Initial commit of Descrybe v2 without local scratch artifacts.

Drop one-shot tmp/axe scripts and agent i18n scratch so the Gitea tree is deployable.
2026-08-09 22:47:43 +02:00
commit 8580c996c3
1285 changed files with 325780 additions and 0 deletions
@@ -0,0 +1,109 @@
+package main
+
+import (
+	"context"
+	"database/sql"
+	"encoding/json"
+	"log"
+	"strconv"
+
+	"github.com/google/uuid"
+	"github.com/jackc/pgx/v5/pgxpool"
+)
+
+// migrateFiles copies file *metadata* only.
+//
+// Blobs strategy (cutover):
+//   - Do NOT stream MySQL/local blob bytes through the migrator.
+//   - Preserve legacy url/path in files.path and legacy id in metadata._legacy_file_id.
+//   - Operators re-attach object storage / local volumes under the same relative paths,
+//     or run a separate rsync/S3 sync keyed by legacy id after DNS freeze.
+//   - raw_products.file_id is left unset until a follow-up remapper exists.
+func migrateFiles(
+	ctx context.Context,
+	mysqlDB *sql.DB,
+	pg *pgxpool.Pool,
+	companyMap, userMap, fileMap map[string]string,
+	allow map[string]bool,
+	report map[string]int,
+	dryRun bool,
+) {
+	if !mysqlTableExists(ctx, mysqlDB, "files") {
+		log.Printf("files skipped: table missing (blobs strategy: metadata-only when present)")
+		return
+	}
+
+	clause, cargs := mysqlCompanyFilter("company_id", allow)
+	rows, err := mysqlDB.QueryContext(ctx, `
+		SELECT id, company_id, COALESCE(user_id, ''), file_name,
+			COALESCE(file_type, ''), COALESCE(file_size, 0),
+			COALESCE(status, 'uploaded'), url, metadata
+		FROM files WHERE 1=1`+clause, cargs...)
+	if err != nil {
+		// Older dumps may lack metadata/url/status.
+		rows, err = mysqlDB.QueryContext(ctx, `
+			SELECT id, company_id, COALESCE(user_id, ''), file_name,
+				COALESCE(file_type, ''), COALESCE(file_size, 0),
+				'uploaded', NULL, NULL
+			FROM files WHERE 1=1`+clause, cargs...)
+	}
+	if err != nil {
+		log.Printf("files skipped: %v", err)
+		return
+	}
+	defer rows.Close()
+
+	for rows.Next() {
+		var legacyID int64
+		var companyLegacy, userLegacy, name, fileType, status string
+		var size int64
+		var url sql.NullString
+		var metadata []byte
+		if err := rows.Scan(&legacyID, &companyLegacy, &userLegacy, &name, &fileType, &size, &status, &url, &metadata); err != nil {
+			report["files_skipped"]++
+			continue
+		}
+		cid, ok := companyMap[companyLegacy]
+		if !ok {
+			report["files_skipped"]++
+			continue
+		}
+		meta := map[string]any{}
+		if len(metadata) > 0 {
+			_ = json.Unmarshal(metadata, &meta)
+		}
+		meta["_legacy_file_id"] = legacyID
+		meta["_blob_strategy"] = "metadata_only_resync_paths"
+		if fileType != "" {
+			meta["file_type"] = fileType
+		}
+		metaBytes, _ := json.Marshal(meta)
+
+		newID := uuid.New()
+		legacyKey := strconv.FormatInt(legacyID, 10)
+		fileMap[legacyKey] = newID.String()
+
+		var uid *string
+		if userLegacy != "" {
+			if mapped, ok := userMap[userLegacy]; ok {
+				uid = &mapped
+			}
+		}
+
+		if dryRun {
+			report["files"]++
+			continue
+		}
+		_, err = pg.Exec(ctx, `
+			INSERT INTO files (id, company_id, user_id, name, path, content_type, size_bytes, status, metadata)
+			VALUES ($1, $2, $3::uuid, $4, $5, $6, $7, $8, $9::jsonb)`,
+			newID, cid, uid, name, nullString(url), nullStr(fileType), size, status, string(metaBytes))
+		if err != nil {
+			log.Printf("files insert %d: %v", legacyID, err)
+			delete(fileMap, legacyKey)
+			report["files_skipped"]++
+			continue
+		}
+		report["files"]++
+	}
+}