• Manifest

Edit on GitHub

Manifest

The manifest file serves to declare a python package as an Odoo module and to specify module metadata.

It is a file called __manifest__.py and contains a single Python dictionary, where each key specifies module metadatum.

{
    'name': "A Module",
    'version': '1.0',
    'depends': ['base'],
    'author': "Author Name",
    'category': 'Category',
    'description': """
    Description text
    """,
    # data files always loaded at installation
    'data': [
        'views/mymodule_view.xml',
    ],
    # data files containing optionally loaded demonstration data
    'demo': [
        'demo/demo_data.xml',
    ],
}

Available manifest fields are:

name (str, required)

the human-readable name of the module

version (str)

this module's version, should follow semantic versioning rules

description (str)

extended description for the module, in reStructuredText

author (str)

name of the module author

website (str)

website URL for the module author

license (str, defaults: LGPL-3)

distribution license for the module

category (str, default: Uncategorized)

classification category within Odoo, rough business domain for the module.

Although using existing categories is recommended, the field is freeform and unknown categories are created on-the-fly. Category hierarchies can be created using the separator / e.g. Foo / Bar will create a category Foo, a category Bar as child category of Foo, and will set Bar as the module's category.

depends (list(str))

Odoo modules which must be loaded before this one, either because this module uses features they create or because it alters resources they define.

When a module is installed, all of its dependencies are installed before it. Likewise dependencies are loaded before a module is loaded.

data (list(str))

List of data files which must always be installed or updated with the module. A list of paths from the module root directory

demo (list(str))

List of data files which are only installed or updated in demonstration mode

auto_install (bool, default: False)

If True, this module will automatically be installed if all of its dependencies are installed.

It is generally used for "link modules" implementing synergic integration between two otherwise independent modules.

For instance sale_crm depends on both sale and crm and is set to auto_install. When both sale and crm are installed, it automatically adds CRM campaigns tracking to sale orders without either sale or crm being aware of one another

external_dependencies (dict(key=list(str)))

A dictionary containing python and/or binary dependencies.

For python dependencies, the python key must be defined for this dictionary and a list of python modules to be imported should be assigned to it.

For binary dependencies, the bin key must be defined for this dictionary and a list of binary executable names should be assigned to it.

The module won't be installed if either the python module is not installed in the host machine or the binary executable is not found within the host machine's PATH environment variable.

application (bool, default: False)

Whether the module should be considered as a fully-fledged application (True) or is just a technical module (False) that provides some extra functionality to an existing application module.

css (list(str))

Specify css files with custom rules to be imported, these files should be located in static/src/css inside the module.

images (list(str))

Specify image files to be used by the module.

installable (bool default: False)

Whether a user should be able to install the module from the Web UI or not.

maintainer (str)

Person or entity in charge of the maintenance of this module, by default it is assumed that the author is the maintainer.

{pre_init, post_init, uninstall}_hook (str)

Hooks for module installation/uninstallation, their value should be a string representing the name of a function defined inside the module's __init__.py.

pre_init_hook takes a cursor as its only argument, this function is executed prior to the module's installation.

post_init_hook takes a cursor and a registry as its arguments, this function is executed right after the module's installation.

uninstall_hook takes a cursor and a registry as its arguments, this function is executed after the module's uninstallation.

These hooks should only be used when setup/cleanup required for this module is either extremely difficult or impossible through the api.