Over a million developers have joined DZone.
Refcard #071

Essential PostgreSQL

Become a PostgreSQL Power User

Written by

Leo Hsu leader, icbc
Regina Obe President, Paragon Corporation

Navigates PostgreSQL’s most popular features and hidden gems and covers topics such as configuration, administration, language support, and more.

Free PDF
Section 1

About PostgreSQL

PostgreSQL is an open-source object-relational database with many enterprise-level features. It runs on numerous platforms: Linux, Unix, Windows, and Mac OS X. It is simple and quick to install, fast, and it sports advanced features such as: streaming replication, spatial support via PostGIS, windowing functions, table partitioning, and full-text search. In addition to its enterprise features, it has the added benefit of supporting numerous languages for authoring stored functions. It has an extensible procedural language architecture to introduce new languages. It also has an extensible type and index architecture for introducing new data types, operators, and indexes for these custom types, and support for querying external data sources such as CSV, web services, and other PostgreSQL services via its Foreign Data Wrapper (SQL/MED) support.

Targeted at novices and professionals alike, this Refcard will help you quickly navigate some of PostgreSQL’s most popular features as well as its hidden gems. It will cover topics such as configuration, administration, backup, language support, and advanced SQL features. There will be a special focus on new features in PostgreSQL 9.3 and 9.4.

Section 2


PostgreSQL uses three main configuration files to control overall operations. You can find these files in the initialized data cluster (the folder specified during the initialization process using initdb -d).

All these can be edited with a text editor. They can be edited via PgAmin III if you install the adminpack extension in master postgres db. To do so:CREATE EXTENSION adminpack;

File Purpose
postgresql.conf Controls the listening port, IP, and default query planner settings, memory settings, path settings, and logging settings. Can be queried via pg_settings database view.
pg_hba.conf Controls the authentication models used by PostgreSQL and can be set per user, per database, per IP range, or a combination of all.
pg_indent.conf Controls mapping of an OS user to a PostgreSQL user.


The following settings are all located in the postgresql.conf file. Remember that these are default settings; many of these you can choose to override for each session, for each database, or for each user/role.

Option Description
listen_addresses Use ‘*’ to listen on all IPs of the server, ‘localhost’ to listen on just local, or a comma separated list of IPs to listen on. Requires service restart if changed and can only be set globally.
port Defaults to 5432, but can be changed to allow multiple postgresql daemon clusters/versions to coexist using same IP but different ports.
search_path List of default schemas that don’t need schema qualification. First schema is where non-schema qualified objects are created.
constraint_exclusion Options: on, off, or partial. Partial was introduced in 8.4 and is the new default. Allows planner to skip over tables if constraint ensures query conditions cannot be satisfied by the table. Mostly used for table partitioning via table inheritance.
shared_buffers Controls how much memory is allocated to PostgreSQL and shared across all processes. Requires service restart and can only be set globally.

In PostgreSQL 9.4, a new SQL construction ALTER SYSTEM was introduced that allows you to set these settings at the system level without editing the postgresql.conf. For many, you still need to do a service restart and for others at least a:

SELECT pg_reload_conf();


PostgreSQL supports many authentication schemes to control access to the database. The pg_hba.conf file dictates which schemes are used based on the rules found in this file. You can mix and match various authentication schemes at the same time. The rules are applied sequentially such that the first match fitting a connection is the one that is used. This is important to remember because if you have a more restrictive rule above a less restrictive, then the more restrictive is the one that trumps.

The most commonly used authentication schemes are trust (which allows connections without a password) and md5 (which authenticates with md5 encrypted passwords). Others include: reject, crypt, password (this is plain text), krb5, ident (authenticate simply by identity of user in OS), pam, and ldap.

The example pg_hba.conf entries below allow all local connections to connect to all databases without a password and all remote connections to authenticate via md5.


host all all trust

host all all md5

Section 3

Data Types

PostgreSQL has numerous built-in types. In addition, you can define custom types. Furthermore, all tables are considered to be types in their own right, and can therefore be used within another table’s column. Below are the common built-in types:

Date/Time Types

Type Description
date The date is a datatype to represent dates with no time. Default representation is ISO 8601 e.g. ‘YYYY-MM-DD’. Use datestyle configuration setting to control defaults.
timestamp This includes both date and time and is timezone-neutral. Example: '2009-07-01 23:00'
timestamp with time zone Timestamp with timezone. Example: '2009-07-01 23:00:00-04'
time Time without date. Example: '23:14:20'
time with time zone Example: '23:14:20-04'
interval A unit of time used to add and subtract from a timestamp. Example: SELECT TIMESTAMP '2009-07-01 23:14:20' + INTERVAL '4 months 2 days 10 hours 9 seconds'
daterange, tsrange, tstzrange New in PostgreSQL 9.2; defines a specific time range. Example is a date > 2012-07-01 and <= 2013-08-31
SELECT '(2012-07-01, 2013-08-31]'::daterange;
Constituents of datetime, use date_part function to extract century, day, decade, dow (starts Sunday), doy, epoch, hour, isodow (day of week starts on Monday), minute, month, quarter, week, year.

Numeric Types

Type Description
int, int8 4 byte and 8 byte integers.
serial, serial4, serial8 Sequential integers; this can be used during table creation to specify auto-numbered fields.
numeric(s, p) Decimal numbers; s is scale and p is precision.
double precision Floating point numbers.
numrange, int4range Introduced in 9.2 for defining number ranges. An integer >= 1 and < 10
SELECT '[1,10)'::int4range;
percentile_cont, percentile_disc Continuous and discrete percentile aggregate. Can take a numeric value (between 0 and 1) for percentile rank or can take an array of numeric values between 0 and 1.

String Types

Type Description
varchar(n) (a.k.a. character varying) Max of n characters, no trailing spaces.
char(n) Padded to n characters.
text Unlimited text.

Other Types

array Arrays in PostgreSQL are typed, and you can create an array of any type. To define a column as an array of a specific type, follow with brackets. Example: varchar(30)[]. You can also autogenerate arrays in SQL statements with constructs such as:
SELECT ARRAY['john','jane'];
SELECT ARRAY(SELECT emp_name FROM employees); 
SELECT array_agg(e.emp_name) FROM employees; 
enum Enumerators:
CREATE TYPE cloth_colors AS ENUM ('red','blue','green');
When used in a table, you define the column as the name of the enum. Sorting is always in the order the items appear in the enum definition.
boolean True/false.
bytea Byte array used for storing binary objects, such as files.
lo Large object. Stored in a separate system table with an object ID reference to the large object. Useful for importing files from a file system and storing and exporting back to the file system.
JSON JavaScript Object Notation (JSON) was introduced in PostgreSQL 9.2 and includes built-in validation. JSON stored as plain text. No direct index support. PostgreSQL 9.3 enhanced JSON functionality by providing more functions and operators that work with JSON. PostgreSQL 9.4 enhanced further by providing even more functions and operators.
jsonb Binary form of JSON—introduced in PostgreSQL 9.4. Can be indexed using GIN indexes and supports intersects and containment operators in addition to all the functions and operators JSON supports. Performance is much faster than the JSON type. No duplicate keys per object are allowed; sort of keys per object are not maintained.

Common Global Variables

Type Description
CURRENT_TIMESTAMP, now() Returns current date and time with timezone.
CURRENT_DATE Returns current date without the time.
CURRENT_TIME Returns current time without the date.
Section 4

Commonly Used Functions

Date/Time Functions and Operators

Type Description
age(timestamp, timestamp) Returns an interval spanned between timestamp1 and timestamp2.
age(timestamp) Difference from current time.
date_part(text, timestamp), date_part(text, interval) date_part('day', timestamp '2009-07-04 11:05:45') => 4 date_part('hour', interval '560 minutes') => 9
date_trunc(text, timestamp | timestamptz | date) date_trunc('hour', '2014-01-15 10:30 PM'::timestamp) => 2014-01-15 22:00:00
operators +, -, / (for intervals only) You can add (or subtract) intervals to datetimes. You can perform addition and subtraction between two datetimes. You can divide intervals into smaller intervals.
generate_series(timestamp, timestamp, [interval]) [8.4] Generate rows of timestamps.

String Functions and Operators

Type Description
|| (string || string, string || number) String concatenation.
left, right, substring Returns left x elements, right x elements, or substring from position x for y number of elements.
length Number of characters in string.
lpad, rpad Left and right pad. lpad('A', 5, 'X') => ‘XXXXA’ rpad('A', 5, 'X') => ‘AXXXX’
lower, upper, initcap Lower, upper, proper case.
md5 Calculates the MD5 hash.
quote_ident Quotes keywords and expressions not suitable for identity when unquoted. quote_ident('in') => “in” quote_ident('big') => big
quote_literal Escapes both single and double quotes.
quote_nullable Similar to quote_literal but doesn’t quote NULL.
replace replace('1234abcv', '1234', 'joe ') => joeabcv
split_part Takes a delimited string and returns the nth item. split_part('abc|def', '|', 2) =>def
string_agg SQL aggregate function that aggregates a set of values into a string.
strpos(text, subtext) Returns numeric position of subtext within text.
trim, btrim, ltrim, rtrim Trim spaces in string.

Array Functions and Operators

Type Description
|| Array concatenation. ARRAY[1,2,3] || ARRAY[3,4,5] => {1,2,3,3,4,5}
unnest Converts an array to rows. SELECT anum FROM unnest(ARRAY[1,2,3])
array_agg SQL aggregate function that aggregates a set of values into an array.
array_upper(anyarray, dimension) array_lower(anyarray, dimension) Returns upper/lower bound of the requested array dimension. SELECT array_upper(ARRAY[ARRAY['a'], ARRAY['b']],1); outputs: 2
array_to_string(anyarray, delimiter_text) Converts an array to a text delimited by the delimiter. array_to_string(ARRAY[12,34], '|') => ‘12|34’

Range Functions and Operators

Type Description
lower(anyrange), upper(anyrange) Lower bound and upper bound value of a range:
SELECT lower(a), upper(a)
FROM (SELECT '[1,10]'::int4range AS a) AS f; 
outputs: lower | upper ------+------- 1| 11
@> Contains range or element.
SELECT a @> 1 AS ce,
a @> '[2,3]'::int4range AS cr
FROM (SELECT '[1,10]'::int4range AS   a) AS f;
&& Have elements in common.
+ Union of 2 ranges.
SELECT '[2014-7-20, 2014-10-20]'::daterange + '[2014-6-20, 2014-7-22]'::daterange;
Output: [2014-06-20,2014-10-21)
* Intersection.
SELECT '[2014-7-20, 2014-10-20]'::daterange * '[2014-6-20, 2014-7-22]'::daterange;
Output: [2014-07-20,2014-07-23)
- Difference.
SELECT '[2014-7-20, 2014-10-20]'::daterange - '[2014-6-20,   2014-7-22]'::daterange;
Output: [2014-07-20,2014-10-21)

JSON/JSONB Functions and Operators

Type Description
->> Extract an element of JSON/jsonb as text.
SELECT prod->>'price' AS price
SELECT '{"id":   1,"name": "milk", "price": 2.50}'::json AS prod) AS f;
Outputs: 2.50
-> Extract an element of JSON/jsonb as JSON/jsonb (useful for doing more operations on a complex subelement).
#>> Extract a nested element of JSON/jsonb as text.
SELECT prod#>>'{nutrition,vitamin d}'::text[] AS   vd
SELECT '{"id": 1,"name":   "milk", "price": 2.50, "nutrition": {"vitamin   d": "30%"}}'::json AS prod) AS f;
Outputs: 30%
#> Extract a nested element of JSON/jsonb as JSON/jsonb. Useful for doing more operations such as working with arrays within json.

Window Functions

Type Description
row_number Number of current row from its current partition.
rank, percent_rank, dense_rank Ranking based on order in current partition (dense_rank is without gaps; percent_rank is relative rank).
lead, lag Nth value relative to current, -nth value relative to current (n defaults to 1) in current partition.
first_value, last_value, nth_value Absolute first/last/nth value in a partition based on order regardless of current position.

Other Functions

Type Description
generate_series(int1, int2, [step]) generate_series(timestamp1, timestamp2, [interval]) Returns rows consisting of numbers from int1 to int2 with [step] as gaps. Step is optional and defaults to 1.
min, max, sum, avg, count Common aggregates.
percentile_dist, percentile_cont [9.4] Useful for computing medians.
Section 5

Database Objects

Here is a listing of what you will find in a PostgreSQL server or database. An * means the object lives at the server level, not the database level.

Object Description
Databases* PostgreSQL supports more than one database per service/daemon.
Tablespaces* Logical representation of physical locations where tables are stored. You can store different tables in different tablespaces, and control data storage based on database and user/group role.
Languages These are the procedural languages installed in the database.
Casts PostgreSQL has the unique feature of having an extensible cast system. It has built-in casts, but allows you to define your own and override default casts. Casts allow you to define explicit behavior when casting from one object to another, and allow you to define autocast behavior.
Schemas These are logical groupings of objects. One can think of them as mini-databases within a larger database. An object always resides in a schema.
Tables, Views Views are virtual tables that encapsulate an SQL SELECT statement. In PostgreSQL, tables can inherit from other tables and data can be altered against views. PostgreSQL 9.1+ introduced Foreign Tables, which are references to data from a Foreign source via a foreign data wrapper (FDW). PostgreSQL 9.3 introduced materialized views, which are views that contain the cached data. These need to be refreshed to update the view cache.
Rules Rules are similar to triggers, except they can only be written in SQL, and they rewrite a statement rather than actually updating directly. Views are actually implemented as SELECT rules (and can have DO INSTEAD inserts/update rules to make them updateable).
Functions, triggers, and aggregates These can be written in any enabled language in the database, live in schemas. PostgreSQL allows you to define your own custom aggregate functions. Triggers are special classes of functions that have OLD and NEW variables available that hold a pointer to the OLD and NEW data. Triggers are bound to table. New in PostgreSQL 9.3 are event triggers which are bound to events such as creation of table or deletion of table.
Operators, Operator Classes, Operator Families Live in schemas. Many are predefined, but more can be added and allow you to define things such as +, =, etc. for custom data types.
Sequences Autocreated when defining columns as serial. In PostgreSQL, sequences are objects in their own right and can be shared across many tables.
Types Live in schemas. Don’t forget that you have the flexibility to create your own custom data types in PostgreSQL.
Foreign Data Wrappers, Servers and User Mappings Foreign Data Wrappers are remote data source drivers that allow you to access data in a non-PostgreSQL or remote PostgreSQL table. PostgreSQL 9.1 introduced these. 9.2 improved on general performance, and 9.3 introduced a new FDW called postgresfdw for connecting to other PostgreSQL servers, and also enhanced the API to support Foreign table updates.
Extensions [9.1+] Packaging of functions, tables, and other objects for easy deployment in a database. These are installed using CREATE EXTENSION.
Section 6


PostgreSQL comes bundled with several tools useful for administration and query writing.

Tool Description
psql Command-line client packaged with PostgreSQL. Good for automating SQL jobs, copying data, outputing simple HTML reports.
createdb, dropdb For creating and dropping a database from the OS shell.
pgAdminIII Popular graphical user interface packaged with PostgreSQL.
pg_restore Command-line tool for restoring compressed or .tar backups.
pg_dump Command-line tool for doing backups. Great for automated backups.
pg_dumpall Command-line tool for dumping all databases into a single backup.
pgAgent A daemon/service that can be downloaded from http://www.pgadmin.org/download/pgagent.php. Used for scheduling SQL jobs and batch shell jobs. Jobs can be added easily and monitored using the PgAdmin III job interface.
pg_basebackup Used for doing filesystem hot backup of db data cluster.
pg_upgrade Used for updating in place from one major version of PostgreSQL to another.
Section 7

PSQL Common Tasks

PSQL is a command-line tool that allows you to run ad-hoc queries, scripts, and other useful database management routines. PSQL runs in both a non-interactive mode (straight from the OS shell prompt) and an interactive mode (PSQL terminal prompt). In both modes, the following arguments apply:

Argument Description
-d Database. Defaults to the user (via system identification if no user is specified).
-h Server host. Defaults to localhost if not specified.
-p Port. Defaults to 5432 if not specified.
-U Username you are trying to log in with. Defaults to system user name.

PSQL Non-Interactive Mode

Getting help

$ psql –help

Execute an SQL script stored in a file

$ psql –h localhost -U postgres –p 5432 –f /path/to/pgdumpall.sql

Output data in html format

$ psql -h someserver -p 5432 -U postgres -d dzone -H -c "SELECT * FROM pg_tips" -o tips.html

Execute a single statement against a db

$ psql -U postgres –p 5432 -d dzone -c "CREATE TABLE test(some_id serial PRIMARY KEY, some_text text);"

Execute an SQL batch script against a database and send output to file

$ psql -h localhost -U someuser -d dzone -f /path/to/scriptfile.sql -o /path/to/outputfile.txt

PSQL Interactive Mode

To initiate interactive PSQL, type:

psql –U username –p 5432 –h localhost –d dzone

Once you are in the psql terminal there are a myriad of tasks you can perform. Below are some of the common ones.

Command Task
\q Quit
:q Cancel out of more screen
\? Help on psql commands
\h some_command Help on SQL commands
\connect postgres Switch database
List all databases \l
\dtv p* List tables and views that start with p.
\du List user/group roles and their group memberships and server level permissions.
\d sometable List columns, data types, and constraints for a table.
\i somefile Execute SQL script stored in a file.
\o somefile Output contents to file.
Retrieve prior commands Use up and down arrows.
\timing Toggle query timing on and off; when on, query output includes timing information.
\copy Copy from client computer to server and from server to client computer. Example: The following command string copies data to local client computer in CSV format with header.
\copy (SELECT * FROM sometable) TO 'C:/sometable.csv' WITH HEADER CSV FORCE QUOTE
\copy ... from program Allows you to copy output from an external program such as ls, dir, wget, curl. New in 9.3.
Section 8

Admin Tasks

Backup and Restore

Below are common backup and restore statements.

Create a compressed backup

pg_dump -h someserver -p 5432 -U someuser -F -c -b -v -f "/somepath/somedb.backup" somedb

Create a compressed backup of select tables

pg_dump -h localhost -p 5432 -U someuser -F -c -b -f "C:/somedb.backup" -t "someschema.table1" -t "someschema.table2" -v somedb

Create a compressed backup excluding a particular schema

pg_dump -h localhost -p 5432 -U someuser -F -c -b -f "C:/somedb.backup" -N someschema -v somedb

Restore a compressed backup

pg_restore –h localhost –d db_to_restore_to –U someuser /path/to/somedb.backup

Restore select schemas from backup

pg_restore –h localhost –d db_to_restore_to –U someuser -n someschema1 -n someschema2 /path/to/somedb.backup

Output a table of contents from backup file

pg_restore -l -f "C:/toc.txt" "C:/somedb.backup"

Restore only items in the table of contents

pg_restore -h localhost -d db_to_restore -U someuser -L "C:/toc.txt" "C:/somedb.backup"


Change globally work mem (9.4+)

Requires reload and some require restart.

ALTER SYSTEM SET work_mem TO ‘20MB’;
SELECT pg_reload_conf();

pg_dumpall currently only dumps to plain-text SQL. pg_dumpall backups must be restored with psql. For space savings and flexibility, use pg_dump. With pg_dump compressed and .tar backups, you can selectively restore objects. You cannot selectively restore with plain-text backups.

Below are common switches used with pg_dump [D], pg_restore [R], pg_dumpall [A]. These tools are packaged with PostgreSQL and are in the bin folder. They are also packaged with pgAdmin III and are in the PgAdmin III/version/ folder.

Switch Tool Description
-b, --blobs D Include large objects in dump.
-d, R Specify name of database to restore to.
-F, |t|p|d D R Specify backup file format (c = compressed, t = tar, p = plain text, d = directory). Plain-text backups must be restored with psql. Directory new in [9.2].
-c, --clean D R A Clean (drop) schema prior to create (for pg_dumpall drop database prior to create).
-g, --globals-only A Dump only global objects (roles, schemas, tablespaces), no databases.
-j, [8.4] , [9.2] D R Use this multiple parallel jobs to restore. This is especially useful for large backups and speeds them up significantly in many cases. 8.4 introduced parallel restore (pg_restore). 9.2 introduced (in pg_dump) parallel backup (needs to have format directory based).
-l, --list R Print summarized TOC of the archive.
-L, R Use TOC from this file for selcting/ordering output.
-n, D R Dump/restore only select objects in schema(s).
-N, D R Exclude from dump/restore named schema.
-r, --roles-only A Dump only roles, no database or tablespace.
-t, D Backup only named table(s) along with associated indexes, constraints, and rules.
-T, D Exclude named table(s) from backup.
-v --verbose D R A Controls verbosity.
[9.2] D Exclude dumping table data for specific table.
-s –-data|post-data|data [9.2] D R Dump or restore select parts. Pre-data just backs up or restores structures; post-data restores primary keys, foreign keys, and constraints. Data just restores data.
--if-exists [9.4] D Use IF EXISTS when dropping.

User Rights Management

These are SQL commands you can use to control rights. They can be run in the PSQL interactive, loading an SQL file, or via PgAdmin.

Create a new role with login rights that can create objects
Create a group role with no login rights and members inherit rights of role
Add a role to another role
GRANT somerole TO someotherrole;
Give rights to a role Example uses:
GRANT SELECT, UPDATE ON TABLE sometable TO somerole;
GRANT ALL ON TABLE sometable TO somerole;
somefunction TO   somerole;

-- Grant execute to all users
somefunction TO   public;
Revoke rights
REVOKE ALL ON TABLE sometable FROM somerole;
Give insert/update rights to select columns
GRANT INSERT, UPDATE (somecolumn) ON sometable TO somerole;
Grant rights to all future tables in a schema
Grant rights to all existing tables in a schema
Section 9

Data Definition (DDL)

Many of the examples we have below use named schemas. If you leave out the schema, objects created will be in the first schema defined in the search_path and dropped by searching the search path sequentially for the named object.

Create a new database
CREATE DATABASE postgresql_dzone;
Install extension in a database
Create a schema
CREATE SCHEMA someschema;
Changing database schema search path Sets the default schema to someschema.
ALTER DATABASE postgresql_dzone SET search_path = someschema, public;
Dropping objects with no dependencies A drop without a CASCADE clause will not drop an object if there are objects that depend on it, such as views, functions, and tables. For drop database you should be connected to a database other than the one you’re dropping.
DROP DATABASE postgresql_dzone;
DROP VIEW someview;
ALTER TABLE sometable DROP COLUMN somecolumn;
DROP FUNCTION somefunction;
Dropping object and all dependencies. (Use with caution.)
Create a table
CREATE TABLE test_scores(student varchar(100), 
score integer, test_date date DEFAULT CURRENT_DATE,
CONSTRAINT pk_test_scores PRIMARY KEY (student, test_date));
Create a child table
CREATE TABLE somechildtable (CONSTRAINT pk_somepk PRIMARY KEY (id)) INHERITS (someparenttable);
Create a check constraint
ALTER TABLE sometable ADD CONSTRAINT somecheckcontraint CHECK (id > 0);
Create or alter a view
[Prior to version 8.4, adding new columns to a view requires dropping and recreating].
Create a materialized view
SELECT * FROM sometable;
Refresh materialized view
Refresh materialized view without read blocking [9.4]
Create a view (doesn’t allow insert if data would not be visible in view) [9.4]
SELECT * FROM sometable 
Add a column to a table
Add a functional index to a table
CREATE INDEX idx_someindex ON sometable USING btree (upper(somecolumn));
Create a new type
CREATE TYPE sometype AS (somecolumn integer, someothercolumn integer[]);
Create a trigger
RETURNS trigger AS

IF OLD.somecolumn <> NEW.somecolumn OR 
(OLD.somecolumn IS NULL AND 
NEW.somecolumn IS NOT NULL) THEN
NEW.sometimestamp   := CURRENT_TIMESTAMP;

Add trigger to table
EXECUTE PROCEDURE sometriggerupdate();
Suppress redundant updates A built-in trigger that prevents updates that would not change any data.
CREATE TRIGGER trig_01_suppress_redundant BEFORE UPDATE ON sometable FOR EACH ROW
EXECUTE PROCEDURE suppress_redundant_updates_trigger();

A table can have multiple triggers, and each trigger for a particular event on a table is run in alphabetical order of the named trigger. So if order is important, name your triggers such that they are sorted in the order you need them to run.

Section 10

Query and Update (DML)

These are examples that show case some of PostgreSQL popular or unique query features.

Adding and Updating Data

Insert statement with multirows
INSERT INTO   test_scores(student,score,test_date)
VALUES ('robe', 95, '2014-01-15'),
('lhsu', 99, '2014-01-15'),
('robe', 98, '2014-07-15'),
('lhsu', 92, '2014-07-15'), 
('lhsu', 97,'2014-08-15');
Insert statement from SELECT, only load items not already in table
INSERT INTO tableA(id,price)
SELECT invnew.id,invnew.price
FROM tableB AS invnew LEFT JOIN tableA AS invold ON   (invnew.id = invold.id)
WHERE invold.price IS NULL; 
Cross update, only update items for a particular store where price has changed
UPDATE tableA 
SET price =   invnew.price
FROM tableB AS invnew
WHERE invnew.id = tableA.id 
AND NOT (invnew.price = tableA.price);
Insert from a tab-delimited file no header
COPY products FROM "/tmp/productslist.txt"   WITH DELIMITER '\t' NULL AS 'NULL';
Insert from a comma-delimited file with header row --these copy from the server’s file system
COPY products FROM "/tmp/productslist.csv"   WITH CSV HEADER NULL AS ‘NULL’;
Copy data to comma-delimited file and include header
--this outputs to the server’s file system
COPY (SELECT * FROM products WHERE product_rating =   ‘A’) TO ‘/tmp/productsalist.csv’ WITH CSV HEADER NULL AS ‘NULL’;

Retrieving Data

View running queries
SELECT * FROM pg_stat_activity;
Select the first record of each distinct set of data
-- this example selects the store and product
-- where the given store has the lowest price 
-- for the product. This uses PostgreSQL
-- DISTINCT ON and an order by to resort 
-- results by product_name.

SELECT r.product_id, r.product_name,   r.product_price 
FROM (SELECT DISTINCT   ON(p.product_id) p.product_id, p.product_name, s.store_name, i.product_price
FROM products AS p INNER JOIN inventory AS i
ON p.product_id = i.product_id 
INNER JOIN store AS s ON i.store_id = s.store_id
ORDER BY p.product_id,   i.product_price) AS r;
Get last date’s score for each student. Returns only one record per student
SELECT DISTINCT ON(student) student, score, test_date
FROM test_scores 
ORDER BY student, test_date DESC;
Use window function to number records and get running average
SELECT row_number() OVER(wt) AS rn, student, test_date,   
(AVG(score)   OVER(wt))::numeric(8,2) AS avg_run
FROM test_scores
WINDOW wt AS (PARTITION BY student ORDER BY test_date);
rn | student | test_date | avg_run ----+---------+------------+--------- 1 | lhsu | 2014-01-15 | 99.00 2 | lhsu | 2014-07-15 | 95.50 3 | lhsu | 2014-08-15 | 96.00 1 | robe | 2014-01-15 | 95.00 2 | robe | 2014-07-15 | 96.50
Get median values [9.4]
SELECT student, percentile_cont(0.5)
WITHIN GROUP (ORDER BY score) AS m_continuous,
WITHIN GROUP (ORDER BY score) AS m_discrete 
FROM test_scores GROUP BY student;
student | m_continuous | m_discrete --------+--------------+------------ lhsu | 97 | 97 robe | 96.5 | 95
Filtered aggregates [9.4] use instead of CASE WHEN (or subselect) (especially useful for aggregates like array_agg which may return nulls with CASE WHEN)
SELECT date_trunc('quarter',test_date)::date AS   qtr_start,
array_agg(score)   FILTER (WHERE student = 'lhsu') AS lhsu,
array_agg(score)   FILTER (WHERE student = 'robe') AS robe
FROM test_scores
GROUP BY date_trunc('quarter',test_date);
qtr_start | lhsu | robe ------------+---------+------ 2014-01-01 | {99} | {95} 2014-07-01 | {92,97} | {98}
Ordered aggregates, list scores in order of test date, one row for each student. Cast to make a string.
SELECT student, 
string_agg(score::text,   ',' ORDER BY test_date DESC) AS scores
FROM test_scores
GROUP BY student;
student | scores ---------+---------- lhsu | 97,92,99 robe | 98,95
Non-Recursive CTE with 2 CTE expressions. Note a CTE expression has only one WITH, each subexpression is separated by a , and the final query follows. Example returns the lowest priced car in each fuel_grade, limiting to just Japan, USA, German
( SELECT country_code, conv_us
FROM country 
WHERE country IN('Japan', 'USA','Germany')
prices AS
(SELECT p.car, p.fuel_grade,   price*c.conv_us AS us_price
FROM cars AS p 
ON   p.country_code = c.country_code
WHERE p.category = 'Cars'
prices.car, us_price
FROM prices
ORDER BY fuel_grade, us_price;
Recursive CTE * inventory, gives full name which includes parent tree name e.g. Paper->Color->Red->20 lbs
(SELECT id, item, parentid, 
CAST(item AS text) AS fullname
FROM products
WHERE parentid IS NULL
SELECT p.id,p.item, p.parentid, 
CAST(t.fullname || '->' 
|| p.item AS text) AS fullname
FROM products AS p
ON (p.parentid = t.id)
SELECT id, fullname 
FROM tree
ORDER BY fullname;
Section 11

Procedural Languages

PostgreSQL stands out from other databases with its extensive and extendable support for different languages to write database-stored functions. It allows you to call out to libraries native to that language. We will list the key language as well as some esoteric ones. The languages with an * are preinstalled with PostgreSQL and can be enabled. Some require further installs in addition to the language handler.

You can create set returning functions, simple scalar functions, triggers, and aggregate functions with most of these languages. This allows for languages that are highly optimized for a particular task to work directly with data without having to always copy it out to process as you normally would need with a simple database storage device. Language handlers can be of two flavors: trusted and untrusted. An untrusted language can access the filesystem directly.

From PostgreSQL 9.1+, languages not enabled by default in database or not built-in are installed using :

CREATE OR REPLACE somename(arg1 arg1type) 
RETURNS result_argtype AS
body goes here
LANGUAGE ‘somelang’;
Language Description Req
SQL* (trusted) Enabled in all databases. Allows you to write simple functions and set returning functions in just SQL. The function internals are visible to the planner, so in many cases it performs better than other functions since the planner can strategize how to navigate based on the bigger query. It is simple and fast, but limited in functionality.
CREATE OR   REPLACE FUNCTION prod_state(prev numeric, e1 numeric, e2 numeric).      
RETURNS   numeric AS        
SELECT   COALESCE($1,0) + COALESCE($2*$3,0);
c* Built in and always enabled. Often used to extend PostgreSQL (e.g. postgis, pgsphere, tablefunc) or, for example, to introduce new windowing functions (introduced in PostgreSQL 8.4). Functions are referenced from a .so or .dll file.
CREATE OR REPLACE FUNCTION st_summary(geometry)
'$libdir/postgis-2.1’, 'LWGEOM_summary'
plpgsql* (trusted) Not always enabled, but packaged so it can be installed.
CREATE FUNCTION cp_upd(p_key integer, p_value varchar)
IF EXISTS(SELECT test_id FROM testtable WHERE test_id =   p_key) THEN 
UPDATE   testtable 
SET   test_stuff = p_value 
WHERE test_id   = p_key;
INSERT INTO   testtable (test_id,
VALUES(p_key,   p_value);
plv8 (trusted) Good for manipulating JSON objects, reusing existing Javascript libraries, numeric processing. Comes packaged with 3 language bindings: Plv8 (aka PL/Javascript), plls (LiveScript), plcoffee (CoffeeScript). To install:
Google v8 engine

Example Functions

This next table demonstrates some examples of writing functions in various languages. For all functions you write, you can use the CREATE or REPLACE FUNCTION construction to overwrite existing functions that take same arguments. We use CREATE FUNCTION here.


Language Example
plperl (trusted), plperlu (untrusted)
my $text_to_quote = shift;
my $qfunc =   $_SHARED{myquote};
return   &$qfunc($text_to_quote);
$$ LANGUAGE plperl;
plpythonu, plpython2u, plpython3u (untrusted)
CREATE FUNCTION fnfileexists(IN fname text) RETURNS   boolean AS
import os
return os.path.exists(fname)
LANGUAGE 'plpythonu' STRICT;
plr Good for doing advanced stats and plotting using the R statistical language.
CREATE FUNCTION   r_quantile(float8[])
RETURNS float8[] AS 
quantile(arg1, probs = seq(0, 1, 0.25),
names = FALSE)
plv8 Allows you to write functions in JavaScript.
fib(n int)   RETURNS int AS $$
function fib(n)   {
return n<2 ? n : fib(n-1) +   fib(n-2)
return fib(n)

Common Procedural Tasks

Create a table trigger and use in table

CREATE OR REPLACE FUNCTION mytable_ft_trigger() 
RETURNS trigger AS $$
NEW.tsv :=
 coalesce(new.field1,'')), 'A') ||
 coalesce(NEW.field2,'')), 'B');
 return NEW;
$$ LANGUAGE plpgsql;

CREATE TRIGGER mytable_trigiu 
ON mytable 
 FOR EACH ROW EXECUTE PROCEDURE mytable_ft_trigger()

Return sets and use of out params

 fn_sqltestmulti(param_subject varchar, 
 OUT test_id integer, 
 OUT test_stuff text) 
 SELECT test_id, test_stuff 
 FROM testtable 
 WHERE test_stuff LIKE $1;

SELECT * FROM fn_sqltestmulti('%stuff%');

Return sets and use of table construct

 fn_sqltestmulti(param_subject varchar) 
 RETURNS TABLE(test_id integer, test_stuff text)
 SELECT test_id, test_stuff 
 FROM testtable 
 WHERE test_stuff LIKE $1;


Extensions extend the capabilities of PostgreSQL by providing additional data types, functions, index types, and more. After installing an extension, you need to run the following command to enable it:

CREATE EXTENSION extension_name;

Notable Extensions

Extension Description Link
PostGIS Adds support for geographic objects allowing location queries to be run using
pg_shard Shards and replicates tables for horizontal scaling and high availability. https://github.com/citusdata/pg_shard
pg_stat_statements Tracks execution statistics of all SQL statements. http://www.postgresql.org/docs/current/static/pgstatstatements.html
cstore_fdw Columnar store for PostgreSQL. https://github.com/citusdata/cstore_fdw
postgresql-hll Distinct value counting with tunable precision. https://github.com/aggregateknowledge/postgresql-hll
pgcrypto Cryptographic functions. http://www.postgresql.org/docs/current/static/pgcrypto.html
dblink Connections to other PostgreSQL databases from a database session. http://www.postgresql.org/docs/current/static/dblink.html

For a full list of extensions shipped with PostgreSQL see: http://www.postgresql.org/docs/current/static/contrib.html To search for third party extensions see: http://pgxn.org/


  • Featured
  • Latest
  • Popular
Design Patterns
Learn design patterns quickly with Jason McDonald's outstanding tutorial on the original 23 Gang of Four design patterns, including class diagrams, explanations, usage info, and real world examples.
213.7k 635.4k
Core Java
Gives you an overview of key aspects of the Java language and references on the core library, commonly used tools, and new Java 8 features.
135k 375.2k
Getting Started with Git
This updated Refcard explains why so many developers are migrating to this exciting platform. Learn about creating a new Git repository, cloning existing projects, the remote workflow, and more to pave the way for limitless content version control.
132.7k 303.4k
Getting Started with Ajax
Introduces Ajax, a group interrelated techniques used in client-side web development for creating asynchronous web applications.
102.9k 216.8k
Foundations of RESTful Architecture
The Representational State Transfer (REST) architectural style is a worldview that elevates information into a first-class element of architectures. REST allows us to achieve the architectural properties of performance, scalability, generality, simplicity, modifiability, and extensibility. This newly updated Refcard explains main HTTP verbs, describes response codes, and lists libraries and frameworks. It also gives additional resources to further explore each topic.
107k 185.4k
Spring Configuration
Catalogs the XML elements available as of Spring 2.5 and highlights those most commonly used: a handy resource for Spring context configuration.
105.3k 271.2k
Scrum is a framework that allows people to productively and creatively deliver products of the highest possible value. With over 70% of Agile teams using Scrum or Scrum hybrid, learn more about its benefits in managing complex product development. This newly updated Refcard explores the details of Scrum, including theory, values, roles, and events. It also includes a sample of a popular approach to deliver Integrated Increments in a scaled environment.
97.6k 261.8k
Core CSS: Part I
Covers Core principles of CSS that will expand and strengthen your professional ability to work with CSS. Part one of three.
91.9k 202.5k
jQuery Selectors
Introduces jQuery Selectors, which allow you to select and manipulate HTML elements as a group or as a single element in jQuery.
94.5k 361.9k
Core Java Concurrency
Helps Java developers working with multi-threaded programs understand the core concurrency concepts and how to apply them.
92.6k 203.8k
Getting Started with Eclipse
Eclipse IDE is a cross-platform, multi-purpose, open-source Integrated Development Environment. It is widely used to develop projects in Java, JavaScript, PHP, C++, Scala, and many others. This newly updated Refcard breaks down installing, setting up, and getting started with Eclipse. It also covers productivity tips, creating new projects and files, accessing Source Control Managers, and debugging configurations.
82.1k 229.4k
Core CSS: Part II
Covers Core principles of CSS that will expand and strengthen your professional ability to work with CSS. Part two of three.
74.9k 144.1k
{{ card.title }}
{{card.downloads | formatCount }} {{card.views | formatCount }}

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}