Build, Continuous Integration, Deploy in python projects?

Question

Python

The question is what tools are accepted and convenient to use to build python projects, what CI is preferred by the python developer community, what is used to deploy code to servers.
The experience of web projects is more interesting, but I will be glad to see examples from other areas.

Answer 1

They use virtualenv and pip to deploy their web projects to the server . Dependencies are usually listed in a pip file.
at the same time, project deployment is reduced to a set of commands: The code that launches the project on a production server must take into account the presence of a virtual environment. If you use Apache and wsgi, this describes how the script should look. Why all these dances with virtualenv? All this is necessary so that several projects with different versions of dependent programs can work quietly on the same server. (for example, each project may have its own version of django)
svn\git clone to <proj_dir>
virtualenv <env_dir>
source <env_dir>/bin/activate
pip install -r <proj_dir>/requrements.pip


Answer 2

For "assembly" I use buildout. Compared to virtualenv, you have to do fewer actions by hand, everything is more or less standardized.
I don't use anything for CI yet. :( I run tests by hand before pushing to the repository.

Answer 3

hg -> project tree with a solid YAML-based configuration -> python-doit tasks -> nose tests -> packaging (python-doit) -> fabric deployment -> version update && data migration (python-doit).

Answer 4

git/hg/svn?

Answer 5

I use git as CVS, for quick deployment to a production server - a small script with a call to rsync. Like all magic.