hyperhive/hive-c0re/src/broker.rs

//! Sqlite-backed message broker. Survives `hive-c0re` restart, and taps every
//! send/recv onto a broadcast channel so the dashboard can stream it.

use std::path::Path;
use std::sync::Mutex;
use std::time::{SystemTime, UNIX_EPOCH};

use anyhow::{Context, Result};
use hive_sh4re::{InboxRow, Message};
use rusqlite::{Connection, OptionalExtension, params};
use serde::Serialize;
use tokio::sync::broadcast;

const SCHEMA: &str = r"
CREATE TABLE IF NOT EXISTS messages (
    id           INTEGER PRIMARY KEY AUTOINCREMENT,
    sender       TEXT    NOT NULL,
    recipient    TEXT    NOT NULL,
    body         TEXT    NOT NULL,
    sent_at      INTEGER NOT NULL,
    delivered_at INTEGER
);
CREATE INDEX IF NOT EXISTS idx_messages_undelivered
    ON messages (recipient, id) WHERE delivered_at IS NULL;
";

/// Capacity of the live event channel. Slow subscribers (e.g. an idle browser)
/// may drop events past this; we send a `lagged` notice in their stream.
const EVENT_CHANNEL: usize = 256;

#[derive(Debug, Clone, Serialize)]
#[serde(rename_all = "snake_case", tag = "kind")]
pub enum MessageEvent {
    Sent {
        from: String,
        to: String,
        body: String,
        at: i64,
    },
    Delivered {
        from: String,
        to: String,
        body: String,
        at: i64,
    },
}

pub struct Broker {
    conn: Mutex<Connection>,
    events: broadcast::Sender<MessageEvent>,
}

impl Broker {
    pub fn open(path: &Path) -> Result<Self> {
        if let Some(parent) = path.parent() {
            std::fs::create_dir_all(parent)
                .with_context(|| format!("create db parent {}", parent.display()))?;
        }
        let conn =
            Connection::open(path).with_context(|| format!("open broker db {}", path.display()))?;
        conn.execute_batch(SCHEMA).context("apply broker schema")?;
        let (events, _) = broadcast::channel(EVENT_CHANNEL);
        Ok(Self {
            conn: Mutex::new(conn),
            events,
        })
    }

    pub fn subscribe(&self) -> broadcast::Receiver<MessageEvent> {
        self.events.subscribe()
    }

    pub fn send(&self, message: &Message) -> Result<()> {
        let conn = self.conn.lock().unwrap();
        conn.execute(
            "INSERT INTO messages (sender, recipient, body, sent_at) VALUES (?1, ?2, ?3, ?4)",
            params![message.from, message.to, message.body, now_unix()],
        )?;
        drop(conn);
        let _ = self.events.send(MessageEvent::Sent {
            from: message.from.clone(),
            to: message.to.clone(),
            body: message.body.clone(),
            at: now_unix(),
        });
        Ok(())
    }

    /// Latest `limit` messages addressed to `recipient`, newest-first.
    /// Includes delivered + undelivered alike — used for the operator
    /// inbox view on the dashboard. Caller decides what to show.
    pub fn recent_for(&self, recipient: &str, limit: u64) -> Result<Vec<InboxRow>> {
        let conn = self.conn.lock().unwrap();
        let limit_i = i64::try_from(limit.min(i64::MAX as u64)).unwrap_or(i64::MAX);
        let mut stmt = conn.prepare(
            "SELECT id, sender, body, sent_at
               FROM messages
              WHERE recipient = ?1
              ORDER BY id DESC
              LIMIT ?2",
        )?;
        let rows = stmt.query_map(params![recipient, limit_i], |row| {
            Ok(InboxRow {
                id: row.get(0)?,
                from: row.get(1)?,
                body: row.get(2)?,
                at: row.get(3)?,
            })
        })?;
        rows.collect::<rusqlite::Result<Vec<_>>>()
            .map_err(Into::into)
    }

    /// Number of undelivered messages addressed to `recipient`. Non-mutating
    /// — used by the harness to surface "N unread" in tool-result status
    /// lines without popping the queue.
    pub fn count_pending(&self, recipient: &str) -> Result<u64> {
        let conn = self.conn.lock().unwrap();
        let n: i64 = conn.query_row(
            "SELECT COUNT(*) FROM messages
              WHERE recipient = ?1 AND delivered_at IS NULL",
            params![recipient],
            |row| row.get(0),
        )?;
        Ok(u64::try_from(n.max(0)).unwrap_or(0))
    }

    /// Long-poll variant of `recv`: returns immediately if there's a
    /// pending message; otherwise waits up to `timeout` for the broker to
    /// emit a `Sent { to: recipient }` event, then retries the pop. Lets
    /// agents react to new mail without polling their socket on a fixed
    /// interval.
    pub async fn recv_blocking(
        &self,
        recipient: &str,
        timeout: std::time::Duration,
    ) -> Result<Option<Message>> {
        if let Some(m) = self.recv(recipient)? {
            return Ok(Some(m));
        }
        let mut rx = self.subscribe();
        let deadline = tokio::time::Instant::now() + timeout;
        loop {
            let Some(remaining) = deadline.checked_duration_since(tokio::time::Instant::now())
            else {
                return Ok(None);
            };
            match tokio::time::timeout(remaining, rx.recv()).await {
                Err(_) => return Ok(None),
                // Channel lagged or closed — fall back to a single direct
                // pop (in case we missed our notification while behind).
                Ok(Err(_)) => return self.recv(recipient),
                Ok(Ok(MessageEvent::Sent { to, .. })) if to == recipient => {
                    if let Some(m) = self.recv(recipient)? {
                        return Ok(Some(m));
                    }
                    // Lost a race (concurrent recv elsewhere). Keep waiting.
                }
                Ok(Ok(_)) => {}
            }
        }
    }

    /// Delete delivered messages older than `older_than_secs`. Undelivered
    /// rows are always kept regardless of age — those are still in flight
    /// from the broker's POV. Returns the number of rows removed.
    pub fn vacuum_delivered(&self, older_than_secs: i64) -> Result<u64> {
        let cutoff = now_unix() - older_than_secs;
        let conn = self.conn.lock().unwrap();
        let n = conn.execute(
            "DELETE FROM messages
              WHERE delivered_at IS NOT NULL
                AND delivered_at < ?1",
            params![cutoff],
        )?;
        Ok(u64::try_from(n).unwrap_or(0))
    }

    pub fn recv(&self, recipient: &str) -> Result<Option<Message>> {
        let conn = self.conn.lock().unwrap();
        let row: Option<(i64, String, String, String)> = conn
            .query_row(
                "SELECT id, sender, recipient, body
                   FROM messages
                  WHERE recipient = ?1 AND delivered_at IS NULL
                  ORDER BY id ASC
                  LIMIT 1",
                params![recipient],
                |row| Ok((row.get(0)?, row.get(1)?, row.get(2)?, row.get(3)?)),
            )
            .optional()?;
        let Some((id, from, to, body)) = row else {
            return Ok(None);
        };
        conn.execute(
            "UPDATE messages SET delivered_at = ?1 WHERE id = ?2",
            params![now_unix(), id],
        )?;
        drop(conn);
        let _ = self.events.send(MessageEvent::Delivered {
            from: from.clone(),
            to: to.clone(),
            body: body.clone(),
            at: now_unix(),
        });
        Ok(Some(Message { from, to, body }))
    }
}

fn now_unix() -> i64 {
    SystemTime::now()
        .duration_since(UNIX_EPOCH)
        .ok()
        .and_then(|d| i64::try_from(d.as_secs()).ok())
        .unwrap_or(0)
}