fix

2026-08-23 22:52:13 +02:00
parent 8983cfc8a1
commit 296dc3c841
32 changed files with 2025 additions and 52 deletions
@@ -0,0 +1,100 @@
+// Command backfill-ai-usage recovers AI cost history for products that were
+// processed before prompt/usage capture existed.
+//
+// It reads the provider usage blocks still stored in processed_products.gpt_response
+// and rolls them into ai_usage_daily with source='backfill', so /admin/ai-costs
+// shows spend from before the feature shipped. See internal/aiaudit/backfill.go for
+// what is and is not recoverable.
+//
+//	go run ./cmd/backfill-ai-usage                 # dry run, reports what it would write
+//	go run ./cmd/backfill-ai-usage -apply
+//	go run ./cmd/backfill-ai-usage -apply -before 2026-08-23
+//
+// Safe to re-run: each pass replaces its own rows and never touches live capture.
+package main
+
+import (
+	"context"
+	"encoding/json"
+	"flag"
+	"fmt"
+	"log"
+	"os"
+	"strings"
+	"time"
+
+	"github.com/descrybe/descrybe-v2/apps/api/internal/aiaudit"
+	"github.com/descrybe/descrybe-v2/apps/api/internal/config"
+	"github.com/jackc/pgx/v5/pgxpool"
+)
+
+func main() {
+	apply := flag.Bool("apply", false, "write rows (default is a dry run)")
+	before := flag.String("before", "", "only products last updated before this date (YYYY-MM-DD); default: the day live capture started")
+	dsn := flag.String("postgres", os.Getenv("DATABASE_URL"), "Postgres URL (defaults to DATABASE_URL / repo .env)")
+	flag.Parse()
+
+	ctx, cancel := context.WithTimeout(context.Background(), 15*time.Minute)
+	defer cancel()
+
+	url := strings.TrimSpace(*dsn)
+	if url == "" {
+		cfg, err := config.Load()
+		if err != nil {
+			log.Fatalf("config: %v", err)
+		}
+		url = cfg.DatabaseURL
+	}
+	pool, err := pgxpool.New(ctx, url)
+	if err != nil {
+		log.Fatalf("db: %v", err)
+	}
+	defer pool.Close()
+
+	cutoff, err := resolveCutoff(ctx, pool, *before)
+	if err != nil {
+		log.Fatalf("cutoff: %v", err)
+	}
+
+	res, err := aiaudit.BackfillFromProcessedProducts(ctx, pool, cutoff, !*apply)
+	if err != nil {
+		log.Fatalf("backfill: %v", err)
+	}
+
+	b, _ := json.MarshalIndent(res, "", "  ")
+	fmt.Println(string(b))
+	fmt.Printf("\nrecovered %s across %d product(s); %d product(s) had no stored usage block\n",
+		usd(res.CostUSD()), res.ProductsUsed, res.ProductsNoUsage)
+	if !*apply {
+		fmt.Println("dry run — re-run with -apply to write these rows")
+	}
+}
+
+// resolveCutoff keeps the backfill from double-counting calls the live recorder
+// already logged: by default it stops at the first day live capture wrote a row.
+func resolveCutoff(ctx context.Context, pool *pgxpool.Pool, before string) (time.Time, error) {
+	if s := strings.TrimSpace(before); s != "" {
+		t, err := time.Parse("2006-01-02", s)
+		if err != nil {
+			return time.Time{}, fmt.Errorf("invalid -before date (YYYY-MM-DD): %w", err)
+		}
+		return t.UTC(), nil
+	}
+	start, err := aiaudit.LiveCaptureStart(ctx, pool)
+	if err != nil {
+		return time.Time{}, err
+	}
+	if start.IsZero() {
+		log.Println("no live capture recorded yet — scanning all history")
+		return time.Time{}, nil
+	}
+	log.Printf("live capture starts %s — backfilling products updated before that", start.Format("2006-01-02"))
+	return start, nil
+}
+
+func usd(v float64) string {
+	if v != 0 && v < 0.01 {
+		return fmt.Sprintf("$%.6f", v)
+	}
+	return fmt.Sprintf("$%.2f", v)
+}