shep-deploy 0.3.0

//! `shep-deploy`: a deploy dog for shep.
//!
//! Watches a git branch, builds a release, swaps to it, and rolls back if it
//! does not come up. This file is the binary entry; the deploy sequence
//! itself lives in [`deploy`], and the layout it works over in [`paths`].
//!
//! # Two invocation modes
//!
//! Supervised, the dog is spawned with no argv at all and two environment
//! entries: `SHEP_HOME`, which every path below is joined onto, and
//! `SHEP_DOG_NAME`, which is what shep registered this process as. That mode
//! runs [`poll::run`], which deploys every `watch = "auto"` target whose
//! branch has moved, once an interval, until the process is asked to stop.
//!
//! The loop starts only when `SHEP_DOG_NAME` is set. A bare run without it is
//! a person at a terminal, and gets usage rather than a second loop beside the
//! adopted dog. Whoever needs the loop outside shep sets `SHEP_DOG_NAME`
//! themselves, which also names the handshake. shep sets it for a bare
//! `shep deploy` typed at a terminal too, and that run polls like the dog.
//!
//! The two modes connect differently. The poll loop connects through a client
//! that survives a daemon handover, because it is the only mode that outlives
//! one, and announces the name shep put in `SHEP_DOG_NAME`; a verb connects
//! through a plain client, announces nothing, and exits. See [`poll_forever`]
//! and [`adopted_as`].
//!
//! Run directly, it takes a verb, or `help` for usage:
//!
//! ```text
//! shep-deploy deploy <sheep>
//! shep-deploy deploy <sheep> --watch auto|manual
//! shep-deploy setup <sheep>
//! shep-deploy survey
//! ```
//!
//! `--watch` changes one setting and returns without deploying. See
//! [`deploy::set_watch`]. `survey` reports where every registered sheep
//! stands and touches nothing; see [`survey::survey`]. `setup` takes a
//! sheep over: it builds the deploy tree and its first release, then
//! re-registers the sheep against `current` and removes the instances it
//! replaced. See [`optin::prepare`] and [`optin::cut_over`] - and note that
//! it is the one deploy that may have downtime, and the one that is not
//! verified against the app's readiness probe.

#![forbid(unsafe_code)]

#[cfg(not(unix))]
compile_error!(
    "shep-deploy is Unix only. This is deliberate rather than an oversight: the deploy model is \
     rename(2) over a symlink, the build's privilege drop is a uid and a gid, and both are Unix \
     concepts this crate uses directly rather than through a portability layer. Windows support \
     is planned and will be scoped separately."
);

mod build;
mod config;
mod daemon;
mod deploy;
mod error;
mod exit;
/// Test helpers shared by every module's own `mod tests`; see its own doc for
/// why a binary crate needs this declared here.
#[cfg(test)]
mod fixtures;
mod flockfile;
mod git;
mod lock;
mod optin;
mod paths;
mod poll;
mod restore;
mod retention;
mod roll;
mod route;
mod shared;
mod smit;
mod state;
mod survey;
mod swap;
mod verify;
mod window;

use std::path::PathBuf;
use std::process::ExitCode;

use shep_client::shep_core::paths::ShepPaths;
use shep_client::{Client, LinkState, RECONNECT_MAX_DELAY, ReconnectingClient};
use tokio::signal::unix::{Signal, SignalKind, signal};

use crate::daemon::Live;
use crate::deploy::Outcome;
use crate::error::Error;
use crate::exit::{code_for, code_for_outcome};
use crate::paths::Tree;
use crate::route::{Route, USAGE, route};
use crate::state::{State, Watch};

/// How long the runtime waits for blocking work after the main future ends.
///
/// One second. The only blocking work is git and the artifact copy, both on
/// tokio's pool through `shared::off_thread`; a stop that lands during a
/// fetch must not wait the fetch out, which dropping the runtime without a
/// timeout would do. The git child left behind is bounded by `git_timeout`
/// and finishes or dies on its own after the dog has gone.
const DRAIN: std::time::Duration = std::time::Duration::from_secs(1);

fn main() -> ExitCode {
    // Answers shep's two probes and ends the process when this run is one of
    // them: `--version` with the version and the protocol number this binary
    // was compiled against, `--schema` with the JSON Schema of
    // `config::Section`. It returns on every ordinary run.
    //
    // First, before the runtime is built and before anything is opened,
    // because that is the whole point of the call. `shep restart deploy`
    // probes the binary on disk to see whether it has been upgraded under
    // the running process, and a dog that does not recognise `--version`
    // ignores it and starts doing its ordinary job instead - here, a poll
    // loop against the live shepherd, killed a second or two later. This
    // dog builds and cuts over releases, so "runs briefly and is then
    // killed" is not a harmless overlap.
    //
    // `route` never sees either flag as a result, which is also why neither
    // needs an arm there: a leading `-` is not a sheep name, so both used to
    // land on `USAGE` and exit 2, and shep read that as a dog with no
    // protocol and no schema.
    //
    // The name and the version are passed rather than read inside
    // shep-client because `env!` expands where it is written, and there it
    // would report shep-client's own version.
    shep_client::dogs::probe::<config::Section>(env!("CARGO_PKG_NAME"), env!("CARGO_PKG_VERSION"));

    let args: Vec<String> = std::env::args().skip(1).collect();
    let args: Vec<&str> = args.iter().map(String::as_str).collect();

    // Built by hand rather than through `#[tokio::main]`, which drops the
    // runtime without a timeout and therefore waits for every blocking task
    // to finish: see `DRAIN`.
    let runtime = match tokio::runtime::Builder::new_current_thread()
        .enable_all()
        .build()
    {
        Ok(runtime) => runtime,
        Err(err) => {
            eprintln!("shep-deploy: cannot start a runtime: {err}");
            return ExitCode::from(1);
        }
    };
    let code = runtime.block_on(run(&args));
    runtime.shutdown_timeout(DRAIN);
    code
}

/// The whole program past argument parsing, on the runtime `main` built.
async fn run(args: &[&str]) -> ExitCode {
    let dog = adopted_as(&|key| std::env::var(key).ok());
    let outcome = match route(args, dog.is_some()) {
        Route::Poll => match &dog {
            Some(name) => poll_forever(name).await,
            None => return usage(),
        },
        Route::Help => {
            println!("{USAGE}");
            return ExitCode::SUCCESS;
        }
        // Its own connection, matching every sibling verb, rather than
        // reaching for a `daemon` that does not exist in this scope. It
        // returns an `ExitCode` directly, not a `Result<u8, Error>`, because
        // it never fails outward - see `on_remove`'s own doc.
        Route::OnRemove => return on_remove().await,
        Route::Survey => survey_once().await,
        Route::Deploy(sheep) => deploy_once(sheep).await,
        Route::Setup(sheep) => setup_once(sheep).await,
        Route::Watch { sheep, mode } => set_watch(sheep, mode),
        Route::Usage => return usage(),
    };

    match outcome {
        Ok(code) => ExitCode::from(code),
        Err(err) => {
            eprintln!("shep-deploy: {err}");
            ExitCode::from(code_for(&err))
        }
    }
}

/// Prints [`USAGE`] to stderr and answers the usage exit code.
fn usage() -> ExitCode {
    eprintln!("{USAGE}");
    ExitCode::from(2)
}

/// The environment entry shep puts an adopted dog's registered name in.
///
/// Set by the daemon itself when it spawns the dog, alongside `SHEP_HOME`.
/// Its absence is the signal that this process is NOT a supervised dog, and
/// so does not poll; see [`adopted_as`].
const DOG_NAME: &str = "SHEP_DOG_NAME";

/// The supervised mode: connect as `name`, the dog shep registered this
/// process as, read its own config section, and poll until the process is
/// asked to stop.
///
/// Only [`run`] reaches this, and only when shep started the process; a bare
/// run by hand is routed to usage first.
///
/// Answers 0 for a stop that was asked for, which is the only way out that
/// is not an error. [`poll::run`] itself never returns.
///
/// # Why this one connection is not a [`Client`]
///
/// This is the only invocation that outlives the daemon it connected to. A
/// `shep upgrade` hands the socket to a successor process, and a dog is an
/// ordinary child that keeps running through it, so its connection has to be
/// re-established or the dog is alive and mute for as long as it is left up.
/// [`ReconnectingClient`] is the client that does that, and
/// `connect_as_dog` is the only public door to a handshake that names a dog.
/// That name is what the daemon marks as handshook, and what it needs to
/// restart the right dog when a successor refuses one. A dog connecting
/// without it serves every request correctly and is still recorded as
/// silent, restarted once, and then written off as stale.
///
/// The four one-shot verbs stay on [`Client`], deliberately. Each runs for
/// one command and exits, so there is nothing to reconnect for, and a
/// command claiming to be this dog would have the daemon record a handshake
/// for a process about to exit - and restart the real dog if the command's
/// own handshake were refused.
///
/// # Errors
/// [`Error::Io`] if `$SHEP_HOME` cannot be resolved, [`Error::Connect`] if
/// the shepherd's socket cannot be reached, [`Error::Refused`] if a
/// successor daemon refuses this dog's handshake, and whatever
/// [`config::Reader::open`] returns - a `[dog.<name>]` section that cannot
/// be parsed stops the dog here rather than being ignored, because a dog
/// running on defaults it was not asked for looks exactly like one
/// honouring the config.
///
/// That is the STARTING read. The loop reads the section again every tick,
/// so an operator who edits it does not have to restart the dog, and a
/// later read that fails keeps the last section that parsed instead of
/// ending a deploy that may be in flight. [`config::Reader::refresh`] says
/// why the two answers differ.
///
/// A target's own failure is NOT one of these. It is reported and the loop
/// carries on to the next target; see [`poll::run`].
async fn poll_forever(name: &str) -> Result<u8, Error> {
    let home = shep_home()?;
    // First, and before anything is awaited. See `Stop`.
    let mut stop = Stop::install();

    let socket = socket()?;
    let client = ReconnectingClient::connect_as_dog(&socket, name).await?;
    let daemon = Live::dog(client);
    let config = config::Reader::open(&daemon).await?;

    tokio::select! {
        result = poll::run(&daemon, &home, config) => result.map(|()| 0),
        // Cancels `poll::run` the same way a stop does, and can land inside
        // a deploy for the same reason - see `Stop::arrives`. It costs
        // nothing here that carrying on would not cost anyway: past a
        // refusal every request this loop makes fails, so the deploy it
        // interrupts had already stopped being able to finish.
        refusal = refused(&daemon) => Err(refusal),
        () = stop.arrives() => {
            println!("shep-deploy: stopping");
            Ok(0)
        }
    }
}

/// The name shep registered this process under, or `None` when nothing did.
///
/// This is the handshake identity, and also what [`run`] reads as "supervised":
/// it polls only when this is `Some`. shep sets the variable for a bare
/// `shep deploy` typed at a terminal as well as for the dog it starts, so the
/// two are not told apart here. It has exactly one source. A name reaching the daemon is a claim about which dog
/// to restart when a handshake is refused, so a guessed one - the binary's
/// own stem, the config section's key, a default - gets some other dog
/// restarted for this process's problem. No name at all is a complete and
/// honest answer, and it is the right one for somebody running the binary by
/// hand: they are not a supervised dog, and the daemon should not record a
/// handshake for them.
///
/// Blank is treated as absent. An empty `SHEP_DOG_NAME` names no dog, and
/// announcing it would claim an identity the daemon cannot match. Anything
/// else is passed through verbatim rather than trimmed or validated: the
/// daemon looks this up against its own registry, so the only useful value
/// is the exact one it set.
///
/// Takes its lookup rather than reading the environment directly, matching
/// [`resolved`], so the decision is testable without a process-global
/// `set_var`.
fn adopted_as(env: &dyn Fn(&str) -> Option<String>) -> Option<String> {
    env(DOG_NAME).filter(|name| !name.trim().is_empty())
}

/// Resolves when a successor daemon has refused this dog's handshake, and
/// never otherwise.
///
/// A refusal is terminal by design: the supervisor stops, and every later
/// request fails with no chance of recovering. The loop would carry on
/// ticking through that, deploying nothing and reporting each target's
/// failure once, which reads like a broken remote rather than a dog that
/// can no longer talk to its shepherd at all. Ending instead puts the reason
/// in the dog's log once and hands the process back to shep, which knows
/// from the handshake which dog this was and can start it again when it is
/// upgraded.
///
/// Polled rather than awaited, because [`LinkState`] is a snapshot and
/// `shep-client` offers no wakeup for it. [`RECONNECT_MAX_DELAY`] is the
/// supervisor's own ceiling between reconnect attempts, so a refusal is
/// noticed about one attempt after it happens, and the poll costs one
/// read of a lock per five seconds.
async fn refused(daemon: &Live) -> Error {
    loop {
        if let Some(LinkState::Refused {
            daemon_version,
            message,
        }) = daemon.link()
        {
            return Error::Refused {
                daemon_version,
                message,
            };
        }
        tokio::time::sleep(RECONNECT_MAX_DELAY).await;
    }
}

/// The two signals that mean stop, registered.
///
/// `SIGTERM` is what shep sends a dog it is stopping; `SIGINT` is what a
/// terminal sends somebody who started the dog by hand to watch it.
///
/// # Why this is a type, and why it is installed by a plain function
///
/// [`signal`] inside an `async fn` does not run when the future is created.
/// It runs when the future is first polled, and `tokio::select!` polls its
/// branches in an order that is randomised per process, so a handler
/// installed that way does not exist yet on about half of starts. The
/// window is the first tick, which opens with a `git fetch` - and a
/// `SIGTERM` arriving in it kills the process on the signal's default
/// disposition, with no message and nothing else in this file running.
///
/// A non-async `install` cannot be lazy, so the handlers exist before the
/// loop is polled at all.
struct Stop {
    /// The `SIGTERM` stream, or `None` if it could not be installed.
    term: Option<Signal>,
    /// The `SIGINT` stream, on the same terms.
    interrupt: Option<Signal>,
}

impl Stop {
    /// Installs both handlers now.
    fn install() -> Self {
        Self {
            term: listen(SignalKind::terminate()),
            interrupt: listen(SignalKind::interrupt()),
        }
    }

    /// Resolves when either signal arrives, and never otherwise.
    ///
    /// A stream that is absent or closed is not a request to stop, and
    /// returning for one would print "stopping" for a stop nobody asked
    /// for. The signal keeps its default disposition in that case, so a
    /// `SIGTERM` still ends the process - without the tidy message.
    ///
    /// # What a stop does NOT interrupt
    ///
    /// Not a tick boundary: cancelling [`poll::run`] can land inside a
    /// deploy, which is acceptable and documented there. It used to be
    /// deferred while a `git` call was in flight, because those ran through
    /// blocking `std::process::Command` on the runtime's one thread; they
    /// run on tokio's blocking pool now (`shared::off_thread`), so a stop
    /// is answered during a fetch and `main` gives the pool [`DRAIN`] to
    /// wind down rather than waiting the fetch out.
    async fn arrives(&mut self) {
        let asked = match (&mut self.term, &mut self.interrupt) {
            (Some(term), Some(interrupt)) => tokio::select! {
                arrived = term.recv() => arrived,
                arrived = interrupt.recv() => arrived,
            },
            (Some(only), None) | (None, Some(only)) => only.recv().await,
            (None, None) => None,
        };

        if asked.is_none() {
            core::future::pending().await
        }
    }
}

/// One signal handler, or `None` and a complaint if it cannot be installed.
///
/// Registering fails only on a runtime with no I/O driver, which
/// `#[tokio::main]` never builds. It is reported rather than fatal because
/// a dog that cannot catch `SIGTERM` still deploys perfectly well; it just
/// dies less tidily.
fn listen(kind: SignalKind) -> Option<Signal> {
    match signal(kind) {
        Ok(stream) => Some(stream),
        Err(err) => {
            eprintln!("shep-deploy: cannot listen for {kind:?}: {err}");
            None
        }
    }
}

/// One deploy of `sheep`, reporting what it did.
///
/// # Errors
/// Whatever [`deploy::deploy`] returns, plus [`Error::Io`] if the target's
/// `deploy.toml` cannot be read and [`Error::Connect`] if the shepherd's
/// socket cannot be reached.
async fn deploy_once(sheep: &str) -> Result<u8, Error> {
    let tree = Tree::for_sheep(&shep_home()?, sheep);
    let mut state = State::read(&tree.state_file())?;

    let client = Client::connect(&socket()?).await?;
    let daemon = Live::command(client);

    let config = config::read(&daemon).await?;
    let outcome = deploy::deploy(&daemon, &tree, &mut state, &config).await?;
    match &outcome {
        Outcome::UpToDate => println!("{sheep} is up to date at {}", deployed(&state)),
        Outcome::Deployed { sha } => println!("{sheep} deployed {sha}"),
        Outcome::RolledBack { to, why } => println!("{sheep} rolled back to {to}: {why}"),
    }

    Ok(code_for_outcome(&outcome))
}

/// Takes `sheep` over: builds its deploy tree and first release, then cuts
/// it over to `current`.
///
/// Prints where the sheep now deploys from, because that path is the one
/// thing an operator has no other way to learn - nothing in `shep flock`
/// names it, and it is where every later release lands.
///
/// # Errors
/// [`Error::Io`] if `$SHEP_HOME` cannot be resolved, [`Error::Connect`] if
/// the shepherd's socket cannot be reached, and whatever
/// [`optin::prepare`] or [`optin::cut_over`] return. [`Error::Stranded`] is
/// the one that is returned AFTER a success is printed: the cutover landed
/// and the instances it replaced did not all go.
async fn setup_once(sheep: &str) -> Result<u8, Error> {
    let client = Client::connect(&socket()?).await?;
    let daemon = Live::command(client);

    let config = config::read(&daemon).await?;
    let prepared = optin::prepare(&daemon, &shep_home()?, sheep, &config).await?;
    // Read before `cut_over` consumes `prepared`.
    let current = prepared.tree.current();

    match optin::cut_over(&daemon, prepared).await {
        Ok(sha) => {
            println!("{sheep} now deploys from {}, at {sha}", current.display());
            Ok(0)
        }
        // The cutover landed and only the cleanup did not, so the operator
        // still needs the path - it is the one thing this command tells
        // them that nothing else will - and then the error says what is
        // left to remove by hand.
        Err(err @ Error::Stranded { .. }) => {
            println!("{sheep} now deploys from {}", current.display());
            Err(err)
        }
        Err(err) => Err(err),
    }
}

/// Reports where every registered sheep stands, and touches nothing.
///
/// # Errors
/// [`Error::Io`] if `$SHEP_HOME` cannot be resolved, and whatever
/// [`survey::survey`] returns.
async fn survey_once() -> Result<u8, Error> {
    let client = Client::connect(&socket()?).await?;
    let daemon = Live::command(client);

    print!("{}", survey::survey(&daemon, &shep_home()?).await?);
    Ok(0)
}

/// The on-remove hook. shep runs this argv before forgetting the dog, under
/// a timeout, and proceeds regardless of the outcome.
///
/// ALWAYS exits 0, including when a sheep could not be restored and
/// including when the shepherd cannot be reached at all. An operator asking
/// to remove something is entitled to have it removed, and a nonzero exit
/// here would be a dog arguing about its own uninstallation. Failures are
/// named in the report instead, which is the output shep pipes to them and
/// the only thing they see about any of this.
async fn on_remove() -> ExitCode {
    let Ok(home) = shep_home() else {
        return ExitCode::SUCCESS;
    };
    // Through `socket()` rather than a second `home.join(...)`: joining by
    // hand here was a second place that knew the control socket's layout,
    // and a change to it would have had to land in both without anything
    // catching the drift.
    let Ok(socket) = socket() else {
        return ExitCode::SUCCESS;
    };
    match Client::connect(&socket).await {
        Ok(client) => {
            let daemon = Live::command(client);
            print!("{}", restore::report(&restore::all(&daemon, &home).await));
        }
        // Nothing was restored and nothing was broken. Said plainly,
        // because silence here is indistinguishable from success.
        Err(err) => println!(
            "no sheep were restored: the shepherd could not be reached ({err}). Any sheep this \
             dog took over is still running from its deploy tree under {}.",
            home.join("deploy").display()
        ),
    }
    ExitCode::SUCCESS
}

/// Sets `sheep`'s watch mode and returns, without deploying.
///
/// # Errors
/// [`Error::Config`] if `mode` is neither `auto` nor `manual`, if `auto` was
/// asked for on a tree the cutover never landed on, or if `deploy.toml` does
/// not parse or fails validation - see [`deploy::set_watch`]. [`Error::Io`]
/// if it cannot be read or written.
fn set_watch(sheep: &str, mode: &str) -> Result<u8, Error> {
    let watch = match mode {
        "auto" => Watch::Auto,
        "manual" => Watch::Manual,
        other => {
            return Err(Error::Config(format!(
                "--watch takes auto or manual, not {other:?}"
            )));
        }
    };

    let tree = Tree::for_sheep(&shep_home()?, sheep);
    // Both the record before and the record after come back from the write:
    // `set_watch` reads immediately before writing, and a read made here
    // would be the older copy the "was already" line must not be true of.
    let (was, state) = deploy::set_watch(&tree, watch)?;

    if was == watch {
        println!("{sheep} was already watch = {}", named(watch));
    } else {
        println!(
            "{sheep} watch: {} -> {}, still deployed at {}",
            named(was),
            named(watch),
            deployed(&state)
        );
    }

    Ok(0)
}

/// `$SHEP_HOME`, absolute.
///
/// Absolute at the point of reading, deliberately: every path this crate
/// builds is joined onto this one, and [`crate::swap::point_at`] writes some
/// of them into symlink targets. A symlink target is resolved against the
/// symlink's own directory rather than this process's working directory, so
/// a relative `SHEP_HOME` would produce links that dangle silently -
/// exactly the failure [`crate::shared::link_into`] canonicalises to avoid,
/// one module over. [`std::path::absolute`] rather than `canonicalize`
/// because this may run before the tree exists, and because resolving
/// through symlinks in `$SHEP_HOME` itself is not this dog's business.
///
/// # Errors
/// [`Error::Io`] if the path cannot be made absolute, which needs the
/// current directory to be readable.
fn shep_home() -> Result<PathBuf, Error> {
    absolute(resolved().home)
}

/// The shepherd's control socket, from the same layout as [`shep_home`].
///
/// Reads `ShepPaths`'s own `socket` field rather than joining `run/shep.sock`
/// onto the home. Both spellings agree today, and the point is that only one
/// of them is shep's to change: the layout belongs to `shep_core::paths`, and
/// a copy here is a second source of truth that drifts silently the day shep
/// moves the socket.
///
/// # Errors
/// As [`shep_home`].
fn socket() -> Result<PathBuf, Error> {
    absolute(resolved().socket)
}

/// The shep layout, as this process's environment spells it.
fn resolved() -> ShepPaths {
    let home = std::env::home_dir().unwrap_or_default();
    ShepPaths::resolve(&|key| std::env::var(key).ok(), &home)
}

/// One path from [`resolved`], made absolute.
///
/// # Errors
/// [`Error::Io`] if the path cannot be made absolute, which needs the current
/// directory to be readable.
fn absolute(path: PathBuf) -> Result<PathBuf, Error> {
    std::path::absolute(&path).map_err(|source| Error::Io { path, source })
}

/// The sha a target is deployed at, for a message.
fn deployed(state: &State) -> &str {
    state.deployed.as_deref().unwrap_or("nothing yet")
}

/// A [`Watch`] as an operator spells it on the command line.
const fn named(watch: Watch) -> &'static str {
    match watch {
        Watch::Auto => "auto",
        Watch::Manual => "manual",
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    /// fails if a supervised dog stops announcing the name shep gave it.
    ///
    /// That name is the whole of the bug this function exists for. Without
    /// it in the handshake the daemon never marks the dog handshook: it
    /// connects, serves every request correctly, and is rendered `silent`,
    /// restarted once after five seconds, then written off as stale and
    /// never started again. Everything an operator can see says the dog is
    /// working.
    #[test]
    fn a_supervised_dog_announces_the_name_shep_gave_it() {
        let env = |key: &str| (key == DOG_NAME).then(|| "log-rotate".to_owned());
        assert_eq!(adopted_as(&env).as_deref(), Some("log-rotate"));
    }

    /// fails if a process nobody supervised claims to be a dog anyway.
    ///
    /// No `SHEP_DOG_NAME` means shep did not spawn this, so there is no dog
    /// to be, and the bare command prints usage instead of polling. The name
    /// must never be guessed here - from the binary's stem, from the
    /// `[dog.<name>]` section, from a default - because it is what the daemon
    /// restarts on a refused handshake, and a guess gets some other dog
    /// restarted for a hand run's problem.
    #[test]
    fn a_hand_run_announces_no_name_at_all() {
        assert_eq!(adopted_as(&|_| None), None);
        // Blank names no dog either, and an identity the daemon cannot
        // match is worse than none: it is a claim.
        assert_eq!(adopted_as(&|_| Some(String::new())), None);
        assert_eq!(adopted_as(&|_| Some("   ".to_owned())), None);
    }

    /// fails if the name stops being read from `SHEP_DOG_NAME` specifically.
    /// `SHEP_NAME` and `SHEP_INSTANCE` sit beside it in a dog's environment
    /// and neither is the registered dog name.
    #[test]
    fn only_shep_dog_name_is_read() {
        let env = |key: &str| (key == "SHEP_NAME").then(|| "web".to_owned());
        assert_eq!(adopted_as(&env), None);
    }

    /// fails if the stop handlers stop being installed eagerly. An
    /// `async fn` that calls `signal()` registers on its first poll, not on
    /// creation, and `select!` polls in a randomised order - so lazily
    /// installed handlers do not exist during the first tick on about half
    /// of starts, and that tick opens with a `git fetch`. A `SIGTERM` in
    /// that window kills the dog on the default disposition.
    ///
    /// `install` being a plain function is what makes it eager; this is
    /// what makes it work.
    #[tokio::test]
    async fn both_stop_handlers_are_installed_up_front() {
        let stop = Stop::install();
        assert!(stop.term.is_some(), "SIGTERM");
        assert!(stop.interrupt.is_some(), "SIGINT");
    }
}