simulator_api/reroute_report/

benchmark.rs

//! The reroute benchmark: how each re-quoted route did against the one L1 landed, as win/loss
//! counts, volume-weighted bps, and the venues both sides routed through.
//!
//! Population is the adapter's to select — landed reroutes whose original also landed, with both
//! realized amounts present and the two mints distinct. [`collect`] adds only what needs the whole
//! population to decide: a signature seen twice is an ambiguous pairing and drops wholesale.
//! ExactOut legs are measured input-side (`amount` is output-token-denominated and outputs tie by
//! construction, so a win is paying less major input), which leaves a non-major-input ExactOut leg
//! unpriceable — those are skipped.
//!
//! Two deliberate differences from the pre-typed-plan reader, both forced by the typing. Hop
//! amounts arrive as strings, so a float-typed string like `"1.5e3"` truncates to an integer
//! instead of voiding the plan — the JSON number/string distinction died with the DOM parse;
//! genuinely unparseable text still voids. And an explicit `"inAmount": null` used to void the
//! whole plan where an absent key merely contributed zero — an `Option<String>` cannot tell the
//! two apart, so both now contribute zero.

use std::{
    cmp::Ordering,
    collections::{BTreeMap, HashMap, HashSet, hash_map::Entry},
};

use serde::Serialize;
use solana_address::Address;

use super::{
    fmt::{commas_u, median},
    prices::Prices,
    record::{LegRecord, SwapMode},
};
use crate::{mints, route_plan::RoutePlan};

/// Venues under this USD exposure on both sides are dust, hidden from the table.
pub const VENUE_DUST_USD: f64 = 1e4;
/// Legs beyond this are treated as pricing artifacts, not routing outcomes.
pub const OUTLIER_BPS: f64 = 2000.0;

#[derive(Default, Debug)]
pub struct LoadStats {
    pub dropped_multi_row: usize,
}

/// Row order is preserved so float accumulation stays deterministic across runs. A signature seen
/// more than once is an ambiguous multi-swap transaction — pairing is by signature — so its rows
/// drop wholesale, and a skipped ExactOut row never claims a signature it could make ambiguous.
pub fn collect(rows: Vec<LegRecord>) -> (Vec<LegRecord>, LoadStats) {
    let mut kept: Vec<LegRecord> = Vec::new();
    let mut first_row: HashMap<String, usize> = HashMap::new();
    let mut dropped: HashSet<usize> = HashSet::new();
    for row in rows {
        if row.swap_mode == SwapMode::ExactOut && !mints::is_major(&row.input_mint) {
            continue;
        }
        match first_row.entry(row.original_signature.clone()) {
            Entry::Vacant(slot) => {
                slot.insert(kept.len());
                kept.push(row);
            }
            Entry::Occupied(slot) => {
                dropped.insert(*slot.get());
            }
        }
    }
    let stats = LoadStats {
        dropped_multi_row: dropped.len(),
    };
    (
        kept.into_iter()
            .enumerate()
            .filter(|(index, _)| !dropped.contains(index))
            .map(|(_, row)| row)
            .collect(),
        stats,
    )
}

/// Total `inAmount` for `mint` across the plan's hops. Amounts are strings on the wire; an integer
/// parse is preferred and a float truncates, matching how numeric JSON amounts were read before
/// plans were typed. A string that parses as neither voids the plan.
fn plan_input(plan: &RoutePlan, mint: &Address) -> Option<i64> {
    let mint = mint.to_string();
    let mut total = 0i64;
    for hop in plan.hops() {
        if hop.input_mint() != Some(mint.as_str()) {
            continue;
        }
        // An absent inAmount contributes zero; a present one that reads as no number at all voids
        // the whole plan.
        total += match hop.swap_info.in_amount.as_deref() {
            None => 0,
            Some(amount) => amount
                .parse::<i64>()
                .ok()
                .or_else(|| amount.parse::<f64>().ok().map(|float| float as i64))?,
        };
    }
    (total > 0).then_some(total)
}

/// `[(label, weight)]`; sequential hops each carry the full amount, splits weighted by percent.
/// Unlabeled hops keep their weight under `None` so dropped flow stays visible.
fn venue_weights(plan: &RoutePlan) -> Vec<(Option<&str>, f64)> {
    plan.hops()
        .iter()
        .map(|hop| (hop.label(), hop.percent as f64 / 100.0))
        .collect()
}

/// One measured leg: how our route compared and what it was worth.
pub struct Measurement {
    pub outcome: Ordering,
    pub l1_usd: f64,
    pub our_usd: f64,
}

/// ExactIn compares delivered output (buys valued at spend); ExactOut compares major input paid
/// for the identical output. `None` = unpriceable or a |bps| > [`OUTLIER_BPS`] dollar-outlier.
pub fn measure(record: &LegRecord, prices: &Prices) -> Option<Measurement> {
    match record.swap_mode {
        SwapMode::ExactOut => measure_exact_out(record, prices),
        SwapMode::ExactIn => measure_exact_in(record, prices),
    }
}

fn measure_exact_out(record: &LegRecord, prices: &Prices) -> Option<Measurement> {
    let mint = &record.input_mint;
    let ours = plan_input(record.route_plan.as_ref()?, mint)?;
    let l1 = plan_input(record.original_route_plan.as_ref()?, mint)?;
    let l1_usd = prices.usd(mint, l1 as f64)?;
    // Same output either way, so credit the input saved: l1 + (l1 - ours).
    let our_usd = prices.usd(mint, (2 * l1 - ours) as f64)?;
    if l1_usd <= 0.0 || (our_usd / l1_usd - 1.0).abs() * 1e4 > OUTLIER_BPS {
        return None;
    }
    Some(Measurement {
        outcome: l1.cmp(&ours),
        l1_usd,
        our_usd,
    })
}

fn measure_exact_in(record: &LegRecord, prices: &Prices) -> Option<Measurement> {
    let ours = record.realized_output_amount.unwrap_or(0) as i64;
    let l1 = record.original_realized_output_amount.unwrap_or(0) as i64;
    let bps = (ours as f64 / l1 as f64 - 1.0) * 1e4;
    let scaled = |usd: f64| usd * (1.0 + bps / 1e4);
    // Majors first, so a leg with a priced side is valued on that side rather than through the
    // exotic map the last two arms fall back to.
    let (l1_usd, our_usd) = if mints::is_major(&record.output_mint) {
        (
            prices.usd(&record.output_mint, l1 as f64),
            prices.usd(&record.output_mint, ours as f64),
        )
    } else if mints::is_major(&record.input_mint) {
        let spend = prices.usd(&record.input_mint, record.amount as f64);
        (spend, spend.map(scaled))
    } else if let Some(l1_usd) = prices.usd(&record.output_mint, l1 as f64) {
        (Some(l1_usd), prices.usd(&record.output_mint, ours as f64))
    } else {
        let spend = prices.usd(&record.input_mint, record.amount as f64)?;
        (Some(spend), Some(scaled(spend)))
    };
    let (l1_usd, our_usd) = (l1_usd?, our_usd?);
    if l1_usd <= 0.0 || bps.abs() > OUTLIER_BPS {
        return None;
    }
    Some(Measurement {
        outcome: ours.cmp(&l1),
        l1_usd,
        our_usd,
    })
}

#[derive(Default, Clone, Serialize)]
pub struct SideTotals {
    pub vol: f64,
    pub net: f64,
}

#[derive(Default, Clone, Serialize)]
pub struct VenueTotals {
    pub us: f64,
    pub l1: f64,
    pub net: f64,
}

/// Accumulated metrics for one range or a cross-range aggregate.
#[derive(Default, Clone, Serialize)]
pub struct RangeReport {
    pub n: u64,
    pub better: u64,
    pub same: u64,
    pub worse: u64,
    pub vol: f64,
    pub net: f64,
    pub vol_better: f64,
    pub vol_same: f64,
    pub vol_worse: f64,
    pub bps: Vec<f64>,
    pub majors: SideTotals,
    pub exotic: SideTotals,
    pub venues: BTreeMap<String, VenueTotals>,
}

impl RangeReport {
    pub fn from_records(records: &[LegRecord], prices: &Prices) -> Self {
        let mut report = Self::default();
        for record in records {
            if let Some(measurement) = measure(record, prices) {
                report.record(record, &measurement);
            }
        }
        report
    }

    fn record(&mut self, leg: &LegRecord, measurement: &Measurement) {
        let delta = measurement.our_usd - measurement.l1_usd;
        self.n += 1;
        let (count, vol) = match measurement.outcome {
            Ordering::Greater => (&mut self.better, &mut self.vol_better),
            Ordering::Equal => (&mut self.same, &mut self.vol_same),
            Ordering::Less => (&mut self.worse, &mut self.vol_worse),
        };
        *count += 1;
        *vol += measurement.l1_usd;
        self.vol += measurement.l1_usd;
        self.net += delta;
        if measurement.l1_usd > 0.0 {
            self.bps.push(delta / measurement.l1_usd * 1e4);
        }
        let side = if mints::trades_only_majors(&leg.input_mint, &leg.output_mint) {
            &mut self.majors
        } else {
            &mut self.exotic
        };
        side.vol += measurement.l1_usd;
        side.net += delta;
        self.attribute_venues(leg, measurement.l1_usd, delta);
    }

    /// Split the leg's USD across its route's venues, per-leg-normalized so every leg contributes
    /// its USD exactly once, for both our winning route and L1's landed one.
    fn attribute_venues(&mut self, leg: &LegRecord, l1_usd: f64, delta: f64) {
        for (ours, plan) in [(true, &leg.route_plan), (false, &leg.original_route_plan)] {
            let Some(plan) = plan else {
                continue;
            };
            let weights = venue_weights(plan);
            let total: f64 = weights.iter().map(|(_, weight)| weight).sum();
            let total = if total == 0.0 { 1.0 } else { total };
            for (label, weight) in weights {
                let label = label.unwrap_or("(unlabeled)");
                let entry = self.venues.entry(label.to_string()).or_default();
                let share = weight / total;
                if ours {
                    entry.us += l1_usd * share;
                    entry.net += delta * share;
                } else {
                    entry.l1 += l1_usd * share;
                }
            }
        }
    }

    pub fn merge(&mut self, other: &Self) {
        self.n += other.n;
        self.better += other.better;
        self.same += other.same;
        self.worse += other.worse;
        self.vol += other.vol;
        self.net += other.net;
        self.vol_better += other.vol_better;
        self.vol_same += other.vol_same;
        self.vol_worse += other.vol_worse;
        self.bps.extend_from_slice(&other.bps);
        self.majors.vol += other.majors.vol;
        self.majors.net += other.majors.net;
        self.exotic.vol += other.exotic.vol;
        self.exotic.net += other.exotic.net;
        for (venue, totals) in &other.venues {
            let entry = self.venues.entry(venue.clone()).or_default();
            entry.us += totals.us;
            entry.l1 += totals.l1;
            entry.net += totals.net;
        }
    }

    pub fn vw_bps(&self) -> f64 {
        if self.vol == 0.0 {
            0.0
        } else {
            self.net / self.vol * 1e4
        }
    }

    pub fn render(&self, name: &str) -> String {
        if self.vol == 0.0 {
            return format!("### {name}: no measurable legs\n");
        }
        let mut lines = self.summary_lines(name);
        lines.push(String::new());
        lines.extend(self.venue_table());
        lines.push(String::new());
        lines.join("\n")
    }

    fn summary_lines(&self, name: &str) -> Vec<String> {
        let mut bps = self.bps.clone();
        let (q1_at, q3_at) = (bps.len() / 4, 3 * bps.len() / 4);
        let med = median(&mut bps);
        let mut lines = vec![
            format!("### {name}"),
            format!(
                "vw vs L1: **{:+.2} bps** (net {}$ on ${})",
                self.vw_bps(),
                commas_signed(self.net),
                commas(self.vol)
            ),
            format!(
                "b/s/w: {} / {} / {}  (by volume {:.1} / {:.1} / {:.1}%)",
                commas_u(self.better),
                commas_u(self.same),
                commas_u(self.worse),
                self.vol_better / self.vol * 100.0,
                self.vol_same / self.vol * 100.0,
                self.vol_worse / self.vol * 100.0
            ),
            format!(
                "per-leg bps: median {med:+.2}, IQR [{:+.2}, {:+.2}]",
                bps[q1_at], bps[q3_at]
            ),
        ];
        for (label, side) in [
            ("majors<->majors", &self.majors),
            ("exotic-involved", &self.exotic),
        ] {
            if side.vol > 0.0 {
                lines.push(format!(
                    "{label}: {:+.2} bps (net {}$ on ${})",
                    side.net / side.vol * 1e4,
                    commas_signed(side.net),
                    commas(side.vol)
                ));
            }
        }
        lines
    }

    fn venue_table(&self) -> Vec<String> {
        let mut lines = vec![
            "| Venue | our share | L1 share | vw bps (our legs) |".to_string(),
            "|---|---|---|---|".to_string(),
        ];
        let total_us: f64 = self.venues.values().map(|e| e.us).sum::<f64>().max(1.0);
        let total_l1: f64 = self.venues.values().map(|e| e.l1).sum::<f64>().max(1.0);
        let mut ranked: Vec<_> = self.venues.iter().collect();
        ranked.sort_by(|a, b| b.1.us.total_cmp(&a.1.us));
        for (venue, e) in ranked {
            if e.us < VENUE_DUST_USD && e.l1 < VENUE_DUST_USD {
                continue;
            }
            let vw = if e.us == 0.0 { 0.0 } else { e.net / e.us * 1e4 };
            lines.push(format!(
                "| {venue} | {:.1}% | {:.1}% | {vw:+.2} |",
                e.us / total_us * 100.0,
                e.l1 / total_l1 * 100.0
            ));
        }
        lines
    }
}

fn commas(x: f64) -> String {
    commas_u(x.round().abs() as u64)
}

fn commas_signed(x: f64) -> String {
    format!("{}{}", if x < 0.0 { "-" } else { "+" }, commas(x))
}

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

    use super::*;
    use crate::{
        reroute_report::{
            Samples,
            testing::{hop, plan, record_of},
        },
        route_plan::RouteHop,
    };

    #[rstest]
    #[case::rounds_up("+1,235", 1_234.6)]
    #[case::negative("-1,235", -1_234.6)]
    fn signed_commas_carry_the_sign(#[case] expected: &str, #[case] x: f64) {
        assert_eq!(commas_signed(x), expected);
    }

    fn hop_of(input_mint: &Address, in_amount: u64) -> RouteHop {
        hop(
            None,
            "Venue",
            &input_mint.to_string(),
            &mints::USDT.to_string(),
            in_amount,
        )
    }

    /// The wire carries amounts as text, which a fixture with a `u64` cannot express.
    fn amount_hop(in_amount: Option<&str>) -> RouteHop {
        let mut hop = hop_of(&mints::USDC, 0);
        hop.swap_info.in_amount = in_amount.map(str::to_string);
        hop
    }

    #[rstest]
    #[case::integer_amount(Some("1000"), Some(1_000))]
    #[case::float_amount_truncates(Some("1000.9"), Some(1_000))]
    #[case::float_string_truncates(Some("1.5e3"), Some(1_500))]
    #[case::unparseable_voids_the_plan(Some("abc"), None)]
    #[case::an_absent_amount_alone_sums_to_nothing(None, None)]
    fn plan_input_reads_integer_and_float_amounts(
        #[case] amount: Option<&str>,
        #[case] expected: Option<i64>,
    ) {
        let plan = plan(vec![amount_hop(amount)]).expect("a plan");
        assert_eq!(plan_input(&plan, &mints::USDC), expected);
    }

    #[test]
    fn plan_input_sums_only_the_asked_mints_hops() {
        let plan = plan(vec![
            hop_of(&mints::USDC, 300),
            hop_of(&Address::default(), 999),
            hop_of(&mints::USDC, 700),
        ])
        .expect("a plan");
        assert_eq!(plan_input(&plan, &mints::USDC), Some(1_000));

        // Deserialization bypasses `RoutePlan::new`, so a plan off the wire can still be empty.
        let empty: RoutePlan = serde_json::from_str("[]").expect("an empty plan");
        assert_eq!(plan_input(&empty, &mints::USDC), None);
    }

    #[test]
    fn an_absent_amount_contributes_zero_without_voiding() {
        let plan = plan(vec![amount_hop(None), hop_of(&mints::USDC, 700)]).expect("a plan");
        assert_eq!(plan_input(&plan, &mints::USDC), Some(700));
    }

    /// Fixes SOL at $100 from the stream's own swaps; the stable legs below price off decimals
    /// alone, so nothing else needs deriving.
    fn prices() -> Prices {
        let mut samples = Samples::default();
        for n in 1..=3u64 {
            samples.push(&LegRecord {
                original_realized_output_amount: Some(n * 100_000_000),
                ..record_of(mints::SOL, mints::USDC, n * 1_000_000_000, None, None, None)
            });
        }
        samples.prices().expect("anchored")
    }

    fn stable_record(mode: SwapMode, ours: u64, l1: u64) -> LegRecord {
        LegRecord {
            swap_mode: mode,
            realized_output_amount: Some(ours),
            original_realized_output_amount: Some(l1),
            route_plan: plan(vec![hop_of(&mints::USDC, ours)]),
            original_route_plan: plan(vec![hop_of(&mints::USDC, l1)]),
            ..record_of(mints::USDC, mints::USDT, 1_000_000, None, None, None)
        }
    }

    #[test]
    fn exact_out_credits_the_input_saved() {
        let m = measure(
            &stable_record(SwapMode::ExactOut, 999_000, 1_000_000),
            &prices(),
        )
        .expect("measurable");
        assert_eq!(m.outcome, Ordering::Greater);
        assert!((m.l1_usd - 1.0).abs() < 1e-9);
        assert!((m.our_usd - 1.001).abs() < 1e-9);
    }

    #[rstest]
    #[case::win(1_010_000, 1_000_000, Some(Ordering::Greater))]
    #[case::tie(1_000_000, 1_000_000, Some(Ordering::Equal))]
    #[case::exactly_at_the_outlier_bound_stays(1_200_000, 1_000_000, Some(Ordering::Greater))]
    #[case::past_the_outlier_bound_drops(1_210_000, 1_000_000, None)]
    fn exact_in_measures_major_output_with_the_outlier_gate(
        #[case] ours: u64,
        #[case] l1: u64,
        #[case] expected: Option<Ordering>,
    ) {
        let m = measure(&stable_record(SwapMode::ExactIn, ours, l1), &prices());
        assert_eq!(m.map(|m| m.outcome), expected);
    }

    /// Pairing is by signature, so a signature seen twice is ambiguous and drops wholesale.
    #[test]
    fn collect_drops_multi_row_signatures_wholesale() {
        let mut a = stable_record(SwapMode::ExactIn, 1_000_000, 999_000);
        a.original_signature = "sigA".to_string();
        let mut a2 = a.clone();
        a2.amount = 2_000_000;
        let mut b = stable_record(SwapMode::ExactIn, 1_000_000, 999_000);
        b.original_signature = "sigB".to_string();

        let (kept, stats) = collect(vec![a, a2, b]);
        assert_eq!(kept.len(), 1);
        assert_eq!(kept[0].original_signature, "sigB");
        assert_eq!(stats.dropped_multi_row, 1);
    }

    /// An ExactOut leg is measured on its input, so a non-major input has no price to measure
    /// against — and dropping it before the signature bookkeeping keeps it from making a
    /// measurable row ambiguous.
    #[test]
    fn collect_skips_exact_out_legs_it_cannot_price() {
        let exotic = LegRecord {
            input_mint: Address::default(),
            ..stable_record(SwapMode::ExactOut, 1_000_000, 999_000)
        };
        let major = stable_record(SwapMode::ExactOut, 1_000_000, 999_000);

        let (kept, stats) = collect(vec![exotic, major]);
        assert_eq!(kept.len(), 1);
        assert_eq!(kept[0].input_mint, mints::USDC);
        assert_eq!(stats.dropped_multi_row, 0, "a skipped row is not a pairing");
    }
}