Prometheus-API is the Flask (About Flask) powered RESTful API behind Prometheus (About Prometheus).
Prometheus-API has been tested and known to work on the following configurations:
- MacOS X 10.9.5
- Ubuntu 14.04 LTS
- Python 2.7, 3.5, and 3.6
- Database abstraction with SQLAlchemy.
- Script support with Flask-Script.
- Database validation with SAValidation
- RESTful API generation with Flask-Restless
Clone the repo
git clone [email protected]:reubano/prometheus-api.git
Install requirements
cd prometheus-api
pip install -r base-requirements.txt
Run API server
manage serve
Now view the API documentation at http://localhost:5000
Prometheus-API comes with a built in script manager manage.py
. Use it to start the
server, run tests, and initialize the database.
manage <command> [command-options] [manager-options]
Start server
manage serve
Run tests
manage test
Run linters
manage lint
Initialize the dev database
manage initdb
Populate the production database
manage popdb -m Production
-m MODE, --cfgmode=MODE set the configuration mode, must be one of ['Production', 'Development', 'Test'] defaults to 'Development'. See `config.py` for details -f FILE, --cfgfile=FILE set the configuration file (absolute path)
checkstage Checks staged with git pre-commit hook cleardb Removes all content from database createdb Creates database if it doesn't already exist initdb Removes all content from database and creates new tables serve Runs the Flask development server i.e. app.run() runtests Run nose tests shell Runs a Python shell inside Flask application context.
Type manage <command> -h
to view any command's options
manage manage serve -h
usage: manage serve [-h] [-t HOST] [-p PORT] [--threaded] [--processes PROCESSES] [--passthrough-errors] [-d] [-r] Runs the Flask development server i.e. app.run() optional arguments: -h, --help show this help message and exit -t HOST, --host HOST -p PORT, --port PORT --threaded --processes PROCESSES --passthrough-errors -d, --no-debug -r, --no-reload
Start production server on port 1000
manage serve -p 1000 -m Production
The following configurations settings are available in config.py
:
variable | description | default value |
---|---|---|
__YOUR_EMAIL__ | your email address | <user>@gmail.com |
API_METHODS | allowed HTTP verbs | ['GET', 'POST', 'DELETE', 'PATCH', 'PUT'] |
API_ALLOW_FUNCTIONS | allow sqlalchemy function evaluation | TRUE |
API_ALLOW_PATCH_MANY | allow patch requests to effect all instances of a given resource | TRUE |
API_MAX_RESULTS_PER_PAGE | the maximum number of results returned per page | 1000 |
API_URL_PREFIX | string to prefix each resource in the api url | '' |
See the Flask-Restless docs for a complete list of settings.
Prometheus-API will reference the SECRET_KEY
environment variable in config.py
if it is set on your system.
To set this environment variable, do the following:
echo 'export SECRET_KEY=value' >> ~/.profile
For a list of available resources, example requests and responses, and code samples, view the online documentation. View the Flask-Restless guide for more request/response examples and directions on making search queries.
The Prometheus API is separated into different services, each responsible for performing a specific set of tasks.
service | description |
---|---|
Hermes | price/event data aggregate |
Cronus | portfolio performance analytics and allocation engine |
Icarus | portfolio risk profiler (coming soon) |
Oracle | random portfolio generator (coming soon) |
Lynx | portfolio x-ray engine (coming soon) |
Rosetta | 3rd party portfolio data converter (coming soon) |
Ideally, you should install python modules for every project into a virtual environment. This setup will allow you to use different versions of the same module in different projects without worrying about adverse interactions.
sudo pip install virtualenv virtualenvwrapper
Add the following to your ~/.profile
export WORKON_HOME=$HOME/.virtualenvs
export PIP_VIRTUALENV_BASE=$WORKON_HOME
export PIP_RESPECT_VIRTUALENV=true
source /usr/local/bin/virtualenvwrapper.sh
Create your new API virtualenv
cd prometheus-api
mkvirtualenv --no-site-packages prometheus-api
sudo easy_install pip
sudo pip install -r base-requirements.txt
Getting Gevent up and running is a bit tricky so follow these instructions carefully.
To use gevent
, you first need to install libevent
.
Linux
apt-get install libevent-dev
Mac OS X via homebrew
brew install libevent
Mac OS X via macports
sudo port install libevent
Mac OS X via DMG
download on Rudix
Now that libevent is handy, install the remaining requirements
sudo pip install -r requirements.txt
Or via the following if you installed libevent from macports
sudo CFLAGS="-I /opt/local/include -L /opt/local/lib" pip install gevent
sudo pip install -r requirements.txt
Finally, install foreman
sudo gem install foreman
Now, you can run the application locally
foreman start
You can also specify what port you'd prefer to use
foreman start -p 5555
If you haven't signed up for Heroku, go ahead and do that. You should then be able to add your SSH key to Heroku, and also heroku login from the commandline.
Install heroku and create your app
sudo gem install heroku
heroku create -s cedar app_name
Add the database
heroku addons:add heroku-postgresql:dev
heroku pg:promote HEROKU_POSTGRESQL_COLOR
Push to Heroku and initialize the database
git push heroku master
heroku run python manage.py createdb -m Production
Start the web instance and make sure the application is up and running
heroku ps:scale web=1
heroku ps
Now, we can view the application in our web browser
heroku open
And anytime you want to redeploy, it's as simple as git push heroku master
.
Once you are done coding, deactivate your virtualenv with deactivate
.
tree . | sed 's/+----/├──/' | sed '/.pyc/d' | sed '/.DS_Store/d'
prometheus-api
├──Procfile (heroku process)
├──README.rst (this file)
├──app
| ├──__init__.py (main app module)
| ├──helper.py (manager/test helper functions)
| ├──LICENSE
| ├──MANIFEST.in (pypi includes)
| ├──models
| | ├──__init__.py
| | ├──cronus.py (portfolio analytics engine models)
| | ├──hermes.py (price/event data aggregator models)
| ├──README.rst (symlink for pypi)
| ├──setup.py (pypi settings)
| ├──tests
| ├──__init__.py (main tests module)
| ├──standard.rc (pylint config)
| ├──test.sh (git pre-commit hook)
| ├──test_hermes.py (hermes model tests)
| ├──test_site.py (site tests)
├──app.db (app development database)
├──config.py (app config)
├──manage.py (flask-script)
├──requirements.txt (python module requirements)
├──runtime.txt (python version)
├──schema.png (database relationship model)
├──setup.cfg (unit test settings)
First time
- Fork
- Clone
- Code (if you are having problems committing because of git pre-commit
hook errors, just run
manage checkstage
to see what the issues are.) - Use tabs not spaces
- Add upstream
git remote add upstream https://github.com/reubano/prometheus-api.git
- Rebase
git rebase upstream/master
- Test
manage runtests
- Push
git push origin master
- Submit a pull request
Continuing
- Code (if you are having problems committing because of git pre-commit
hook errors, just run
manage checkstage
to see what the issues are.) - Use tabs not spaces
- Update upstream
git fetch upstream
- Rebase
git rebase upstream/master
- Test
manage runtests
- Push
git push origin master
- Submit a pull request
git shortlog -sn
commits: 430
430 Reuben Cummings
Prometheus tells you how your stock portfolio has performed over time, gives insight into how to optimize your asset allocation, and monitors your portfolio for rebalancing or performance enhancing opportunities.
Flask is a BSD-licensed microframework for Python based on Werkzeug, Jinja2 and good intentions.
Prometheus API is distributed under the BSD License, the same as Flask on which this program depends.