mosql 0.5

It lets you use the common Python's data structures to build SQLs, and provides a convenient model of result set.

The full version of this documentation is at mosql.mosky.tw.

MoSQL — More than SQL

It lets you use the common Python’s data structures to build SQLs, and provides a convenient model of result set.

The main features:

1. Easy-to-learn — Everything is just plain data structure or SQL keyword. See The SQL Builders.

2. Convenient — It makes result set more easy to use. See The Model of Result Set.

3. Secure — It prevents the SQL injection from both identifier and value.

4. Faster — It just builds the SQLs from Python’s data structure and then send it via the connector.

It is just “More than SQL”.

NOTE: The versions after v0.2 is a new branch and it does not provide backward-compatibility for v0.1.x.

The SQL Builders

>>> from mosql import build
>>> build.select('author', {'email like': '%mosky%@%'})
SELECT * FROM "author" WHERE "email" LIKE '%mosky%@%'

It is very easy to build a query by Python’s data structures and mosql.build.

There is more explanation of the builders — mosql.build.

It also provides mosql.result.Model for result set, and you can use the same way to make queries to database.

The Model of Result Set

Here is a SQL and the result set:

mosky=> select * from detail where person_id in ('mosky', 'andy') order by person_id, key;
 detail_id | person_id |   key   |           val
-----------+-----------+---------+--------------------------
         5 | andy      | email   | [e-mail]
         3 | mosky     | address | It is my first address.
         4 | mosky     | address | It is my second address.
         1 | mosky     | email   | [e-mail]
         2 | mosky     | email   | [e-mail]
        10 | mosky     | email   | [e-mail]
(6 rows)

Then, use the model configured (the module, detail, is in the examples) to do so:

>>> from detail import Detail
>>> for detail in Detail.arrange({'person_id': ('mosky', 'andy')}):
...     print detail
...
{'detail_id': [5],
 'key': 'email',
 'person_id': 'andy',
 'val': ['[e-mail]']}
{'detail_id': [3, 4],
 'key': 'address',
 'person_id': 'mosky',
 'val': ['It is my first address.', 'It is my second address.']}
{'detail_id': [1, 2, 10],
 'key': 'email',
 'person_id': 'mosky',
 'val': ['[e-mail]', '[e-mail]', '[e-mail]']}

Here I use arrange for taking advantages from the model configured, so the result sets are grouped into three model instances, but the plain methods, such as select, are also available.

It converts the result set to column-oriented models, and the column can be squashed. The non-list value above is just the squashed column. See mosql.result for more information.

There is more explanation of the model — mosql.result.

Installation

It is easy to install MoSQL with pip:

$ sudo pip install mosql

Or clone the source code from Github:

$ git clone git://github.com/moskytw/mosql.git