Pydici is a software for consulting/IT services company to manage:
- commercial leads workflow (from business detection to sales and simple CRM)
- workload schedule to forecast people activity
- operation management: timesheet, margin control, billing
It is written in Python using the Django framework.
Pydici is available under the GNU Affero Public License v3 or newer (AGPL 3+). http://www.gnu.org/licenses/agpl-3.0.html
Sébastien Renard (Sebastien.Renard@digitalfox.org).
Pydici can be installed as any Django project. Just drop the code somewhere and setup Apache.
Python version >= 3.6 is required
To install all python prerequisites, please do the following: pip install -r requirements.txt. It is strongly advised to use a virtual env.
Drop source code in a directory readable by your apache user git clone https://github.com/digitalfox/pydici.git
Create a virtual env in a directory readable by your apache user and activate it virtual-env venv . venv/bin/activate
Install prerequisites :
pip install -r <path to pydici source code>/requirements.txt
Setup your favorite database (mysql/mariaDB or postgresql) and create a schema/base (with UTF-8 character set please) with a valid user that can create/alter/select/delete/update its objects.
Configure your database in pydici/settings.py. Look at django docs to understand the various database options.
Create tables with :
./manage.py migrate
Generate a new secret key with ./manage.py generate_secret_key and put it in pydici/settings.py
Collect static files with ./manage.py collectstatic
Setup your apache virtual env:
- activate mod_wsgi.
- activate ssl
- active mod_expires
- add Alias to /media and /static
- define auth backend. By default, pydici is designed to work with an http front end authentication. Look at https://docs.djangoproject.com/en/dev/howto/auth-remote-user/
Setup in cron (or your favorite scheduler) the followings tasks (adapt the schedule and options to your needs):
5 2 * * * <path to pydici>/venv/python -W ignore::DeprecationWarning <path to pydici>/manage.py clearsessions # Purge expired sessions in database
0 6 1 * * <path to pydici>/venv/python -W ignore::DeprecationWarning <path to pydici>/batch/timesheet_check.py # Warn for incomplete and surbooking timesheet for past month
0 6 2-31 * * <path to pydici>/venv/python -W ignore::DeprecationWarning <path to pydici>/batch/timesheet_check.py -w # Warn for incomplete timesheet for past month
0 6 21-31 * * <path to pydici>/venv/python -W ignore::DeprecationWarning <path to pydici>/batch/timesheet_check.py -m current -d 20 # Warn for incomplete timesheet on current month for 20th first days
0 * * * * <path to pydici>/venv/python -W ignore::DeprecationWarning <path to pydici>/manage.py process_tasks -d 3600 # Process tasks for 1 hour
After pulling the latest changes, you need to update the database.
There are three possible situations, depending on whether your installation uses Django Migration, uses South or does not use a migration system at all.
Your installation uses South if there is a south_migrationhistory
-
Your installation already uses Django Migration. Run these commands:
./manage.py migrate
-
Your installation uses South:
Firstly you need to checkout the last pydici version using South "pydici/master", then run these commands:
./manage.py syncdb ./manage.py migrate
After that you can update to to a newer version of pydici then :
./manage.py migrate auth ./manage.py migrate contenttypes ./manage.py migrate --fake # for internal pydici apps
Once this is done, future updates will be handled as situation #1.
-
Your installation does not use Django Migration or South yet.
Firstly you need to checkout the last pydici version using South "pydici/master", then run these commands:
./manage.py syncdb ./manage.py migrate --all 0001 --fake ./manage.py migrate
After that you can update to to a newer version of pydici by running:
./manage.py migrate --fake
Once this is done, future updates will be handled just like situation #1.
Scikit learn is a machine learning framework for python. It is an optional Pydici deps that can predict leads tags and state. Some comments:
- You need to install scikit-learn, numpy and scipy (pip install -r requirements-sklearn.txt)
- You might need to add the following directive in your apache virual host file to avoid some nasty deadlock during init when using scikit learn : WSGIApplicationGroup %{GLOBAL}
- For proper model caching, you might need to increase memcached object size (1m => 10m) as well as your python client memcache (hardcoded in lib for python-memcached...sic).
If you want to run javascript tests, you need to install node, phantomjs and casperjs. Node may be included in your distribution. For the two others, here's how to install localy, without any root access:
npm install phantomjs
npm install casperjs
On source:
./manage.py dumpdata -o dump.json -e contenttypes -e auth.Permission -e admin.LogEntry --natural-foreign
On target:
create empty database and play migrations
./manage.py loaddata dump.json