Voodoo is a template system for project skeletons (similar to the template part of PasteScript):
- It can make a copy of a project skeleton processing some files, filter others, etc.
- Works with local paths and git/hg URLs.
- It generates a beautiful output and take care of not overwrite existing files, unless instructed to do so.
- BSD License. See
LICENSE
for more details.
Above, Voodoo sample output as used in a program
Requirements:
- Pypy or Python 2.6, 2.7, 3.3 or newer.
- It also uses the Jinja2 and Colorama Python libraries.
Install it by doing:
$ pip install voodoo
Then, whenever you want to clone a project template to make a new project, do:
$ voodoo <Git URL or path to another folder> <New project name>
Example:
$ voodoo [email protected]:lucuma/voodoo-flask.git myapp
First of all install it and/or add it to your requirements.txt:
$ pip install voodoo
The API is very simple. A render_skeleton
function that takes two
absolute paths: the project skeleton to process, and where to copy it.:
from voodoo import render_skeleton
render_skeleton(skeleton_path, new_project_path)
It also provide a prompt
and prompt_bool
functions that take
user input, to help you to make interactive scripts.
The content of the files inside the skeleton are copied to the destination without changes, unless are suffixed with the extension '.tmpl'. In that case, the templating engine will be used to render them.
A slightly customized Jinja2 templating is used. The main difference is
that variables are referenced with [[ name ]]
instead of
{{ name }}
and blocks are [% if name %]
instead of
{% if name %}
. To read more about templating see the Jinja2
documentation.
Use the data argument to pass whatever context you want to be available in the templates. The arguments can be any valid Python value, even a function:
from hashlib import sha256
from os import urandom
from voodoo import render_skeleton
data = {
'package': 'demo',
'py3': True,
'make_secret': lambda: sha256(urandom(48)).hexdigest()
}
render_skeleton(skeleton_path, new_project_path, data=data)
so in your template you can have:
import [[ package ]]
secret = '[[ make_secret() ]]'
[% if py3 %]
msg = 'Python 3!'
[% else %]
msg = 'meh'
[% endif %]
Your folders and files can also use the [[ name ]]
notation: mytemplate/[[ foo ]]/something/[[ bar ]].txt
It's easy to integrate Voodoo with your own scripts. The following example it's a classic make new project script found in many popular frameworks:
from os.path import join, dirname, basename
from voodoo import render_skeleton
default_context = {
'foo': 'bar',
}
SKELETON_PATH = join(dirname(__file__), '..', 'tests', 'demo')
def new_project(path, options):
data = default_context.copy()
data['project_name'] = basename(path)
render_skeleton(SKELETON_PATH, path, data=data, **options)
if __name__ == '__main__':
import argparse
parser = argparse.ArgumentParser(description='Create a new project')
parser.add_argument('path', help='The name or fullpath of the new project')
parser.add_argument('-p', '--pretend', action='store_true',
help='Run but do not make any changes')
parser.add_argument('-f', '--force', action='store_true',
help='Overwrite files that already exist, without asking')
parser.add_argument('-s', '--skip', action='store_true',
help='Skip files that already exist, without asking')
parser.add_argument('-q', '--quiet', action='store_true',
help='Suppress status output')
args = parser.parse_args()
da = vars(args)
new_project(da.pop('path'), da)
You can se this example working in the examples folder. Play with it, generate a new project and manually update some files. Then run the script again to see how it detects what files has changed, and what files are identical and with no need of regeneration.
An interactive version of this script could be made using the
voodoo.prompt
and/or the voodoo.prompt_bool
helper functions.
Voodoo supports rendering project skeletons from a version control system repository. Git and Mercurial are supported. It requires a working VCS command on your path: git or hg.
The forms of the URL is the same that the pip
installer uses,
detecting the type of VCS using URL prefixes: "git+" or "hg+".
Currently supports cloning over git
, git+https
and git+ssh
:
git+git://git.myproject.org/MyProject git+https://git.myproject.org/MyProject git+ssh://git.myproject.org/MyProject [email protected]:MyProject
Passing branch names, a commit hash or a tag name is possible like so:
git://git.myproject.org/MyProject.git@master git://git.myproject.org/[email protected] git://git.myproject.org/MyProject.git@da39a3ee5e6b4b0d3255bfef95601890afd80709
The supported schemes are: hg+http
, hg+https
, hg+static-http
and hg+ssh
:
hg+http://hg.myproject.org/MyProject hg+https://hg.myproject.org/MyProject hg+ssh://hg.myproject.org/MyProject
You can also specify a revision number, a revision hash, a tag name or a local branch name like so:
hg+http://hg.myproject.org/MyProject@da39a3ee5e6b hg+http://hg.myproject.org/MyProject@2019 hg+http://hg.myproject.org/[email protected] hg+http://hg.myproject.org/MyProject@special_feature
render_skeleton (src_path, dst_path, data=None, filter_ext=None, pretend=False, force=False, skip=False, quiet=False, envops=None)
- src_path:
- Absolute path to the project skeleton
- dst_path:
- Absolute path to where to render the skeleton
- data:
- Data to be passed to the templates, as context.
- filter_this:
- A list of names or shell-style patterns matching files or folders that musn't be copied. The default is:
['.*', '~*', '*.py[co]']
- include_this:
- A list of names or shell-style patterns matching files or folders that must be included, even if its name are in the filter_this list. Eg:
['.gitignore']
. The default is an empty list. - pretend:
- Run but do not make any changes
- force:
- Overwrite files that already exist, without asking
- skip:
- Skip files that already exist, without asking
- quiet:
- Suppress the status output
- envops:
- Extra options for the Jinja template environment.
prompt (text, default=None)
Ask a question via raw_input() and return their answer.
- text:
- prompt text
- default:
- default value if no answer is provided.
prompt_bool (text, default=False, yes_choices=None, no_choices=None)
Ask a yes/no question via raw_input() and return their answer.
- text:
- prompt text
- default:
- default value if no answer is provided.
- yes_choices:
- default
['y', 'yes', '1', 'on', 'true', 't']
- no_choices:
- default
['n', 'no', '0', 'off', 'false', 'f']
copyright: | Juan-Pablo Scaletti. |
---|---|
license: | BSD, see LICENSE for more details. |