New 2.0 Context Implementation #1008
Conversation
…issing coverage lines column
… the new context.py
… use it for portability
…ed v1 overwrite call structure.
|
Come to think of it, now that I am more familiar with the front-end of Cookiecutter (pre-jinja2 rendering) it might be more stream-lined to get rid of the code that processes v1 templates, and just convert all v1 files internally to version 2 files and then do the version 2 processing. The conversion is straightforward and it simplifies the code paths and maintenance. |
…sed by other variable fields
|
In the course of being a user of this fork and having a cookiecutter.json file that is rather complex, I have added a few more features that make conditional processing of variables more intuitive. These features are: I am in the middle of testing these now... |
# Conflicts: # cookiecutter/__init__.py # cookiecutter/generate.py # cookiecutter/main.py # setup.cfg # tests/test_main.py
New cookiecutter context
Bring your implementation up to date with cookiecutter 2.0
|
I'm closing this PR due to its age and size. |
For your consideration - an implementation of version 2 context format as specified by hackebrot's pull request 848 and using hackebrot's context.py module from his new context format branch here.
My detailed dev notes on the implementation are available here along with Travis and AppVeyor reports. The dev notes cover implementation goals, implementation details, and testing.
The high-altitude summary is - one new file added, two existing files modified, and created 48 new tests for code coverage. Currently not passing Python 2.7, but 3.3 - 3.6 are green. I might need help with 2.7. Assuming of course, this pull request gains traction.
Added a ReadTheDocs project that documents the version 2 Cookiecutter template format that this rull requests supports.
Also, I have a repo containing a command line utility that will convert a version 1 template file to a version 2 file.