Load plugins from a --plugins-dir=plugins/ directory

[simonw]

In #14 and 33c7c53 I've added working support for setuptools entry_points plugins. These can be installed from PyPI using pip install ....

I imagine some projects will benefit from being able to add plugins without first publishing them to PyPI. Datasette already supports loading custom templates like so:

datasette serve --template-dir=mytemplates/ mydb.db

I propose an additional option, --plugins-dir= which specifies a directory full of blah.py files which will be loaded into Datasette when the application server starts.

datasette serve --plugins-dir=myplugins/ mydb.db

This will also need to be supported by datasette publish as those Python files should be copied up as part of the deployment.

[simonw]

The easiest way to implement this in Python 2 would be execfile(...) - but that was removed in Python 3. According to https://stackoverflow.com/a/437857/6083 2to3 replaces that with this, which ensures the filename is associated with the code for debugging purposes:

with open("somefile.py") as f:
    code = compile(f.read(), "somefile.py", 'exec')
    exec(code, global_vars, local_vars)

Implementing it this way would force this kind of plugin to be self-contained in a single file. I think that's OK: if you want a more complex plugin you can use the standard pluggy-powered setuptools mechanism to build it.

[simonw]

This needs unit tests. I also need to manually test the datasette package and datesette publish commands.

[simonw]

Here's the result of running:

datasette publish now fivethirtyeight.db \
    --plugins-dir=plugins/ --title="FiveThirtyEight" --branch=plugins-dir

https://datasette-phjtvzwwzl.now.sh/fivethirtyeight-2628db9?sql=select+convert_units%28100%2C+%27m%27%2C+%27ft%27%29

Where plugins/pint_plugin.py contains the following:

from datasette import hookimpl
import pint

ureg = pint.UnitRegistry()

@hookimpl
def prepare_connection(conn):
    def convert_units(amount, from_, to_):
        "select convert_units(100, 'm', 'ft');"
        return (amount * ureg(from_)).to(to_).to_tuple()[0]
    conn.create_function('convert_units', 3, convert_units)

[simonw]

This worked as well:

datasette package fivethirtyeight.db \
    --plugins-dir=plugins/ --title="FiveThirtyEight" --branch=plugins-dir

[simonw]

Added unit tests in 33c6bca

[simonw]

Here's the result of running this:

datasette publish heroku fivethirtyeight.db \
    --plugins-dir=plugins/ --title="FiveThirtyEight" --branch=plugins-dir

https://intense-river-24599.herokuapp.com/fivethirtyeight-2628db9?sql=select+convert_units%28100%2C+%27m%27%2C+%27ft%27%29