type (type-name &rest args)

Creates a composite type with a type-name and two or more columns. E.g.

(query (:create-composite-type 'fullname (first-name text) (last-name text)))

sql-op :exists (query)

The EXISTS operator. Takes a query as an argument, and returns true or false depending on whether that query returns any rows. In the example below, it is applied to a subquery.

(query (:select 'id 'name
                :from 'regions
                :where (:exists
                        (:select 'region-id
                         :from 'countries
                         :where (:and
                                  (:= 'countries.name "Costa Rica")
                                  (:= 'regions.id 'countries.region-id))))))

sql-op :is-false (arg)

Test whether a boolean value is false.

(query (:select 'ta :from 'a :where (:is-false 'ta)))
(select-dao 'account (:is-false 'active))

sql-op :is-true (arg)

Test whether a boolean value is true.

(query (:select 'ta :from 'a :where (:is-true 'ta)))
(select-dao 'account (:is-true 'active))

sql-op :is-null (arg)

Test whether a value is null.

(query (:select 'ta :from 'a :where (:not (:is-null 'ta))))

sql-op :not-null (arg)

Test whether a value is not null.

(query (:select 'ta :from 'a :where (:not-null 'ta)))

sql-op :in (value set)

Test whether a value is in a set of values.

(query (:select 'name
                :from 'countries
                :where (:in 'id
                            (:set 20 21 23))))

(query (:select 'region 'product (:as (:sum 'quantity) 'product-units)
                (:as (:sum 'amount) 'product-sales)
                :from 'orders
                :where (:in 'region (:select 'region :from 'top-regions))
                :group-by 'region 'product))

Beginning in Postmodern v. 1.33.7, you can use lists with :in.

(query (:select '* :from 'employee :where (:in 'id '(1 3 4))))

(let ((emp-ids '(1 2)))
  (query (:select 'name :from 'employee :where (:in 'id emp-ids))))

IMPORTANT REGARDING PARAMETERIZED QUERIES: For Postmodern versions before 1.33.7, you cannot use a list in a parameterized statement. You have to convert the list to a vector and use :any* rather than :in. See S-SQL#sql-op-any for more details.

(query (:select 'name :from 'employee :where (:= 'id (:any* '$1)))
            #(1 3) :column)
 '("Jason" "Celia")

;; Beginning with Postmodern version 1.33.7 you can also use lists

(query (:select 'name :from 'employee :where (:= 'id (:any* '$1)))
            '(1 3) :column)
 '("Jason" "Celia")

sql-op :not-in (value set)

Inverse of the above.

(query (:select '* :from 'table1 :where (:not-in 'id (:set 1 2 3))))

Beginning in Postmodern v. 1.33.7, you can use lists with :in.

(query (:select '* :from 'table1 :where (:not-in 'name '("alpha" "beta" "ceta"))))

sql-op :set (&rest elements)

Denote a set of values. This operator has two interfaces. When the elements are known at compile-time, they can be given as multiple arguments to the operator. When they are not, a single argument that evaluates to a list should be used.

The following would be the syntax in postmodern sql where the set is a list. If you want to use a vector, then you need to use Any:

The following are equivalent

(query (:select 'name
                :from 'countries
                :where (:in 'id
                            (:set 20 21 23))))

(let ((x (list 20 21 23)))
  (query (:select 'name
                  :from 'countries
                  :where (:in 'id
                              (:set x)))))

(query (:select 'name
                :from 'countries
                :where (:in 'id (:set (list 20 21 23)))))

However, the following will generate an error.

(query (:select 'name
                :from 'countries
                :where (:in 'id
                            (:set '(20 21 23)))))

Now with selecting a dao

(select-dao 'countries
            (:in 'id
                 (:set (list 20 21 23

Now with selecting from a vector. Note both the use of any* and := instead of :in.

(let ((x (vector 20 21 23)))
  (query (:select 'name
                  :from 'countries
                  :where (:= 'id (:any* x)))))

Note that the responses will still come back in a list of lists

IMPORTANT: :SET DOES NOT WORK IN PARAMETERIZED QUERIES. YOU NEED TO USE :ANY* RATHER THAN :IN. See S-SQL for more details.

sql-op :array (query)

This is used when calling a select query into an array. See array-notes.html for more detailed notes on the use of arrays.

(query (:order-by
        (:select 'r.rolename
                 (:as (:array
                       (:select 'b.rolename
                                :from (:as 'pg_catalog.pg-auth-members 'm)
                                :inner-join (:as 'pg-catalog.pg-roles 'b)
                                :on (:= 'm.roleid 'b.oid)
                                :where (:= 'm.member 'r.oid )))
                      'memberof)

                 :from (:as 'pg-catalog.pg-roles 'r))
        1))

sql-op :array[] (&rest args)

This is the general operator for arrays. It also handles statements that include functions in the query such as (:+ 1 2), (:pi) in the array. See array-notes.html for more detailed notes on the use of arrays.

(query (:select (:array-prepend 1 (:array[] 2 3))))

((#(1 2 3)))

(query (:select (:array-prepend 1 (:array[] 2 3)))
       :single)

#(1 2 3)

sql-op :[] (form start &optional end)

Dereference an array value. If end is provided, extract a slice of the array. Sample usage below, but also see array-notes.html for more detailed notes on the use of arrays.

(query (:select 'receipe-id (:[] 'tags 2 3)
                :from 'receipe-tags-array
                :where (:= 'receipe-id 3)))

sql-op :extract (unit form)

Extract a field from a date/time value. For example, (:extract :month (:now)).

(query (:order-by
        (:select 'facid
                 (:as (:extract 'month 'starttime) 'month)
                 (:as (:sum 'slots) 'total-slots)
                 :from 'cd.bookings
                 :where (:and (:>= 'starttime "2012-01-01")
                              (:< 'starttime "2013-01-01"))
                 :group-by 'facid 'month)
        'facid 'month))

sql-op :case (&rest clauses)

A conditional expression. Clauses should take the form (test value). If test is :else, an ELSE clause will be generated.

(query (:select 'name
                (:as (:case ((:> 'monthlymaintenance 100) "expensive")
                            (:else "cheap")) 'cost)
                :from 'cd.facilities))

sql-op :between (n start end)

Test whether a value lies between two other values.

(query (:select 'name
                :from 'countries
                :where (:between 'latitude -10 10))
       :column)

sql-op :between-symmetric (n start end)

Works like :between, except that the start value is not required to be less than the end value.

sql-op :dot (&rest names)

Can be used to combine multiple names into a name of the form A.B to refer to a column in a table, or a table in a schema. Note that you can also just use a symbol with a dot in it.

sql-op :raw (string)

Insert a string as-is into the query. This can be useful for doing things that the syntax does not support, or to re-use parts of a query across multiple queries:

(let* ((test (sql (:and (:= 'foo 22) (:not-null 'bar))))
       (rows (query (:select '* :from 'baz :where (:raw test)))))
  (query (:delete-from 'baz :where (:raw test)))
  (do-stuff rows))

sql-op :fetch (form amount &optional offset)

Fetch is a more efficient way to do pagination instead of using limit and offset. Fetch allows you to retrieve a limited set of rows, optionally offset by a specified number of rows. In order to ensure this works correctly, you should use the order-by clause. If the amount is not provided, it assumes you only want to return 1 row. https://www.postgresql.org/docs/current/sql-select.html

Examples:

(query (:fetch (:order-by (:select 'id :from 'historical-events) 'id) 5))

((1) (2) (3) (4) (5))

(query (:fetch (:order-by (:select 'id :from 'historical-events) 'id) 5 10))

((11) (12) (13) (14) (15))

sql-op :limit (query amount &optional offset)

In S-SQL limit is not part of the select operator, but an extra operator that is applied to a query (this works out better when limiting the union or intersection of multiple queries, same for sorting). It limits the number of results to the amount given as the second argument, and optionally offsets the result by the amount given as the third argument.

(query (:limit (:order-by (:select 'surname :distinct :from 'cd.members) 'surname) 10))

sql-op :order-by (query &rest exprs)

Order the results of a query by the given expressions. See :desc for when you want to invert an ordering. Note: This is not the same as passing an :order-by parameter to an aggregation operator. For that see Aggregation Operators.

(query (:order-by (:select 'id 'name 'city 'salary (:every (:like 'name "J%"))
                                          :from 'employee
                                          :group-by 'name 'id 'salary 'city)
                                 'name))

sql-op :values

Values computes a row value or set of row values for use in a specific query. See the postgresql docs at: https://www.postgresql.org/docs/current/static/queries-values.html and https://www.postgresql.org/docs/current/static/sql-values.html Example:

(query (:select '*
                :from (:as (:values (:set 1 "one")
                                    (:set 2 "two")
                                    (:set 3 "three"))
                           (:t1 'num 'letter))))

(query (:select 'a 'b 'c (:cast (:as (:* 50 (:random)) 'int))
                :from (:as (:values (:set "a") (:set "b")) (:d1 'a))
                (:as (:values (:set "c") (:set "d")) (:d2 'b))
                (:as (:values (:set "e") (:set "f")) (:d3 'c))))

(query
 (:with-recursive
  (:as (:t1 'n)
       (:union-all (:values (:set 1))
                   (:select (:+ 'n 1)
                            :from 't1
                            :where (:< 'n 100))))
  (:select (:sum 'n) :from 't1))
 :single)

sql-op :empty-set

This is a fudge. It returns a string "()" where something like '() would return "false" or :() would throw an error. Example:

(query (:select 'appnumber 'day (:sum 'inserts)
                (:sum 'updates) (:sum 'deletes) (:sum 'transactions)
                :from 'db-details
                :group-by (:grouping-sets (:set 'appnumber 'day (:empty-set)))))

sql-op :group-by

https://www.postgresql.org/docs/current/static/queries-table-expressions.html#QUERIES-GROUPING-SETS The GROUP BY Clause is used to group together those rows in a table that have the same values in all the columns listed. The order in which the columns are listed does not matter. The effect is to combine each set of rows having common values into one group row that represents all rows in the group. This is done to eliminate redundancy in the output and/or compute aggregates that apply to these groups. Example:

(query (:order-by
        (:select 'mems.surname 'mems.firstname 'mems.memid (:as (:min 'bks.starttime) 'starttime)
                 :from (:as 'cd.bookings 'bks)
                 :inner-join (:as 'cd.members 'mems)
                 :on (:= 'mems.memid 'bks.memid)
                 :where (:>= 'starttime "2012-09-01")
                 :group-by 'mems.surname 'mems.firstname 'mems.memid)
        'mems.memid))

sql-op :grouping-sets

https://www.postgresql.org/docs/current/static/queries-table-expressions.html#QUERIES-GROUPING-SETS More complex grouping operations are possible using the concept of grouping sets. The data selected by the FROM and WHERE clauses is grouped separately by each specified grouping set, aggregates computed for each group just as for simple GROUP BY clauses, and then the results returned. This operator requires postgresql 9.5 or later. For example:

(query (:select 'city (:as (:extract 'year 'start-date)  'joining-year) (:as (:count 1) 'employee_count)
                :from 'employee
                :group-by (:grouping-sets (:set 'city (:extract 'year 'start-date)))))

Time, Date and Interval Operators

sql-op :interval (arg)

Creates an interval data type, generally represented in postmodern as an alist

sql-op :current-date ()

(query (:select (:current-date)) :single)

sql-op :current-time ()

sql-op :current-timestamp ()

sql-op :timestamp (arg)

sql-op :age (&rest args)

sql-op :date (arg)

sql-op :make-interval (&rest args)

Takes lists of (time-unit value) and returns a timestamp type. Example:

(query (:select (:make-interval ("days" 4) ("hours" 10) ("secs" 1.2)))
       :single)

sql-op :make-timestamp (&rest args)

Takes lists of (time-unit value) and returns a timestamptz type. Example:

(query (:select
          (:make-timestamptz ("year" 2014) ("month" 1) ("mday" 13)
                             ("hour" 21) ("min" 50) ("sec" 0)))
       :single)

sql-op :make-timestamptz (&rest args)

Takes lists of (time-unit value) and returns a timestamptz type. Example:

(query (:select
          (:make-timestamptz ("year" 2014) ("month" 1) ("mday" 13)
                             ("hour" 21) ("min" 50) ("sec" 0) ("timezone" "Asia/Tokyo")))
       :single)

Aggregation Operators

sql-op :count (&rest args)

Count returns the number of rows for which the expression is not null. It can be the number of rows collected by the select statement as in:

(query (:select (:count '*)
                :from 'table1
                :where (:= 'price 100)))

or it can be a smaller number of rows based on the allowed keyword parameters :distinct and :filter or some other type of condition as in:

(query (:select (:count 'memid :distinct)
                :from 'cd.bookings))

or

(query (:select (:as (:count '* :distinct) 'unfiltered)
                (:as (:count '* :filter (:= 1 'bid))
                     'filtered)
                :from 'testtable))

Note that if used, the filter must be last in the count args. If distinct is used, it must come before filter. Unlike standard sql, the word 'where' is not used inside the filter clause. E.g.

(query (:select (:count '*)
                (:count '* :filter (:= 1 'bid))
                'id
                :from 'pbbench-history))

See tests.lisp for examples.

sql-op :avg (&rest rest args)

Avg calculates the average value of a list of values. Note that if the filter keyword is used, the filter must be last in the avg args. If distinct is used, it must come before filter. E.g. See tests.lisp for more examples.

(query (:select (:avg '*) (:avg '* :filter (:= 1 'bid)) 'id
                :from 'pbbench-history))

sql-op :sum (&rest rest args)

Sum calculates the total of a list of values. Note that if the keyword filter is used, the filter must be last in the sum args. If distinct is used, it must come before filter. Unlike standard sql, the word 'where' is not used inside the filter clause (s-sql will properly expand it). See tests.lisp for more examples.

(query (:select (:sum '*) (:sum '* :filter (:= 1 'bid)) 'id
                :from 'pbbench-history))

sql-op ::max (&rest args)

max returns the maximum value of a set of values. Note that if the filter keyword is used, the filter must be last in the max args. If distinct is used, it must come before filter. Unlike standard sql, the word 'where' is not used inside the filter clause (s-sql will properly expand it). See tests.lisp for more examples.

(query (:select (:max '*) (:max '* :filter (:= 1 'bid)) 'id
                :from 'pbbench-history))

sql-op ::min (&rest args)

min returns the minimum value of a set of values. Note that if the filter keyword is used, the filter must be last in the min args. If distinct is used, it must come before filter. Unlike standard sql, the word 'where' is not used inside the filter clause (s-sql will properly expand it). See tests.lisp for more examples.

(query (:select (:min '*) (:min '* :filter (:= 1 'bid)) 'id
                :from 'pbbench-history))

sql-op ::every (&rest args)

Every returns true if all input values are true, otherwise false. Note that if the filter keyword is used, the filter must be last in the every args. If distinct is used, it must come before filter. Unlike standard sql, the word 'where' is not used inside the filter clause (s-sql will properly expand it). See tests.lisp for more examples.

(query (:select '* (:every (:like 'studname "%h"))
                :from 'tbl-students
                :group-by 'studname 'studid 'studgrades))

sql-op :percentile-cont (&rest args)

Requires Postgresql 9.4 or higher. Percentile-cont returns a value corresponding to the specified fraction in the ordering, interpolating between adjacent input items if needed. There are two required keyword parameters :fraction and :order-by. If the fraction value is an array, then it returns an array of results matching the shape of the fractions parameter, with each non-null element replaced by the value corresponding to that percentile. Examples:

(query (:select (:percentile-cont :fraction 0.5 :order-by 'number-of-staff)
                :from 'schools))

(query (:select (:percentile-cont :fraction array[0.25 0.5 0.75 1]
                                  :order-by 'number-of-staff)
                :from  'schools))

sql-op :percentile-dist (&rest args)

Requires Postgresql 9.4 or higher. There are two required keyword parameters :fraction and :order-by. Percentile-dist returns the first input value whose position in the ordering equals or exceeds the specified fraction. If the fraction parameter is an array eturns an array of results matching the shape of the fractions parameter, with each non-null element replaced by the input value corresponding to that percentile. Examples:

(query (:select (:percentile-dist :fraction 0.5
                                  :order-by 'number-of-staff)
                :from 'schools))

(query (:select (:percentile-dist :fraction array[0.25 0.5 0.75 1]
                                  :order-by 'number-of-staff)
                :from  'schools))

sql-op :corr (y x)

The corr function returns the correlation coefficient between a set of dependent and independent variables. Example:

(query (:select (:corr 'height 'weight)
                :from 'people))

sql-op :covar-pop (y x)

The covar-pop function returns the population covariance between a set of dependent and independent variables. Example:

(query (:select (:covar-pop 'height 'weight)
                :from 'people))

sql-op :covar-samp (y x)

(query (:select (:covar-samp 'height 'weight)
                :from 'people))

The covar-samp function returns the sample covariance between a set of dependent and independent variables. Example:

sql-op :string-agg (&rest args)

String-agg allows you to concatenate strings using different types of delimiter symbols. Allowable optional keyword parameters are :distinct, :order-by and :filter Note that order-by in string-agg requires postgresql 9.0 or later. Filter requires postgresql 9.4 or later. See tests.lisp for more examples.

(query (:select (:as (:string-agg 'bp.step-type \",\" )
                     'step-summary)
                :from 'business-process))

(query (:select 'mid (:as (:string-agg  'y \",\" :distinct :order-by (:desc 'y))
                          'words)
                :from 'moves))

(query (:select (:string-agg  'name "," :order-by (:desc 'name) :filter (:< 'id 4))
                :from 'employee))

sql-op :array-agg (&rest args)

Array-agg returns a list of values concatenated into an arrays. Allowable optional keyword parameters are :distinct, :order-by and :filter.

Note that order-by in array-agg requires postgresql 9.0 or later. Filter requires postgresql 9.4 or later. See array-notes.html for more detailed notes on the use of arrays.

Example with Filter:

(query (:select 'g.id
                (:as (:array-agg 'g.users :filter (:= 'g.canonical \"Y\"))
                     'canonical-users)
                (:as (:array-agg 'g.users :filter (:= 'g.canonical \"N\"))
                     'non-canonical-users)
                :from (:as 'groups 'g)
                :group-by 'g.id))

sql-op :mode (&rest args)

Mode is used to find the most frequent input value in a group. See e.g. https://www.postgresql.org/docs/10/static/functions-aggregate.html#FUNCTIONS-ORDEREDSET-TABLE and article at https://tapoueh.org/blog/2017/11/the-mode-ordered-set-aggregate-function

(query (:select (:mode 'items)
                :from 'item-table))

sql-op :regr_avgx (y x)

The regr_avgx function returns the average of the independent variable (sum(X)/N) Example:

(query (:select (:regr_avgx 'height 'weight)
                :from 'people))

sql-op :regr_avgy (y x)

The regr_avgy function returns the average of the dependent variable (sum(Y)/N). Example:


(query (:select (:regr_avgy 'height 'weight) :from 'people))

sql-op :regr_count (y x)

The regr_count function returns the number of input rows in which both expressions are nonnull. Example:

(query (:select (:regr_count 'height 'weight)
                :from 'people))

sql-op :regr_intercept (y x)

The regr_intercept function returns the y-intercept of the least-squares-fit linear equation determined by the (X, Y) pairs. Example:

(query (:select (:regr_intercept 'height 'weight)
                :from 'people))

sql-op :regr_r2 (y x)

The regr_r2 function returns the square of the correlation coefficient. Example:

(query (:select (:regr_r2 'height 'weight)
                :from 'people))

sql-op :regr_slope (y x)

The regr_slope function returns the slope of the least-squares-fit linear equation determined by the (X, Y) pairs. Example:

(query (:select (:regr_slope 'height 'weight)
                :from 'people))

sql-op :regr_sxx (y x)

The regr_sxx function returns the sum(X^2) - sum(X)^2/N (“sum of squares” of the independent variable). Example:

(query (:select (:regr_sxx 'height 'weight)
                :from 'people))

sql-op :regr_sxy (y x)

The regr_sxy function returns the sum(X*Y) - sum(X) * sum(Y)/N (“sum of products” of independent times dependent variable). Example:

(query (:select (:regr_sxy 'height 'weight)
                :from 'people))

sql-op :regr_syy (y x)

The regr_syy function returns the sum(Y^2) - sum(Y)^2/N (“sum of squares” of the dependent variable). Example:

(query (:select (:regr_syy 'salary 'age)
                :from 'employee))

sql-op :stddev (&rest args)

The stddev function returns the the sample standard deviation of the input values. It is a historical alias for stddev-samp. Example:

(query (:select (:stddev 'salary)
                :from 'employee))

sql-op :stddev-pop (&rest args)

The stddev-pop function returns the population standard deviation of the input values. Example:

(query (:select (:stddev-pop 'salary)
                :from 'employee))

sql-op :stddev-samp (&rest args)

The stddev-samp function returns the sample standard deviation of the input values. Example:

(query (:select (:stddev-samp 'salary)
                :from 'employee))

sql-op :variance (&rest args)

Variance is a historical alias for var_samp. The variance function returns the sample variance of the input values (square of the sample standard deviation). Example:

(query (:select (:variance 'salary)
                :from 'employee))

sql-op :var-pop (&rest args)

The var-pop function returns the population variance of the input values (square of the population standard deviation). Example:

(query (:select (:var-pop 'salary)
                :from 'employee)
       :single)

sql-op :var-samp (&rest args)

The var-samp function returns the sample variance of the input values (square of the sample standard deviation). Example:

(query (:select (:var-samp 'salary)
                :from 'employee)
       :single)

Window Functions

sql-op :range-between (&rest args)

Range-between allows window functions to apply to different segments of a result set. It accepts the following keywords: :order-by, :rows-between, :range-between, :unbounded-preceding, :current-row and :unbounded-following. Use of :preceding or :following will generate errors. See https://www.postgresql.org/docs/current/sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS for Postgresql documentation on usage.

An example which calculates a running total could look like this:

(query
 (:select (:as 'country 'country-name)
          (:as 'population 'country-population)
          (:as (:over (:sum 'population)
                      (:range-between :order-by 'country
                                      :unbounded-preceding :current-row))
               'global-population)
  :from 'population
  :where (:and (:not-null 'iso2)
               (:= 'year 1976))))

sql-op :rows-between (&rest args)

Rows-between allows window functions to apply to different segments of a result set. It accepts the following keywords: :order-by, :rows-between, :range-between, :preceding, :unbounded-preceding, :current-row, :unbounded-following and :following. See https://www.postgresql.org/docs/current/sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS for Postgresql documentation on usage.

An example could look like this :

(query
 (:select (:as 'country 'country-name)
          (:as 'population 'country-population)
          (:as (:over (:sum 'population)
                      (:rows-between :order-by 'country :preceding 2 :following 2))
               'global-population)
  :from 'population
  :where (:and (:not-null 'iso2)
               (:= 'year 1976))))

sql-op :over (form &rest args)

Over, partition-by and window are so-called window functions. A window function performs a calculation across a set of table rows that are somehow related to the current row and adds that as an additional column to the result. The following collects individual salaries and the total salaries.

(query (:select 'salary (:over (:sum 'salary))
                :from 'empsalary))

A more complicated version that calculates a running total might look like:

(query
 (:select 'name
          (:as 'salary 'individual-salary)
          (:as (:over (:sum 'salary)
                      (:range-between :order-by 'name :unbounded-preceding
                       :current-row))
               'running-total-salary)
  :from 'empsalary))

sql-op :partition-by (&rest args)

Args is a list of one or more columns to partition by, optionally followed by other keywords. Partition-by accepts the following keywords: :order-by, :rows-between, :range-between, :preceding, :unbounded-preceding, :current-row, :unbounded-following and :following. See https://www.postgresql.org/docs/current/sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS for Postgresql documentation on usage.

(query (:select 'depname 'subdepname 'empno 'salary
                (:over (:avg 'salary)
                       (:partition-by 'depname 'subdepname))
                :from 'empsalary))

Note the use of :order-by without parens:

(query (:select 'depname 'empno 'salary
                (:over (:rank)
                       (:partition-by 'depname :order-by (:desc 'salary)))
                :from 'empsalary))

The following example shows a query for country population in 1976 with running total population by region.

(query
 (:select (:as 'population.country 'country-name)
          (:as 'population 'country-population)
          'region-name
          (:as (:over (:sum 'population)
                      (:partition-by 'region-name :order-by 'region-name
                       :rows-between :unbounded-preceding :current-row))
               'regional-population)
          :from 'population
          :inner-join 'regions
          :on (:= 'population.iso3 'regions.iso3)
          :where (:and (:not-null 'population.iso2)
                       (:= 'year 1976))))

sql-op :window (form)

(query (:select (:over (:sum 'salary) 'w)
                (:over (:avg 'salary) 'w)
                :from 'empsalary :window
                (:as 'w (:partition-by 'depname :order-by (:desc 'salary)))))

sql-op :with (&rest args)

With provides a way to write auxillary statements for use in a larger query, often referred to as Common Table Expressions or CTEs.

(query (:with (:as 'upd
                   (:parens
                    (:update 'employees :set 'sales-count (:+ 'sales-count 1)
                             :where (:= 'id
                                        (:select 'sales-person
                                                 :from 'accounts
                                                 :where (:= 'name "Acme Corporation")))
                             :returning '*)))
              (:insert-into 'employees-log
                            (:select '* (:current-timestamp) :from
                                     'upd))))

sql-op :with-recursive (&rest args)

Recursive modifier to a WITH statement, allowing the query to refer to its own output.

(query (:with-recursive
      (:as (:t1 'n)
           (:union-all (:values (:set 1))
                       (:select (:+ 'n 1)
                                :from 't1
                                :where (:< 'n 100))))
      (:select (:sum 'n) :from 't1)))

(query (:with-recursive
      (:as (:included_parts 'sub-part 'part 'quantity)
           (:union-all
            (:select 'sub-part 'part 'quantity
                     :from 'parts
                     :where (:= 'part "our-product"))
            (:select 'p.sub-part 'p.part 'p.quantity
                     :from (:as 'included-parts 'pr)
                     (:as 'parts 'p)
                     :where (:= 'p.part 'pr.sub-part) )))
      (:select 'sub-part (:as (:sum 'quantity) 'total-quantity)
               :from 'included-parts
               :group-by 'sub-part)))

(query (:with-recursive
      (:as (:search-graph 'id 'link 'data 'depth)
           (:union-all (:select 'g.id 'g.link 'g.data 1
                                :from (:as 'graph 'g))
                       (:select 'g.id 'g.link 'g.data (:+ 'sg.depth 1)
                                :from (:as 'graph 'g) (:as 'search-graph 'sg)
                                :where (:= 'g.id 'sg.link))))
      (:select '* :from 'search-graph)))

(query (:with-recursive
      (:as (:search-graph 'id 'link 'data'depth 'path 'cycle)
           (:union-all
            (:select 'g.id 'g.link 'g.data 1
                     (:[] 'g.f1 'g.f2) nil
                     :from (:as 'graph 'g))
            (:select 'g.id 'g.link 'g.data (:+ 'sg.depth 1)
                     (:|| 'path (:row 'g.f1 'g.f2))
                     (:= (:row 'g.f1 'g.f2)
                         (:any* 'path))
                     :from (:as 'graph 'g)
                     (:as 'search-graph 'sg)
                     :where (:and (:= 'g.id 'sg.link)
                                  (:not 'cycle)))))
      (:select '* :from 'search-graph)))

sql-op :with-ordinality, :with-ordinality-as

Selects can use :with-ordinality or :with-ordinality-as parameters. Postgresql will give the new ordinality column the name of ordinality. :with-ordinality-as allows you to set different names for the columns in the result set.

(query (:select '*
        :from (:generate-series 4 1 -1)
        :with-ordinality))

(query (:select 't1.*
        :from (:json-object-keys "{\"a1\":\"1\",\"a2\":\"2\",\"a3\":\"3\"}")
        :with-ordinality-as (:t1 'keys 'n)

Table Functions

sql-op :for-update (query &key of nowait)

Locks the selected rows against concurrent updates. This will prevent the rows from being modified or deleted by other transactions until the current transaction ends. The :of keyword should be followed by one or more table names. If provided, PostgreSQL will lock these tables instead of the ones detected in the select statement. The :nowait keyword should be provided by itself (with no argument attached to it), after all the :of arguments. If :nowait is provided, PostgreSQL will throw an error if a table cannot be locked immediately, instead of pausing until it's possible.

(query (:for-update (:select :* :from 'foo 'bar 'baz) :of 'bar 'baz :nowait))

sql-op :for-share (query &key of nowait)

Similar to :for-update, except it acquires a shared lock on the table, allowing other transactions to perform :for-share selects on the locked tables.

sql-op :insert-into (table &rest rest)

You can use insert-into when you are:

  1. Inserting from a select clause and you do not need to specify specific columns:
(query (:insert-into 'table1
         (:select 'c1 'c2 :from 'table2)))
  1. Inserting from a select clause and you specifying the columns which will be filled with values from the select clause
(query (:insert-into 't11
        :columns 'region 'subregion 'country
        (:select (:as 'region-name 'region)
                 (:as 'sub-region-name 'subregion)
                 'country
         :from 'regions)))

or

  1. You are alternating specific columns and values for a single row:
(query (:insert-into 'my-table :set 'field-1 42 'field-2 "foobar"))

You can use parameterized variables in the insert statement.

(let ((name "test-cat4"))
  (query (:insert-into 'categories :set 'name '$1) name))

It is possible to add :returning, followed by a list of field names or expressions, at the end of the :insert-into form. This will cause the query to return the values of these expressions as a single row.

(query (:insert-into 'my-table
        :set 'field-1 42 'field-2 "foobar"
        :returning '*))

(query (:insert-into 'my-table
        :set 'field-1 42 'field-2 "foobar"
        :returning 'id))

In Postgresql versions 9.5 and above, it is possible to add :on-conflict-do-nothing (if the item already exists, do nothing). If you want to specify the unique column to be checked for conflict, use :on-conflict 'column-name :do-nothing. If you do not want to specify the unique column name, use :on-conflict-do-nothing.

(query (:insert-into 'test-table :set 'column-A '$1 'column-B '$2
                     :on-conflict 'column-A :do-nothing
                     :where (:= 'test-table.column-A '$1)
                     :returning '*)
        "c" 37)

(query (:insert-into 'test-table :set 'column-A '$1 'column-B '$2
                     :on-conflict-do-nothing 'column-A
                     :where (:= 'test-table.column-A '$1)
                     :returning '*)
        "c" 37)

If your insertion is setting a column that is an identity column with a value normally created by the system and you want to override that, you can use the :overriding-system-value keyword:

(query (:insert-into 'table1
        :set 'c1 "A" 'c2 "B"
        :overriding-system-value))

To create what is commonly known as an upsert, use :on-conflict-update (if the item already exists, update the values) followed by a list of field names which are checked for the conflict then using :update-set followed by a list of field names or expressions following the syntax for updating a table. This is sometimes called an "upsert". Note that as per the postgresql sql documentation you must prepend the table name to the column in the where statement if you are updating.

(query (:insert-into 'test-table :set 'column-A '$1 'column-B '$2
                     :on-conflict-update 'column-A
                     :update-set 'column-B '$2
                     :where (:= 'test-table.column-A '$1)
                     :returning '*)
        "c" 37)

If the destination table has identity columns and you want to override those identity columns with specific values, you should specify :overriding-system-value.

(query (:insert-into 'test-table
        :set 'column-A '$1 'column-B '$2
        :overriding-system-value
        :on-conflict-update 'column-A
        :update-set 'column-B '$2
        :where (:= 'test-table.column-A '$1)
        :returning '*)
        "c" 37)

If you are selecting from another table which has column names the same as your destination table and you want to keep the destination table's identity column values, then you can use :overriding-user-value. E.g.

(query (:insert-into 'table1
         :overriding-user-value
         (:select 'c1 'c2 :from 'table2)))

sql-op :insert-rows-into (table &rest rest)

Insert-rows-into provides the ability to insert multiple rows into a table without using a select statement. (Insert-rows-into keeps the VALUES key word in the resulting sql. If you do use a select statement, Postgresql requires that it only return one row.)

Specify the columns first with the keyword :columns then provide a list of lists of the values as a parameter to the keyword :values. Example:

(query (:insert-rows-into 'my-table
        :columns 'field-1 'field-2
        :values '((42 "foobar") (23 "foobaz"))))

An example using a select statement returning one row:

(squery (:insert-rows-into 't6
         :columns 'tags
         :values '(((:select 'id
                     :from 't5)))))

If you will use the default columns, this can be simplified and the :columns parameters can be dropped. Example:

(query (:insert-rows-into 'my-table
        :values '((42 "foobar") (23 "foobaz"))))

If your insertion is setting a column that is an identity column with a value normally created by the system and you want to override that, you can use the :overriding-system-value keyword:

(query (:insert-rows-into 'table1
                   :columns 'c1 'c2
                   :overriding-system-value
                   :values '((1 "a") (2 "b"))))

(query (:insert-rows-into 'table1
        :overriding-system-value
        :values '(((:select 'c1 'c2 :from 'table2)))))

Similarly to :insert-into, :insert-rows-into allows the "upsert" use of :on-conflict. Again, if you want to specify the unique column to be checked for conflict, use :on-conflict 'column-name :do-nothing. If you do not want to specify the unique column name, use :on-conflict-do-nothing. The following example uses :on-conflict-do-nothing

(query (:insert-rows-into 'distributors
        :columns 'did 'dname
        :values '((10 "Conrad International"))
        :on-conflict-do-nothing
        :where 'is-active))

(query (:insert-rows-into 'distributors
        :columns 'did 'dname
        :values '((10 "Conrad International"))
        :on-conflict 'did
        :do-nothing
        :where 'is-active))

or :on-conflict-update

(query (:insert-rows-into 'distributors
        :columns 'did 'dname
        :values '((5 "Gizmo Transglobal") (6 "Associated Computing Inc."))
        :on-conflict-update 'did
        :update-set 'dname 'excluded.dname))

You can use :on-conflict-on-constraint to check for conflicts on constraints.

(query (:insert-rows-into 'test :columns 'some-key 'some-val
                                :values '(("a" 3) ("b" 6) ("c" 7))
                                :on-conflict-on-constraint 'somekey
                                :do-nothing
                   :returning '*))

(query (:insert-rows-into 'test :columns 'some-key 'some-val
                                :values '(("a" 2) ("b" 6) ("c" 7))
                                :on-conflict-on-constraint 'somekey
                                :update-set 'some-val 'excluded.some-val
                   :returning '*))

sql-op :update (table &rest rest)

Update values in a table. There are two ways to update the values

The first method uses the keyword :set and any number of alternating field names and values, like for :insert-into. Next comes the optional keyword :from, followed by at least one table name and then any number of join statements, like for :select. After the joins, an optional :where keyword followed by the condition, and :returning keyword followed by a list of field names or expressions indicating values to be returned as query result.

(query (:update 'weather
        :set 'temp-lo (:+ 'temp-lo 1)
             'temp-hi (:+ 'temp-lo 15)
             'prcp :default
        :where (:and (:= 'city "San Francisco")
                     (:= 'date "2003-07-03"))
        :returning 'temp-lo 'temp-hi 'prcp))

The second method uses the :columns keyword to specify which columns get created and allows the use of either :set or :select (both of which need to be enclosed in a form) to provide the values, allowing update queries like:

(query (:update 'weather
        :columns 'temp-lo 'temp-hi 'prcp
                 (:set (:+ 'temp-lo 1)  (:+ 'temp-lo 15) :DEFAULT)
        :where (:and (:= 'city "San Francisco")
                     (:= 'date "2003-07-03"))))

(query (:update 't1
        :columns 'database-name 'encoding
                 (:select 'x.datname 'x.encoding
                  :from (:as 'pg-database 'x)
                  :where (:= 'x.oid 't1.oid))))

sql-op :delete-from (table &rest rest)

Delete rows from the named table. Can be given a :where argument followed by a condition, and a :returning argument, followed by one or more expressions that should be returned for every deleted row.

(query (:delete-from 'cd.bookings :where (:= 'id 5)))

The following more complicated version shows the application of the :using and :where clauses:

(query (:delete-from 'members
        :using 'producers
        :where (:and (:= 'members.id 'producers.id)
                     (:= 'members.name "Steve"))))

(sql (:delete-from 'members
      :using 'producers 'films
                   :where (:and (:= 'members.id 'producers.id)
                                (:= 'members.name "Steve")
                                (:= 'producers.films-id 'films.id)))

sql-op :create-table (name (&rest columns) &rest options)

Create a new table. The simplest example would pass two parameters, the table name and a list of lists providing information for each column. For example:

(query (:create-table 'george ((id :type integer))))

where 'george is the name of the table, it has 1 column named id which is limited to integers. There are no indexes or keys in this example.

See create-tables.html for more detailed examples.

Column Definition parameters

After the table name a list of column definitions follows, which are lists that start with a name, followed by one or more of the following keyword arguments:

  • :type

This one is required. It specifies the type of the column. Use a type like (or db-null integer) to specify a column that may have NULL values.

  • :default

Provides a default value for the field.

  • :unique

If this argument is non-nil, the values of the column must be unique.

  • :primary-key

When non-nil, the column is a primary key of the table.

  • :check

Adds a constraint to this column. The value provided for this argument must be an S-SQL expression that returns a boolean value. It can refer to other columns in the table if needed.

  • :references

Adds a foreign key constraint to this table. The argument provided must be a list of the form (target &optional on-delete on-update). When target is a symbol, it names the table to whose primary key this constraint refers. When it is a list, its first element is the table, and its second element the column within that table that the key refers to. on-delete and on-update can be used to specify the actions that must be taken when the row that this key refers to is deleted or changed. Allowed values are :restrict, :set-null,

  • :set-default, :cascade, and :no-action.

Table Constraints

After the list of columns, zero or more extra options (table constraints) can be specified. These are lists starting with one of the following keywords:

  • :check

Adds a constraint to the table. Takes a single S-SQL expression that produces a boolean as its argument.

  • :primary-key

Specifies a primary key for the table. The arguments to this option are the names of the columns that this key consists of.

  • :unique

Adds a unique constraint to a group of columns. Again, the arguments are a list of symbols that indicate the relevant columns.

  • :foreign-key

Create a foreign key. The arguments should have the form (columns target &optional on-delete on-update), where columns is a list of columns that are used by this key, while the rest of the arguments have the same meaning as they have in the :references option for columns. Every list can start with :constraint name to create a specifically named constraint.

Note that, unlike most other operators, :create-table expects most of its arguments to be unquoted symbols. The exception to this is the value of :check constraints: These must be normal S-SQL expressions, which means that any column names they contain should be quoted. When programmatically generating table definitions, sql-compile is usually more practical than the sql macro.

Here is an example of a :create-table form:

(:create-table enemy
  ((name :type string :primary-key t)
   (age :type integer)
   (address :type (or db-null string) :references (important-addresses :cascade :cascade))
   (fatal-weakness :type text :default "None")
   (identifying-color :type (string 20) :unique t))
  (:foreign-key (identifying-color) (colors name))
  (:constraint enemy-age-check :check (:> 'age 12)))

For more detail and examples on building tables using the s-sql approach, see create-tables.html

sql-op :alter-table (name action &rest args)

Alters named table. Currently changing a column's data type is not supported. The meaning of args depends on action:

  • :add-column

Adds column to table. args should be a column in the same form as for :create-table.

(query (:alter-table "packages" :add-column 'system-data-p :type (or boolean db-null)))
  • :set-default

Adds or changes a default value for a column

(query (:alter-table 'countries :alter-column 'updated-at :set-default  (:now)))
  • :drop-column

Drops a column from the table.

(query (:alter-table "test-uniq" :drop-column 'address))
  • :add-constraint

Adds a named constraint to the table.

(query (:alter-table "test-uniq" :add-constraint silly-key :primary-key 'code 'title))

(query (:alter-table enemy :add-constraint enemy-age-check :check (:> 'age 21)))
  • :drop-constraint

Drops constraint. First of args should name a constraint to be dropped; second, optional argument specifies behaviour regarding objects dependent on the constraint and it may equal :cascade or :restrict.

(query (alter-table enemy :drop-constraint enemy-age-check))
  • :add

Adds an unnamed constraint to table. args should be a constraint in the same form as for :create-table. (This is for backwards-compatibility, you should use named constraints.)

  • :rename

Adds the ability to rename a table.

  • :rename-column

Adds the ability to rename a column of a table.

(query (:alter-table "test-uniq" :rename-column 'address 'city))

sql-op :drop-table (name)

Drops the named table. You may optionally pass :if-exists before the name to suppress the error message if the table does not exist. You can also optionally pass :cascade after the name to indicate that it should also drop any other tables, indices, etc which depend on that table. Accepts strings, variable or symbol as the identifier.

(query (:drop-table 'table1))
(query (:drop-table :if-exists 'table1))
(query (:drop-table :if-exists 'table1 :cascade))
(query (:drop-table (:if-exists 'table1-with-longer-name) :cascade))
(let ((table-var1 "table1"))
    (is (equal (sql (:drop-table :if-exists table-var1 :cascade))
               "DROP TABLE IF EXISTS table1 CASCADE"))
(let ((table-var1 'table-1))
    (is (equal (sql (:drop-table :if-exists table-var1 :cascade))
               "DROP TABLE IF EXISTS table_1 CASCADE"))

sql-op :truncate (&rest args)

Truncates one or more tables, deleting all the rows. Optional keyword arguments are allowed in the following order. Note that :continue-identity and :restart-identity make no sense if both are included.

  • :only (if not specified, the table and its descendants are truncated).
  • :continue-identity (the values of sequences will not be changed. This is the default)
  • :restart-identity (the values of sequences owned by the table(s) will be restarted)
  • :cascade (will cascade the truncation through tables using foreign keys.)

Example calls would be:

(query (:truncate 'bigtable 'fattable))

(query (:truncate 'bigtable 'fattable :only))

(query (:truncate 'bigtable 'fattable :only :continue-identity))

(query (:truncate 'bigtable 'fattable :restart-identity))

(query (:truncate 'bigtable 'fattable :only :restart-identity :cascade ))

sql-op :create-index (name &rest args)

Create an index on a table. After the name of the index the keyword :on should follow, with the table name after it. Then the keyword :fields, followed by one or more column names. Other optional parameters are :using (to use index methods other than b-tree) :with (for storage parameters) or :where (for a condition can be added at the end to make a partial index). See S-SQL-C for more examples.

(query (:create-index 'films_idx :on 'films :fields 'title))

(query (:create-index 'films-idx :on "films" :fields 'title 'id))

(query (:create-index 'gin-idx :on "historical-events" :using 'gin :fields 'data))

Expressions can be used in the :fields specification:

(query (:create-index 'films_idx :on "films" :fields (:lower 'title)))

Opclasses can be used within the fields specification:

(query (:create-index 'pointloc2 :on 'points :using 'gist
        :fields (:asc (:box 'location 'location)) (:nulls-last 'name)))

Storage parameters can be set using a :with clause:

(query (:create-index 'gin-idx :on 'documents-table :using gin
        :fields 'locations :with (:= 'deduplicate-items off) (:= 'fillfactor 70)))

Partial indexes can be set using a :where clause:

(query (:create-index 'orders_unbilled_idx :on 'orders :fields 'order-nr
        :where (:and (:is-not-true 'billed)
                            (:< 'order_nr 1000)))))

Indexing a jsonb column named metadata:

(query (:create-index 'items-idx :on 'test-items :using 'gin
        :fields (:jsonb-path-ops 'metadata)))

sql-op :create-unique-index (name &rest args)

Works like :create-index, except that the index created is unique.

sql-op :drop-index (name)

Drop an index. Takes :if-exists and/or :cascade arguments like :drop-table. Accepts strings, variable or symbol as the identifier.

(query (:drop-index 'index1))
(query (:drop-index :if-exists 'index1))
(query (:drop-index :if-exists 'index1 :cascade))
(let ((table-var1 "table1"))
    (is (equal (sql (:drop-index :if-exists table-var1 :cascade))
               "DROP INDEX IF EXISTS table1 CASCADE"))

sql-op :create-sequence (name &key increment min-value max-value start cache cycle)

Create a sequence with the given name. The rest of the arguments control the way the sequence selects values.

sql-op :alter-sequence (name)

Alters a sequence. See Postgresql documentation for parameters.

  • :increment

Sets the amount by which each subsequent increment will be increased.

  • :min-value
  • :max-value
  • :no-min
  • :no-max
  • :start
  • :restart
  • :cache
  • :cycle
  • :no-cycle
  • :owned-by
  • :if-exists before the name to suppress the error message.

sql-op :drop-sequence (name)

Drop a sequence. Takes :if-exists and/or :cascade arguments like :drop-table. Accepts strings, variable or symbol as the identifier.

(query (:drop-sequence 'sequence1))
(query (:drop-sequence :if-exists 'sequence1))
(query (:drop-sequence :if-exists 'sequence1 :cascade))

sql-op :create-view (name query)

Create a view from an S-SQL-style query.

sql-op :drop-view (name)

Drop a view. Takes optional :if-exists argument. Accepts strings, variable or symbol as the identifier.

sql-op :set-constraints (state &rest constraints)

Configure whether deferrable constraints should be checked when a statement is executed, or when the transaction containing that statement is completed. The provided state must be either :immediate, indicating the former, or :deferred, indicating the latter. The constraints must be either the names of the constraints to be configured, or unspecified, indicating that all deferrable constraints should be thus configured.

sql-op :listen (channel)

Tell the server to listen for notification events on channel channel, a string, on the current connection.

sql-op :unlisten (channel)

Stop listening for events on channel.

sql-op :notify (channel &optional payload)

Signal a notification event on channel channel, a string. The optional payload string can be used to send additional event information to the listeners.

sql-op :create-role (role &rest args)

Create a new role (user). Following the role name are optional keywords arguments:

  • :options

One or more of the no-parameter options to PostgreSQL's CREATE ROLE SQL command.

  • :password

Sets the role's password. (A password is only of use for roles having the LOGIN attribute, but you can nonetheless define one for roles without it.) If you do not plan to use password authentication you can omit this option. If no password is specified, the password will be set to null and password authentication will always fail for that user.

  • :connection-limit

If role can log in, this specifies how many concurrent connections the role can make. -1 (the default) means no limit.

  • :valid-until

The :valid-until clause sets a date and time after which the role's password is no longer valid. If this clause is omitted the password will be valid for all time.

  • :role

Lists one or more existing roles which are automatically added as members of the new role. (This in effect makes the new role a “group”.)

  • :in-role

Lists one or more existing roles to which the new role will be immediately added as a new member.

Here is an example of a :create-role form:

(query (:create-role 'user23
                     :options 'SUPERUSER 'NOINHERIT 'LOGIN
                     :password "mypassword"
                     :connection-limit 100 :role 'users))

sql-op :create-database (name)

Create a new database with the given name.

sql-op :drop-database (name)

Drops the named database. You may optionally pass :if-exists before the name to suppress the error message. Examples:

(query (:drop-database 'database-name))
(query (:drop-database :if-exists 'database-name))
(let ((var 'my-database)) (query (:drop-database var)))

sql-op :copy (table &rest args)

Move data between Postgres tables and filesystem files. Table name is required followed by one or more of the following keyword arguments. Documentation for the copy command provides a full reference. An example from the Greenplum tutorial:

(query
   (:copy 'faa.d_airlines
    :columns 'airlineid 'airline_desc
    :from "/home/gpadmin/gpdb-sandbox-tutorials/faa/L_AIRLINE_ID.csv"
    :on-segment t
    :binary t
    :oids t
    :header t
    :delimiter ","
    :null "NULL"
    :escape "my-escape-string"
    :newline "CR"
    :csv t
    :log-errors t
    :segment-reject-limit 100 'ROWS))

Analyze

The official Postgresql documents note that analyze collects statistics about a database and provides various options, whether you want specific tables and columns or the entire database, etc. The following is a series of examples of how it can be used in s-sql with increasing levels of complexity. Options can be set to true with t, 1 or :on and set to false with :nil 0 or off.

(query (:analyze))

(query (:analyze :verbose))
;; Now specifying just analyzing table t1
(query (:analyze :verbose :tables 't1))

(query (:analyze :verbose :tables :t1))
;; Now specifying just analyzing tables t1 and t2
(query (:analyze :verbose :tables :t1 :t2))

(query (:analyze :verbose :tables 't1 't2))
;; Now specifying columns in tables t1 and t2
(query (:analyze :verbose
        :tables (:t1 't1c1 't1c2) (:t2 't2c1 't2c2 't2c3)))
;; Alternative syntax for the columns
(query (:analyze :verbose :tables (:t1 :t1c1 :t1c2)))
;; Starting to look at more specific options
(query (:analyze :option (:verbose t)
                 :tables (:t1 't1c1 't1c2) (:t2 't2c1 't2c)))
;; The following will set option verbose to true, skip-locked to false
;; and buffer-usage-limit to 70MB
(query (:analyze :option (:verbose 1) (:skip-locked 0) (:buffer-usage-limit "70MB")
                 :tables (:t1 't1c1 't1c2) (:t2 't2c1 't2c)))
;; The following will set option verbose to true, skip-locked to false
;; and buffer-usage-limit to 70MB
(query (:analyze :option (:verbose t) (:skip-locked nil) (:buffer-usage-limit "70MB")
                 :tables (:t1 't1c1 't1c2) (:t2 't2c1 't2c)))

Dynamic Queries, Composition and Parameterized Queries

Overview

The question gets asked how to build dynamic or composable queries in postmodern. First we need to understand the context - is the programmer building the query or are you taking data from a user and using that to build a query?

Programmer Built Queries

The question gets asked how to build dynamic or composable queries in postmodern. First we need to understand the context - is the programmer building the query or are you taking data from a user and using that to build a query? We need to remember that the query macro assumes that everything that is not a list starting with a keyword will evaluate to a string.

In any case you will need to ensure that either you have control over the inputs or they still result in parameterized queries. If not you have opened yourself up to an sql injection attack.

IMPORTANT REGARDING PARAMETERIZED QUERIES: For Postmodern versions before 1.33.7, you cannot use a list in a parameterized statement. You have to convert the list to a vector and use :any* rather than :in. See S-SQL for more details.

If you are not using s-sql, then it becomes easy. The query macro assumes that everything that is not a list starting with a keyword will evaluate to a string. That means you can build it with a simple format string

(query (format nil "select ~a from ~a where ~a"  "carrots" "garden" "length > 3"))

With s-sql, there are generally three approaches to building dynamic or composible queries: pass symbols and values as variables, use sql-compile or use :raw.

For purposes of this example, we will use the following employee table:

(query (:create-table employee ((id :type int)
                                (name :type text)
                                (salary :type numeric)
                                (start_date :type date)
                                (city :type text)
                                (region :type char)
                                (age :type int))))

(query (:insert-rows-into 'employee
                          :columns 'id 'name 'salary 'start-date 'city 'region 'age
                          :values '((1 "Jason" 40420 "02/01/94" "New York" "W" 29)
                                    (2 "Robert" 14420 "01/02/95" "Vancouver" "N" 21)
                                    (3 "Celia" 24020 "12/03/96" "Toronto" "W" 24)
                                    (4 "Linda" 40620 "11/04/97" "New York" "N" 28)
                                    (5 "David" 80026 "10/05/98" "Vancouver" "W" 31)
                                    (6 "James" 70060 "09/06/99" "Toronto" "N" 26)
                                    (7 "Alison" 90620 "08/07/00" "New York" "W" 38)
                                    (8 "Chris" 26020 "07/08/01" "Vancouver" "N" 22)
                                    (9 "Mary" 60020 "06/08/02" "Toronto" "W" 34))))
  • Approach #1 Using symbols in variables
    • Select Statements

      Consider the following two toy examples where we determine the table and columns to be selected using symbols (either keyword or quoted) inside variables.

      (let ((table 'employee) (col1 :id) (col2 :name) (id 3))
          (query (:select col1 col2 :from table :where (:= 'id '$1)) id))
      ((3 "Celia"))
      
      (let ((table 'employee) (col1 'name) (col2 'salary) (id 3))
          (query (:select col1 col2 :from table :where (:= 'id '$1)) id))
      (("Celia" 24020))
      

      This will not work if you use strings instead of symbols because sql-expand will wrap the strings in the variables in escape format as if they were string constants and Postgresql will throw an error because it is not expecting string constants in the middle of a select statement.

    • Update Statements

      This works with update statements as well

      (let ((table 'employee) (col1 :id) (col2 :name) (new-name "Celeste") (id 3))
          (query (:update table :set col2 new-name :where (:= col1 '$1)) id)
          (query (:select col1 col2 :from table :where (:= 'id '$1)) id))
      ((3 "Celeste"))
      
    • Insert Statements

      This works with insert-into statements as well

      (let ((table 'employee) (col1 'id) (col2 'name) (new-name "Rochelle")
            (id 10) (col3 'salary) (col3-value 3452) (col4 'start-date)
            (col4-value "02/01/03") (col5 'city) (col5-value "Victoria")
            (col6 'region) (col6-value "N") (col7 'age) (col7-value 32))
        (query (:insert-into table :set col1 id col2 new-name col3 col3-value
                             col4 col4-value col5 col5-value col6 col6-value
                             col7 col7-value)))
      
      (query (:select 'id 'name 'salary :from 'employee :where (:= 'id 10 )))
      ((10 "Rochelle" 3452))
      
  • Delete Statements

    This works with delete statements as well

    (let ((table 'employee) (col1 :id) (col1-value 10))
       (query (:delete-from table :where (:= col1 col1-value))))
    

Approach #2 Use sql-compile

Sql-compile does a run-time compilation of an s-sql expression. In the following example, we create a function that accepts a where-clause, a table-name, 3 columns to select and two parameters to go into the where clause.

(defun toy-example (where-clause table-name col1 col2 col3 arg1 arg2)
       (with-test-connection
       (query (sql-compile
                (append `(:select ,col1 ,col2 ,col3 :from ,table-name :where)
                       where-clause))
              arg1 arg2)))

(toy-example '((:and (:= 'city '$1) (:> 'salary '$2))) 'employee 'id 'name 'city "Toronto" 45000)

((6 "James" "Toronto") (9 "Mary" "Toronto"))

If we just look at what this call to sql-compile in toy-example generates, it would look like:

"(SELECT id, name, city FROM employee WHERE ((city = $1) and (salary > $2)))"

This example is still a parameterized query but for security reasons you will need to be very careful how you generate the where clause.

Another example with sql-compile and append, in this case updating a table and setting two columns to NULL.

(sql-compile (append '(:update :table1 :set)
                      (loop for a in '("col1" "col2")
                            collect a
                            collect :NULL)))

 "UPDATE table1 SET E'col1' = NULL, E'col2' = NULL"

Lets think about it differently. What if we know the universe of columns we want to select, but want to conditionally select some of them. Suppose we know our targetted table has columns:

'id 'name 'salary 'start-date 'city 'region 'age.

We may decide we always want name, city and age, but salary and start-date are conditional.

(defun toy-example-2 (salaryp start-date-p)
  (sql-compile
   (remove nil `(:select 'name 'city 'age
                         ,(if salaryp 'salary nil)
                         ,(if start-date-p 'start-date nil)
                         :from 'employee))))

(query (toy-example-2 t t))

(("Jason" "New York" 29 40420 #<SIMPLE-DATE:DATE 01-02-1994>)
 ("Robert" "Vancouver" 21 14420 #<SIMPLE-DATE:DATE 02-01-1995>)
 ("Celia" "Toronto" 24 24020 #<SIMPLE-DATE:DATE 03-12-1996>)
 ("Linda" "New York" 28 40620 #<SIMPLE-DATE:DATE 04-11-1997>)
 ("David" "Vancouver" 31 80026 #<SIMPLE-DATE:DATE 05-10-1998>)
 ("James" "Toronto" 26 70060 #<SIMPLE-DATE:DATE 06-09-1999>)
 ("Alison" "New York" 38 90620 #<SIMPLE-DATE:DATE 07-08-2000>)
 ("Chris" "Vancouver" 22 26020 #<SIMPLE-DATE:DATE 08-07-2001>)
 ("Mary" "Toronto" 34 60020 #<SIMPLE-DATE:DATE 08-06-2002>))

(query (toy-example-2 t nil))

(("Jason" "New York" 29 40420) ("Robert" "Vancouver" 21 14420)
 ("Celia" "Toronto" 24 24020) ("Linda" "New York" 28 40620)
 ("David" "Vancouver" 31 80026) ("James" "Toronto" 26 70060)
 ("Alison" "New York" 38 90620) ("Chris" "Vancouver" 22 26020)
 ("Mary" "Toronto" 34 60020))

You could skip the (remove nil… portion and substitute t for nil. E.g.

(defun toy-example-2 (salaryp start-date-p)
        (sql-compile
         `(:select 'name 'city 'age
                   ,(if salaryp 'salary t)
                   ,(if start-date-p 'start-date t)
                   :from 'employee)))

But I prefer to remove those segments completely from the query.

Following on this same thread of thought, you can define a portion of the sql in a let clause:

(let ((sql1 '(:= name "Jason")))
        (query (sql-compile
          `(:select 'name 'city 'age :from 'employee :where ,sql1))))

      (("Jason" "New York" 29))

An example of this would be getting more columns depending on the postgresql server versionr:

(defun more-table-info (table-name)
        "Returns variable amounts of information depending on the postgresql server version"
        (let* ((version>11 (postgresql-version-at-least "12.0" =*database*=))
               (version>10 (postgresql-version-at-least "11.0" =*database*=))
               (select-query (sql-compile
                              `(:order-by
                                (:select (:as 'a.attnum 'ordinal-position)
                                         (:as 'a.attname 'column-name)
                                         (:as 'tn.typname 'data-type)
                                         ,(if version>10 'a.attidentity t)
                                         ,(if version>11 'a.attgenerated t)
                                         :from (:as 'pg_class 'c)
                                         (:as 'pg_attribute 'a)
                                         (:as 'pg_type 'tn)
                                         :where (:and
                                                 (:= 'c.relname '$1)
                                                 (:> 'a.attnum 0)
                                                 (:= 'a.attrelid 'c.oid)
                                                 (:= 'a.atttypid 'tn.oid)))
                                'a.attnum))))
          (query select-query
                 (to-sql-name table-name))))

Approach #3 Use :raw

To quote Marijn, the :raw keyword takes a string and inserts it straight into the query. I try to stay away from :raw if possible, but sometimes…

(query (:select (:raw "tmp1.name") :from (:as 'baz (:raw "tmp1"))))

Queries with User Input

In any of the above approaches to building queries you will need to ensure that either you have control over the inputs or they still result in parameterized queries. If not, you have opened yourself up to an sql injection attack.