Cvcreator was a bit of a hack that was thrown together in the early days of the company.
Since I had time last half year, I've have taken the time to refactor the code.
The update required that we updated the specification for our content YAML files.
And considering that had to happen, some other changes were also made. These include:
1. Changing the content format from `.yaml` to `.toml`
2. New CLI name `cv` instead of `cvcreator`
3. Limitations to the choice of technical skills
1. Changing the content format from `.yaml` to `.toml`
In the backend the (home brewed) engine that filled text into template filled content was replaced with Jinja2.
This reduces a lot of edge cases, drastically reducing the amount of code, and simplifies the whole code base.
However, the template engine is pretty much incompatible with the "ad-hoc" format choices for the yaml content files.
At the minimum we had to make some small changes here and there.
Since we needed to alter the source files, we chose to go further to change the format as well to toml.
Toml is similar to yaml in capability, but have a more config-ini feel, and was
chosen since it was a good compromise between simplicity and expressiveness.
2. New CLI name `cv` instead of `cvcreator`
The `cvcreator` command line interface originally only did one thing. Now it
does many things that does not quite fit into a single command.
So the CLI was refactored to include subparsers where creating documents were only one feature.
The best way to get an overview of the new tool suite, is to just run the help command for the new CLI:
sh
$ cv --help
Usage: cv [OPTIONS] COMMAND [ARGS]...
Command line tool for creating curriculum vitae from TOML source files.
Options:
--help Show this message and exit.
Commands:
create Create CV as .pdf file
example Create example .toml file
latex Create CV as .tex file
skills List allowed technical skills
txt2yaml Convert old .txt to .yaml
yaml2toml Convert old .yaml to .toml
3. Limitations to the choice of technical skills
As the company grows, we want to create relevant statistics on the skills we have.
To do this, we need to be a bit more strict about the technical skills we can list in our CVs.
The idea is that we don't have mixing of variations of the same thing, like: `scikit-learn Scikitlearn Sklearn`.
We want only one way of writing the same thing, and we want to enforce that rule.
In this release, all skills that people are currently using have been included in some variation.
But should there come a need for new skills, this is okay, but require creating a pull-request against the Cvcreator repository.
Description of how to create such a PR is noted in the README.
For a current list of available technical skills, run:
sh
cv skills