sphinxalchemy – SQLAlchemy dialect for SphinxQL

This package provides a dialect for SQLAlchemy for interfacing with Sphinx search engine via SphinxQL. It allows reusing core SQLAlchemy parts to generate SphinxQL language constructs programmatically via sqlalchemy.sql and to maintain a pool of connections to Sphinx via sqlalchemy.pool.

Basic usage

You need sphinxalchemy to be installed to allow SQLAlchemy to pick up SphinxQL dialect classes. To install sphinxalchemy do:

% pip install sphinxalchemy

or if you like easy_install command better:

% easy_install sphinxalchemy

Now, as Sphinx uses MySQL protocol for interfacing via SphinxQL, you need to decide which one of supported by sphinxalchemy MySQL connectivity libraries to use – MySQLdb or mysqlconnector-python. Install one of those and use:

sphinx+mysqldb://[email protected]:port

to connect to Sphinx using MySQLdb or:

sphinx+mysqlconnector://[email protected]:port

to use mysqlconnector-python. You can now create engine using sqlalchemy.create_engine() function as usually:

from sqlalchemy import create_engine, MetaData

engine = create_engine("sphinx+mysqlconnector://[email protected]:port")
metadata = MetaData(bind=engine)

To define indexes (analogs of tables in relational databases) you should use sphinxalchemy.schema module:

from sphinxalchemy.schema import Index, Attribute, ArrayAttribute

documents = Index("documents", metadata,
  Attribute("created"),
  ArrayAttribute("tag_ids"))

Now you can query your documents index:

results = engine.execute(
  documents.select()
    .match("We think in generalities, but we live in details")
    .where(documents.c.tag_ids.in_([42, 1])))

API reference

class sphinxalchemy.schema.Index(*args, **kwargs)

Sphinx index metadata

Accepted arguments are the same as sqlalchemy.schema.Table accepts.

class sphinxalchemy.schema.Attribute(*args, **kwargs)

Sphinx index scalar attribute metadata

Accepted arguments are the same as sqlalchemy.schema.Column accepts.

class sphinxalchemy.schema.ArrayAttribute(*args, **kwargs)

Sphinx index array attribute metadata

Accepted arguments are the same as sqlalchemy.schema.Column accepts.

class sphinxalchemy.sphinxql.Select(columns, *args, **kwargs)

SphinxQL SELECT construct.

See corresponding doc section.

match(query)

Provide full text query for index

Sphinx uses extended query syntax in SphinxQL.

options(*args, **kwargs)

Provide options to execute query

Available options described here.

class sphinxalchemy.sphinxql.Replace(table, values=None, inline=False, bind=None, prefixes=None, returning=None, **kwargs)

Represent an REPLACE construct.

The Replace object is created using the replace() function.

See also:

Insert Expressions

sphinxalchemy.sphinxql.select(columns=None, whereclause=None, from_obj=[], **kwargs)

Factory for creating Select constructs.

Ressembles sqlalchemy.sql.expression.select().

sphinxalchemy.sphinxql.replace(table, values=None, inline=False, **kwargs)