goods/api/internal/adminstore/quality.go

package adminstore

import (
	"context"
	"math"
	"time"

	"github.com/jackc/pgx/v5"
)

// Quality weights mirror ingestion/opengoods/etl/quality.py.
const (
	wCompleteness = 0.4
	wSourceTrust  = 0.3
	wAgreement    = 0.2
	wFreshness    = 0.1
)

// queryer is satisfied by both *pgxpool.Pool and pgx.Tx.
type queryer interface {
	QueryRow(ctx context.Context, sql string, args ...any) pgx.Row
}

func agreementFromSources(n int) float64 {
	switch {
	case n <= 1:
		return 0.5
	case n == 2:
		return 0.8
	default:
		return 1.0
	}
}

func freshnessFromAge(ageDays *float64) float64 {
	if ageDays == nil {
		return 0.5
	}
	d := *ageDays
	switch {
	case d <= 30:
		return 1.0
	case d <= 180:
		return 0.8
	case d <= 365:
		return 0.6
	case d <= 730:
		return 0.4
	default:
		return 0.2
	}
}

func (s *Store) computeQuality(ctx context.Context, q queryer, productID string) (float64, error) {
	var name, country *string
	var gtin *string
	var brandID, categoryID *string
	var netCanonical *float64
	var ingredients *string
	var hasNutri, hasImage bool
	err := q.QueryRow(ctx, `
SELECT p.name, p.gtin, p.brand_id, p.category_id, p.net_content_canonical,
       p.country_of_origin, f.ingredients_text,
       (f.nutriments IS NOT NULL AND f.nutriments::text <> '{}'),
       EXISTS (SELECT 1 FROM product_image pi WHERE pi.product_id = p.id)
FROM product p LEFT JOIN food_detail f ON f.product_id = p.id
WHERE p.id = $1`, productID).Scan(
		&name, &gtin, &brandID, &categoryID, &netCanonical, &country,
		&ingredients, &hasNutri, &hasImage)
	if err != nil {
		return 0, err
	}

	present := 0
	bump := func(ok bool) {
		if ok {
			present++
		}
	}
	bump(name != nil && *name != "")
	bump(gtin != nil && *gtin != "")
	bump(brandID != nil)
	bump(categoryID != nil)
	bump(netCanonical != nil)
	bump(country != nil && *country != "")
	bump(hasNutri)
	bump(ingredients != nil && *ingredients != "")
	bump(hasImage)
	completeness := float64(present) / float64(len(CompletenessFields))

	var sourceCount int
	var sourceTrust *float64
	var lastFetched *time.Time
	err = q.QueryRow(ctx, `
SELECT count(DISTINCT ps.source_id), COALESCE(max(s.trust_weight),0), max(ps.fetched_at)
FROM product_source ps LEFT JOIN source s ON s.id = ps.source_id
WHERE ps.product_id = $1`, productID).Scan(&sourceCount, &sourceTrust, &lastFetched)
	if err != nil {
		return 0, err
	}
	trust := 0.0
	if sourceTrust != nil {
		trust = *sourceTrust
	}

	var ageDays *float64
	if lastFetched != nil {
		d := time.Since(*lastFetched).Hours() / 24.0
		if d < 0 {
			d = 0
		}
		ageDays = &d
	}

	raw := wCompleteness*completeness + wSourceTrust*trust +
		wAgreement*agreementFromSources(sourceCount) + wFreshness*freshnessFromAge(ageDays)
	raw = math.Max(0, math.Min(1, raw))
	return math.Round(raw*1000) / 1000, nil
}

func (s *Store) recomputeQualityTx(ctx context.Context, tx pgx.Tx, productID string) (float64, error) {
	v, err := s.computeQuality(ctx, tx, productID)
	if err != nil {
		return 0, err
	}
	_, err = tx.Exec(ctx, "UPDATE product SET quality_score=$1 WHERE id=$2", v, productID)
	return v, err
}

func (s *Store) recomputeQuality(ctx context.Context, productID string) (float64, error) {
	v, err := s.computeQuality(ctx, s.pool, productID)
	if err != nil {
		return 0, err
	}
	_, err = s.pool.Exec(ctx, "UPDATE product SET quality_score=$1 WHERE id=$2", v, productID)
	return v, err
}