chevdor / tera-cli

A command line utility on top of the tera templating engine. Takes json|yaml|toml as input and can merge ENV in. You may see it as envsubst on steroid.

tera-cli

Intro

tera is a template engine written in Rust and inspired by Jinja2. It allows merging some data called context data into a template and produces a new output. This project, tera-cli, is a command line for the tera template engine.

This project is called tera-cli but the command installed on your system is simply tera.

tera-ci offers powerful features related to your environment variables, allowing you to control the output both from the context data you pass but also from the ENV variables set on your system.

Example

Here is a basic example. For instance, you will pass data such as:

.data.json:
{
    "title": "Demo",
    "users": [
        {
            "username": "Alice",
            "url": "http://example.org/alice",
            "fav_colors": ["red", "green", "yellow"]
        },
        {
            "username": "Bob",
            "url": "http://example.org/bob",
            "fav_colors": ["orange"]
        }
    ]
}

as well as a template such as:

.template.tmpl
<title>{% block title %} {{title}} {% endblock title %}</title>

<ul>
{% for user in users -%}
    <li><a href="{{ user.url }}">{{ user.username }}
    {{ user.username }} likes {% for color in user.fav_colors -%}{{ color }} {% endfor %}
    </a></li>
{% endfor %}
</ul>

and a call such as tera --template template.tera data.json will produce:

.result
<title> Demo </title>

<ul>
<li><a href="http://example.org/alice">Alice
    Alice likes red green yellow
    </a></li>
<li><a href="http://example.org/bob">Bob
    Bob likes orange
    </a></li>

</ul>

The tera engine allows way more than the simple replacements shown above. You may check out the doc for more information. To name only a few, tera offers the following:

• variables & expressions (you can do math…​)

• comments

• control structure & loops (if, for, …​)

• filters

• formatting functions (show a file size, format a date, etc…​)

• inheritance, include, etc…​

• built-ins: capitalize strings, replace, trim, etc…​

Install

cargo install --git https://github.com/chevdor/tera-cli

Hot reload

You may find it useful to watch a folder with your templates and run tera if a template changes. For this to work, it is recommended to name you template as foobar.md.tera if your template expands into a markdown file for instance. You may then use fswatch and watch a templates folder using:

fswatch templates -e ".*\.md$" | \
    xargs -n1 -I{} \
    tera --include-path templates \
        --template templates/template.md.tera context.json

Execute as Docker container

You can find a tera Docker image at chevdor/tera. The image is very small and should be less than 8MB.

You can test it with:

docker run --rm -it chevdor/tera --version

The Docker image mentioned above is not yet built by the CI so you may not find the very latest version from time to time.

Build container image

docker build --tag tera-cli .

Execute tera from the Docker container

Check the tera help

docker run -it --rm tera-cli --help

Parse a template

docker run -it --rm \
    --volume="$(pwd)/templates:/templates" \
    --read-only \
    --env=FOO=BAR \
    tera-cli --template /templates/env-debug.txt --env-only --env-key env

What can I do with that anyway ?

Well…​ if you have data and you want to format them, this tool will likely be a great companion.

• You may generate beautiful changelogs in markdown, asciidoc, restructured text, etc…​

• You may generate some more human views of your data

• You may…​ make a blog with that…​

• You may generate k8s config files…​.

Features

Supported formats

You may pass the context data either as file of into stdin.

Current stdin supports only json.

ENV support

There are several options related to the environment variables.

Enable ENV variables injection

By default, the environments variables are not merged in. You can turn this feature on with --env.

Collisions

Now that you enabled the merging of the ENV variables, it is important to understand that, in some cases, your ENV may collide with your context data. This can be convenient if you want your ENV to override the context data.

ENV injection priority

If you prefer the context data to overwrite the ENV, you may use --env-first. As a result, the ENV will be applied first to the context and your context data will be loaded afterward.

Collisions handling

You may perfer to consider collisions as failures. This is what --fail-on-collision is for. If a collision is detected, the program will exit with a status code of 1 and an appropriate message.

ENV only

You may also want to ONLY load ENV variables as context data. This is what --env-only does.

ENV sub key

By default, your ENV variables will be loaded at the root of the context data. For instance, the HOME ENV variable will be then available in your tera template as {{ HOME }}. As we just mentioned, collisions may be an issue. There is an easy to prevent them entirely: you may move the ENV into a sub key in the context data. This is allowed thanks to the --env-key <name> option. For instance, using --env-key env will make your HOME ENV variable available in the tera template as {{ env.HOME }}.

While the syntax is a little more verbose, paired with --fail-on-collision, this option allows ensuring that nothing happens in your back.

External files

Using the --include flag, the command will scan recursively for files that could be included, used as macros or for inheritance. By default, it will scan the folder where the main template is located, unless the --include-path option is given.

From this repository, you can test the include feature with the command:

USER="[YOURNAME]" tera --template data/include/hello.txt --include --env-only

and test the inheritance feature with:

USER="[YOURNAME]" tera --template data/inheritance/child.txt --inherit --env-only

Content escaping

Passing the -a | --escape flag allows escaping the content.

Usage

Command line utility for the tera templating engine. You need to provide a template using the tera syntax as well as some data (various format are supported)

Usage: tera [OPTIONS] --template <TEMPLATE> [CONTEXT]

Arguments:
  [CONTEXT]  Location of the context data. This file can be of the following type: json | toml | yaml. If you prefer to pass the data as stdin, use `--stdin`

Options:
  -t, --template <TEMPLATE>          Location of the template
  -i, --include                      This flag tells the command to parse all templates found in the same path where the given template is located [aliases: inherit]
      --include-path <INCLUDE_PATH>  Option to define a different path from which search and parse templates [aliases: inherit-path]
  -s, --stdin                        The context data can be passed using stdin
  -e, --env                          If true, the current ENV will be appended to the data under the --env-key key
      --env-key <ENV_KEY>            By default, if --env is set, the environment variables will be attached at the root of the context. This is convenient but may end up conflicting with your data. To prevent collisions, you can provide a custom key with this option
      --env-first                    By default, the context is made of the data you pass and the ENV is applied afterwards. Setting this option will apply the ENV first. This is interesting if you prefer your data to override the ENV
      --fail-on-collision            if you prefer your data to override the ENV
      --env-only                     If you want to solely use the ENV as context, you may pass this option. This will prevent an error about no context being passed to be raised
  -o, --out <OUT>                    Optional output file. If not passed, using stdout
  -a, --escape                       Auto-escape rendered content. This is useful for HTML output
  -h, --help                         Print help
  -V, --version                      Print version