Breaking Changes

init_xtask now takes an XtaskArgs parameter and the argument parsing is done with a dedicated function parse_args<C: clap::Subcommand>. You need to update the call to init_xtask into two function calls. This allows to mutate the command arguments before actually initializing xtask.

Replace:
```
let args = init_xtask::<Command>()?;
```
With:
```
let args = init_xtask::<Command>(parse_args::<Command>()?)?;
```
The execution environment has been renamed to the context which is more accurate and broad. The flag -E, --execution-environment is now -c, --context.

All handle_command functions of base commands now take the environment and the context as parameters.

Before:

pub fn handle_command(args: TestCmdArgs) -> anyhow::Result<()> {}

After:

pub fn handle_command(args: TestCmdArgs, env: Environment, ctx: Context) -> anyhow::Result<()> {}

New features

Automatic sourcing of environment files containing environment variables given the value for the -e,--environment argument:
- .env for any set environment,
- .env.{environment} (example: .env.dev) for the non-sensitive configuration,
- .env.{environment}.secrets (example .env.dev.secrets) for the sensitive configuration like password. These
new command docker integrated with the automatic sourcing of environment variable files. It starts a docker compose stack with the naming scheme docker-compose.{env}.yml, env being the shorthand environment name.
TestCmdArgs accepts new parameters --force and --nocapture.
BuildCmdArgs accepts new parameter --release.