Getting started with Psycopg 3

This section of the documentation will explain how to install Psycopg and how to perform normal activities such as querying the database or loading data using COPY.

Important

If you are familiar with psycopg2 please take a look at Differences from psycopg2 to see what is changed.

Contents:

• Installation
  • Supported systems
  • Binary installation
  • Local installation
  • Pure Python installation
  • Installing the connection pool
  • Handling dependencies
• Basic module usage
  • Main objects in Psycopg 3
  • Shortcuts
  • Connection context
  • Adapting pyscopg to your program
• Passing parameters to SQL queries
  • execute() arguments
  • Danger: SQL injection
  • Binary parameters and results
• Adapting basic Python types
  • Booleans adaptation
  • Numbers adaptation
  • Strings adaptation
  • Binary adaptation
  • Date/time types adaptation
  • JSON adaptation
  • Lists adaptation
  • UUID adaptation
  • Network data types adaptation
  • Enum adaptation
• Adapting other PostgreSQL types
  • Composite types casting
  • Range adaptation
  • Multirange adaptation
  • Hstore adaptation
  • Geometry adaptation using Shapely
• Transactions management
  • Autocommit transactions
  • Transaction contexts
  • Transaction characteristics
  • Two-Phase Commit protocol support
• Using COPY TO and COPY FROM
  • Writing data row-by-row
  • Reading data row-by-row
  • Copying block-by-block
  • Binary copy
  • Asynchronous copy support
  • Example: copying a table across servers
• Differences from psycopg2
  • Server-side binding
  • Multiple statements in the same query
  • Multiple results returned from multiple statements
  • Different cast rules
  • You cannot use IN %s with a tuple
  • Different adaptation system
  • Copy is no longer file-based
  • with connection
  • callproc() is gone
  • client_encoding is gone
  • No default infinity dates handling
  • What’s new in Psycopg 3