Valdo

(like "Vala Do", pronounced like "Waldo")

Create a new Vala project from a repository of templates.

Installation

Packages are available for Fedora, Ubuntu, Debian, and more via OBS.

To install from source, clone this repository and run: meson build && meson install -C build

Example use

valdo new - initializes a new project (from a template new) in the current directory

valdo gtk - initializes a new GTK app

valdo lib - initializes a new library

valdo - lists all available templates

Creating a new a template

Templates should be added here so that they can be used by everyone.

Fork this repository and add a new directory under templates/. The name of the directory is the template name. Then you need to add a template.json describing the template, the variables it uses, and the files that need to be substituted.

Here is what a template might look like:

.
├── meson.build
├── README.md
├── src
│   ├── main.vala
│   └── meson.build
└── template.json

1 directory, 5 files

Your template.json may start off looking like:

{
    "description": "a bare app, with minimal dependencies",
    "variables": {
        "PROGRAM_NAME": {
            "summary": "the name of the program",
            "default": "main"
        }
    },
    "templates": [
        "src/meson.build",
        "meson.build",
        "README.md"
    ]
}

"templates" is a list of template files, containing variables to be substituted.

"variables" is a dictionary mapping each variable name to a short description. Having a default value for a variable is optional. There are at least two variables every template uses, PROJECT_NAME and PROJECT_VERSION, which you don't have to specify.

For every variable listed, the template engine will substitute ${VARIABLE_NAME} in each templated file. If a variable does not have a default value, Valdo will prompt the user. The rest of the files in the template directory will be copied over unmodified.

Once you're done, submit a PR to https://github.com/vala-lang/valdo

Advanced template features

Variable substitution

You can define a variable's default value in terms of another variable like so:

{
    "variables": {
        "API_NAMESPACE": {
            "summary": "the API namespace",
            "default": "MyLib",
            "pattern": "^[A-Za-z][[:word:]]*$"
        },
        "LIBRARY_NAME": {
            "summary": "the name of the library",
            "default": "/${API_NAMESPACE}/\\w+/\\L\\g<0>\\E/",
            "pattern": "^[[:word:]][[:word:]-]*$"
        },
        "APP_ID": {
            "summary": "the application ID",
            "default": "com.${USERNAME}.${API_NAMESPACE}",
            "pattern": "^\\w+(\\.\\w+)*$"
        }
    }
}

Here, this means that LIBRARY_NAME will be auto-generated from the namespace name, unless the user enters something different in a prompt [^1]. In this case, if the user enters nothing for API_NAMESPACE and LIBRARY_NAME, the result will be mylib.

The syntax is to write the variable name, then a regular expression that matches something you want to replace, then a regular expression for the replacement. Here, \L says "make lowercase", \g<0> matches group 0 or the entire string matched, and \E ends the conversion. Consult the docs for GLib.Regex.replace() for more information about how regex syntax is used in GLib.

If you wanted to substitute another pattern in the text, you just have to add another pair of regexes separated by a /. For example:

"default": "/${API_NAMESPACE}/\\w+/\\L\\g<0>\\E/bob/greta/"

After converting everything to lowercase, then substitute "bob" for "greta". The substitutions will be applied in the order they appear, one after another.

[^1]: You can specify "auto": true for the variable to make it completely automatic, so there's no prompting the user. This is useful when generating variable names in build scripts, for example.