boxmaker/docs/plans/M5a/files/crates/memoryd/src/recall.rs

//! `recall`: search the index and answer with tagged hits (M5a spec, section 5). The query
//! reaches SQLite only as quoted terms, so no FTS5 operator from the model is ever parsed.

use serde::Serialize;

use proto::{DataClass, RecallArgs, Timestamp};
use rusqlite::params;

use crate::files::Store;
use crate::index::{Index, class_name, class_of};

/// The most characters of one hit's text.
pub const SNIPPET: usize = 600;

/// The answer to the model, and the label the result carries.
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct Recalled {
    pub text: String,
    pub class: DataClass,
    pub untrusted: bool,
}

/// The FTS5 query for `query`: every run of letters and digits, double-quoted, joined by `OR`.
/// `None` when there is no such run.
pub fn terms(query: &str) -> Option<String> {
    // Split `query` on every char that is not alphanumeric (`!c.is_alphanumeric()`), drop the empty
    // pieces, wrap each in double quotes (format!("\"{t}\"")) and join them with " OR ". None when
    // there is no piece. So `a NEAR(b` is `"a" OR "NEAR" OR "b"`: no FTS5 operator from the model
    // is ever parsed.
    todo!()
}

/// `2026-09-20 14:02 UTC` from `2026-09-20T14:02:00.000Z` or `2026-09-20T14:02Z`.
fn when(time: &str) -> String {
    // The first 10 characters (the day) and characters 11 to 16 (the clock), as format!("{day}
    // {clock} UTC"): both 2026-09-20T14:02:00.000Z and 2026-09-20T14:02Z give 2026-09-20 14:02 UTC.
    // When either `get` is None, the time unchanged.
    todo!()
}

fn day(time: &str) -> &str {
    // The first 10 characters of `time` (`get(..10)`), or all of it when shorter.
    todo!()
}

struct Hit {
    kind: String,
    session: String,
    epoch: i64,
    record: i64,
    time: String,
    role: String,
    class: DataClass,
    untrusted: bool,
    text: String,
}

impl Hit {
    /// `identity`, `memory` or `memory-tainted`.
    fn tag(&self) -> &'static str {
        // kind "moved" -> "identity"; "dropped" -> "memory"; any other kind: "memory-tainted" when
        // untrusted, else "memory".
        todo!()
    }

    fn header(&self, n: usize) -> String {
        // `from` by kind: "moved" -> format!("moved {} from {}", day(time), role); "dropped" -> the
        // same with "dropped"; "noted" -> format!("note, {}, session {}", when(time), session); a
        // session record -> format!("session {}, {}, {said}", session, when(time)) where said is
        // "the owner said" for role "owner", "the assistant called" for "call", and "the assistant
        // said" otherwise. Return format!("[{n}] {} · {from} · {}", self.tag(),
        // class_name(self.class)).
        todo!()
    }
}

#[derive(Serialize)]
struct LoggedHit<'a> {
    kind: &'a str,
    session: &'a str,
    epoch: i64,
    record: i64,
}

#[derive(Serialize)]
struct Logged<'a> {
    time: String,
    session: &'a str,
    query: &'a str,
    hits: Vec<LoggedHit<'a>>,
}

/// Search for `args` on behalf of `session`, log the recall, and answer.
pub fn recall(
    index: &Index,
    store: &Store,
    args: &RecallArgs,
    session: &str,
    now: Timestamp,
) -> Result<Recalled, String> {
    // 1. `args.check()?`.
    // 2. hits: `terms(&args.query)` None -> none; Some(fts) -> `search(index, &fts, args.k)?`.
    // 3. Log it: a Logged { time: now.to_rfc3339(), session, query, hits: a LoggedHit (kind,
    //    session, epoch, record) per hit } as one JSON line appended to
    //    format!("recalls/{month}.jsonl") (month: the first 7 characters of now.to_rfc3339()),
    //    after `create_dir_all(store.path("recalls"))`.
    // 4. No hits: Ok(Recalled { text: format!("(no matches for \"{}\")", args.query), class
    //    Private, untrusted false }).
    // 5. Otherwise, for each hit with i from 0: class = class.max(its class) starting at Private;
    //    untrusted |= its tag is "memory-tainted"; its text cut to its first SNIPPET characters;
    //    the item is format!("{}\n{text}", hit.header(i + 1)). Ok(Recalled { the items joined by
    //    newlines, class, untrusted }).
    todo!()
}

fn search(index: &Index, fts: &str, k: u64) -> Result<Vec<Hit>, String> {
    // `index.conn().prepare("SELECT s.kind, s.session, s.epoch, s.record, s.time, s.role, s.class,
    // s.untrusted, snippet(hits, 0, '', '', '…', 64) FROM hits JOIN sources s ON s.rowid =
    // hits.rowid WHERE hits MATCH ?1 ORDER BY bm25(hits) LIMIT ?2")`, then `query_map(params![fts,
    // limit], ..)` building a Hit from columns 0 to 8 in that order (the class through `class_of`);
    // limit is k as an i64 (`i64::try_from`, 10 when it does not fit). Collect into Vec<Hit>. Every
    // rusqlite error as format!("index: {e}").
    todo!()
}