feat(search): PDF content via own ES index and oo index (#42)

.env.example

@@ -41,21 +41,7 @@ ONLYOFFICE_PROJECT_ID=33
 # ONLYOFFICE_ES_URL=http://127.0.0.1:9200
 # ONLYOFFICE_ES_INDEX=files_file
 # ONLYOFFICE_TENANT=
-
+#
-# Read-only SQL file store over the Community Server database (see
-# docs/community-server-db.md). The live portal runs MySQL; ONLYOFFICE_DSN is
-# `user:pass@tcp(host:port)/onlyoffice?parseTime=true`, or a `postgres://` URL.
-# ONLYOFFICE_DSN=
-# ONLYOFFICE_PG_DRIVER=          # postgres | mysql (auto-detected from DSN)
-# ONLYOFFICE_PG_TENANT=1         # falls back to ONLYOFFICE_TENANT
-# Alternatively build a PostgreSQL DSN from parts:
-# ONLYOFFICE_PG_HOST=
-# ONLYOFFICE_PG_PORT=5432
-# ONLYOFFICE_PG_USER=
-# ONLYOFFICE_PG_PASSWORD=
-# ONLYOFFICE_PG_DBNAME=onlyoffice
-# ONLYOFFICE_PG_SSLMODE=disable
-
 # oo index / oo search --backend own — own full-text index for PDF/scans,
 # filled by `oo index` from internal/docpipe (pdftotext + OCR). Defaults to
 # oo_docs_text. Uses the same ONLYOFFICE_ES_URL.

README.md

@@ -701,11 +701,9 @@ Requires `ONLYOFFICE_ES_URL` (plus optional `ONLYOFFICE_ES_INDEX`,
 
 The OnlyOffice index covers Office formats only, so PDFs (`S1019`-style invoice
 numbers) are not searchable by content. `oo index` extracts PDF text with
-`internal/docpipe` (pdftotext, OCR for scans) — including the text of embedded
+`internal/docpipe` (pdftotext, OCR for scans) into a separate index
-PDF attachments (`pdfdetach`: `<doc>.md`, `.xml`, covers the original/scan and
+(`ONLYOFFICE_ES_TEXT_INDEX`, default `oo_docs_text`); the OnlyOffice server and
-ZUGFeRD e-invoice XML) — into a separate index (`ONLYOFFICE_ES_TEXT_INDEX`,
+its index are **not** modified. Then search it with `--backend own`.
-default `oo_docs_text`); the OnlyOffice server and its index are **not**
-modified. Then search it with `--backend own`.
 
 ```bash
 oo index folder 634 --recursive --exts pdf    # populate (idempotent upsert)

cmd/office/fetch/list.go

@@ -17,18 +17,6 @@ const MailListPageSize = 25
 // Loader fetches list items for a menu subject using the OnlyOffice client.
 type Loader struct {
 	Client *onlyoffice.Client
-
-	// Files is the backend-agnostic file store used for file download, preview
-	// and delete. When nil it falls back to Client.FileStore(ProviderREST).
-	Files onlyoffice.FileStore
-}
-
-// fileStore returns the configured file store, defaulting to REST.
-func (l *Loader) fileStore() onlyoffice.FileStore {
-	if l.Files != nil {
-		return l.Files
-	}
-	return l.Client.FileStore(onlyoffice.ProviderREST)
 }
 
 // List returns items for the given list spec (nav leaf).
@@ -183,7 +171,11 @@ func (l *Loader) executeDelete(ctx context.Context, item model.Item) (string, er
 		}
 		return fmt.Sprintf("Deleted message %s", item.Title), nil
 	case model.KindFile:
-		if err := l.fileStore().Delete(ctx, []string{item.ID}); err != nil {
+		id, err := strconv.Atoi(item.ID)
+		if err != nil {
+			return "", err
+		}
+		if err := l.Client.DeleteFiles(ctx, []int{id}); err != nil {
 			return "", err
 		}
 		return fmt.Sprintf("Deleted file %s", item.Title), nil
@@ -207,7 +199,7 @@ func (l *Loader) executeDownload(ctx context.Context, item model.Item, destPath
 		return "", err
 	}
 	defer f.Close()
-	if _, err := l.fileStore().Download(ctx, item.ID, f); err != nil {
+	if _, err := l.Client.DownloadFile(ctx, item.ID, f); err != nil {
 		return "", err
 	}
 	return fmt.Sprintf("Downloaded to %s", destPath), nil

cmd/office/fetch/preview.go

@@ -5,6 +5,7 @@ import (
 	"context"
 	"fmt"
 
+	onlyoffice "github.com/eslider/go-onlyoffice"
 	"github.com/eslider/go-onlyoffice/cmd/office/model"
 	"github.com/eslider/go-onlyoffice/cmd/office/preview"
 )
@@ -29,11 +30,13 @@ func (l *Loader) filePreviewMarkdown(ctx context.Context, item model.Item) (stri
 		return "", fmt.Errorf("file id missing")
 	}
 	name := item.Title
-	if e, err := l.fileStore().Stat(ctx, item.ID); err == nil && e.Title != "" {
+	if meta, err := l.Client.GetFile(ctx, item.ID); err == nil && meta != nil {
-		name = e.Title
+		if t := onlyoffice.FileEntryTitle(meta); t != "" {
+			name = t
+		}
 	}
 	var buf bytes.Buffer
-	if _, err := l.fileStore().Download(ctx, item.ID, &buf); err != nil {
+	if _, err := l.Client.DownloadFile(ctx, item.ID, &buf); err != nil {
 		return "", err
 	}
 	return preview.FileBytesToMarkdown(name, buf.Bytes())

cmd/oo/dav.go

@@ -3,7 +3,6 @@ package main
 import (
 	"fmt"
 	"os"
-	"time"
 
 	onlyoffice "github.com/eslider/go-onlyoffice"
 	"github.com/spf13/cobra"
@@ -13,8 +12,8 @@ func init() {
 	rootCmd.AddCommand(davCmd())
 }
 
-// davCmd exposes the Documents module through the backend-agnostic FileStore
+// davCmd exposes the Documents module through the same Dav calls that back
-// (DAV backend). The underlying Dav calls are the oo-webdav proven path:
+// oo-webdav (ListDavFolder / MoveDavItems / CopyDavItems / DownloadDavFile).
 // MoveDavItems sends resolveType=Skip + holdResult=true, which the legacy
 // fileops/move call without those params silently ignores (200 without move).
 func davCmd() *cobra.Command {
@@ -62,21 +61,35 @@ func davLsCmd() *cobra.Command {
 				printTable([]string{"id", "title", "filesCount", "foldersCount"}, rows)
 				return nil
 			}
-			entries, err := c.FileStore(onlyoffice.ProviderDAV).List(ctx, args[0])
+			l, err := c.ListDavFolder(ctx, args[0])
 			if err != nil {
 				return err
 			}
-			folders := make([]onlyoffice.Entry, 0, len(entries))
+			if outputFormat == "json" {
-			files := make([]onlyoffice.Entry, 0, len(entries))
+				folders := make([]map[string]any, 0, len(l.Folders))
-			for _, e := range entries {
+				for _, f := range l.Folders {
-				if e.Kind == onlyoffice.Folder {
+					folders = append(folders, map[string]any{
-					folders = append(folders, e)
+						"id":           f.ID,
-				} else {
+						"title":        f.Title,
-					files = append(files, e)
+						"filesCount":   f.FilesCount,
+						"foldersCount": f.FoldersCount,
+					})
 				}
+				files := make([]map[string]any, 0, len(l.Files))
+				for _, f := range l.Files {
+					files = append(files, map[string]any{
+						"id":      f.ID,
+						"title":   f.Title,
+						"size":    f.Size,
+						"updated": f.Updated,
+					})
 				}
-			frows := make([]map[string]any, 0, len(folders))
+				printObject(map[string]any{"folders": folders, "files": files})
-			for _, f := range folders {
+				return nil
+			}
+			if len(l.Folders) > 0 {
+				frows := make([]map[string]any, 0, len(l.Folders))
+				for _, f := range l.Folders {
 					frows = append(frows, map[string]any{
 						"id":           f.ID,
 						"title":        f.Title,
@@ -84,24 +97,23 @@ func davLsCmd() *cobra.Command {
 						"foldersCount": f.FoldersCount,
 					})
 				}
-			rows := make([]map[string]any, 0, len(files))
+				if outputFormat == "table" {
-			for _, f := range files {
+					fmt.Println("folders:")
+				}
+				printTable([]string{"id", "title", "filesCount", "foldersCount"}, frows)
+			}
+			rows := make([]map[string]any, 0, len(l.Files))
+			for _, f := range l.Files {
 				rows = append(rows, map[string]any{
 					"id":      f.ID,
 					"title":   f.Title,
 					"size":    f.Size,
-					"updated": entryUpdated(f),
+					"updated": f.Updated,
 				})
 			}
-			if outputFormat == "json" {
+			if outputFormat == "table" {
-				printObject(map[string]any{"folders": frows, "files": rows})
-				return nil
-			}
-			if len(frows) > 0 {
-				fmt.Println("folders:")
-				printTable([]string{"id", "title", "filesCount", "foldersCount"}, frows)
-			}
 				fmt.Println("files:")
+			}
 			printTable([]string{"id", "title", "size", "updated"}, rows)
 			return nil
 		},
@@ -109,18 +121,6 @@ func davLsCmd() *cobra.Command {
 	return cmd
 }
 
-// entryUpdated prefers the backend-native timestamp string so table/JSON output
-// round-trips what the API returned.
-func entryUpdated(e onlyoffice.Entry) string {
-	if e.Updated != "" {
-		return e.Updated
-	}
-	if e.Modified.IsZero() {
-		return ""
-	}
-	return e.Modified.Format(time.RFC3339)
-}
-
 func davMoveCmd() *cobra.Command {
 	var folderIDs []string
 	cmd := &cobra.Command{
@@ -132,8 +132,7 @@ func davMoveCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			ids := append(append([]string{}, folderIDs...), args[1:]...)
+			if err := c.MoveDavItems(cmd.Context(), folderIDs, args[1:], args[0]); err != nil {
-			if err := c.FileStore(onlyoffice.ProviderDAV).Move(cmd.Context(), ids, args[0]); err != nil {
 				return err
 			}
 			printObject(map[string]any{"moved_files": args[1:], "moved_folders": folderIDs, "dest": args[0]})
@@ -155,8 +154,7 @@ func davCopyCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			ids := append(append([]string{}, folderIDs...), args[1:]...)
+			if err := c.CopyDavItems(cmd.Context(), folderIDs, args[1:], args[0]); err != nil {
-			if err := c.FileStore(onlyoffice.ProviderDAV).Copy(cmd.Context(), ids, args[0]); err != nil {
 				return err
 			}
 			printObject(map[string]any{"copied_files": args[1:], "copied_folders": folderIDs, "dest": args[0]})
@@ -177,7 +175,7 @@ func davMkdirCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			f, err := c.FileStore(onlyoffice.ProviderDAV).CreateFolder(cmd.Context(), args[0], args[1])
+			f, err := c.CreateDavFolder(cmd.Context(), args[0], args[1])
 			if err != nil {
 				return err
 			}
@@ -202,8 +200,7 @@ func davRemoveCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			ids := append(append([]string{}, args...), folderIDs...)
+			if err := c.DeleteDavItems(cmd.Context(), folderIDs, args); err != nil {
-			if err := c.FileStore(onlyoffice.ProviderDAV).Delete(cmd.Context(), ids); err != nil {
 				return err
 			}
 			printObject(map[string]any{"deleted_files": args, "deleted_folders": folderIDs})
@@ -224,7 +221,7 @@ func davRenameFileCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			if err := c.FileStore(onlyoffice.ProviderDAV).Rename(cmd.Context(), args[0], args[1]); err != nil {
+			if err := c.RenameDavFile(cmd.Context(), args[0], args[1]); err != nil {
 				return err
 			}
 			printObject(map[string]any{"id": args[0], "title": args[1]})
@@ -243,7 +240,7 @@ func davRenameFolderCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			if err := c.FileStore(onlyoffice.ProviderDAV).Rename(cmd.Context(), args[0], args[1]); err != nil {
+			if err := c.RenameDavFolder(cmd.Context(), args[0], args[1]); err != nil {
 				return err
 			}
 			printObject(map[string]any{"id": args[0], "title": args[1]})
@@ -264,21 +261,20 @@ func davDownloadCmd() *cobra.Command {
 				return err
 			}
 			ctx := cmd.Context()
-			store := c.FileStore(onlyoffice.ProviderDAV)
+			f, err := c.GetFile(ctx, args[0])
-			e, err := store.Stat(ctx, args[0])
 			if err != nil {
 				return err
 			}
 			path := to
 			if path == "" {
-				path = onlyoffice.SafeLocalFileName(e.Title)
+				path = onlyoffice.SafeLocalFileName(onlyoffice.FileEntryTitle(f))
 			}
 			out, err := os.Create(path)
 			if err != nil {
 				return err
 			}
 			defer out.Close()
-			n, err := store.Download(ctx, args[0], out)
+			n, err := c.DownloadDavFile(ctx, args[0], out)
 			if err != nil {
 				_ = os.Remove(path)
 				return err

cmd/oo/projects_files.go

@@ -3,7 +3,6 @@ package main
 import (
 	"fmt"
 	"os"
-	"path/filepath"
 	"strconv"
 	"time"
 
@@ -160,22 +159,20 @@ func prjFilesDownloadCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			ctx := cmd.Context()
+			f, err := c.GetFile(cmd.Context(), args[0])
-			store := c.Files()
-			e, err := store.Stat(ctx, args[0])
 			if err != nil {
 				return err
 			}
 			path := to
 			if path == "" {
-				path = onlyoffice.SafeLocalFileName(e.Title)
+				path = onlyoffice.SafeLocalFileName(onlyoffice.FileEntryTitle(f))
 			}
 			out, err := os.Create(path)
 			if err != nil {
 				return err
 			}
 			defer out.Close()
-			n, err := store.Download(ctx, args[0], out)
+			n, err := c.DownloadFile(cmd.Context(), args[0], out)
 			if err != nil {
 				_ = os.Remove(path)
 				return err
@@ -202,16 +199,11 @@ func prjFilesRenameCmd() *cobra.Command {
 			if err != nil {
 				return err
 			}
-			store := c.Files()
+			entry, err := c.RenameFile(cmd.Context(), args[0], args[1])
-			if err := store.Rename(cmd.Context(), args[0], args[1]); err != nil {
-				return err
-			}
-			entry, err := store.Stat(cmd.Context(), args[0])
 			if err != nil {
 				return err
 			}
-			entry.Title = args[1]
+			printObject(fileEntryToMap(entry))
-			printObject(entryToMap(entry))
 			return nil
 		},
 	}
@@ -236,7 +228,7 @@ func prjFilesDeleteCmd() *cobra.Command {
 				}
 				ids = append(ids, id)
 			}
-			if err := c.Files().Delete(cmd.Context(), args); err != nil {
+			if err := c.DeleteFiles(cmd.Context(), ids); err != nil {
 				return err
 			}
 			printObject(map[string]any{"deleted": ids})
@@ -325,29 +317,6 @@ func fileEntryToMap(f *onlyoffice.FileEntry) map[string]any {
 	return m
 }
 
-// entryToMap renders a canonical Entry with the same keys as fileEntryToMap.
-func entryToMap(e onlyoffice.Entry) map[string]any {
-	m := map[string]any{
-		"id":            e.ID,
-		"title":         e.Title,
-		"fileExst":      filepath.Ext(e.Title),
-		"contentLength": contentLengthString(e.Size),
-	}
-	if e.Updated != "" {
-		m["updated"] = e.Updated
-	} else if !e.Modified.IsZero() {
-		m["updated"] = e.Modified.Format(time.RFC3339)
-	}
-	return m
-}
-
-func contentLengthString(n int64) string {
-	if n <= 0 {
-		return ""
-	}
-	return strconv.FormatInt(n, 10)
-}
-
 func fileIDStr(f *onlyoffice.FileEntry) string {
 	if f == nil || f.ID == nil {
 		return ""

cmd/oo/search.go

@@ -5,7 +5,6 @@ import (
 	"strings"
 
 	onlyoffice "github.com/eslider/go-onlyoffice"
-	"github.com/eslider/go-onlyoffice/cmd/internal/bootstrap"
 	"github.com/spf13/cobra"
 )
 
@@ -13,9 +12,9 @@ func init() {
 	rootCmd.AddCommand(searchCmd())
 }
 
-// searchCmd queries the OnlyOffice document index through the file facade. The
+// searchCmd queries the OnlyOffice Elasticsearch index directly. The REST
-// REST /api/2.0/files/@search endpoint only searches file names in the database;
+// /api/2.0/files/@search endpoint only searches file names in the database;
-// content search needs Elasticsearch (see docs/elasticsearch.md).
+// content search needs ES (see docs/elasticsearch.md).
 func searchCmd() *cobra.Command {
 	var (
 		content bool
@@ -41,15 +40,13 @@ func searchCmd() *cobra.Command {
 			if asJSON {
 				outputFormat = "json"
 			}
-			bootstrap.LoadEnv()
-			c := onlyoffice.NewClient(onlyoffice.GetEnvironmentCredentials())
 			var (
 				searcher onlyoffice.Searcher
 				err      error
 			)
 			switch strings.ToLower(strings.TrimSpace(backend)) {
 			case "", "oo", "elasticsearch":
-				searcher, err = c.Files().Search()
+				searcher, err = onlyoffice.NewESSearcher(onlyoffice.ESConfigFromEnv())
 			case "own", "es-text":
 				searcher, err = onlyoffice.NewESTextIndex(onlyoffice.ESTextConfigFromEnv())
 			default:

docs/community-server-db.md

@@ -1,130 +0,0 @@
----
-type: reference
-status: current
-related:
-  - README.md
-  - file_pg.go
-  - docs/elasticsearch.md
----
-
-# Community Server DB — прямой SQL-доступ (read-only)
-
-## Что это
-
-Бэкенд `pgStore` (`file_pg.go`) читает файлы и папки **напрямую из БД
-Community Server**, без HTTP-слоя. Реализует `FileStore` (`List`/`Stat`/
-`Download`) и `Searcher` по имени. Запись запрещена: все write-методы
-возвращают `ErrReadOnly`.
-
-## Что за БД (research, live)
-
-Проверено на VM `onlyoffice-v2` (SSH `127.0.0.1:32`):
-
-- Community Server работает на **MySQL 8.0**, не на PostgreSQL.
-  - Хост: `127.0.0.1:3306` внутри VM, база `onlyoffice`.
-  - Конфиг: `/etc/onlyoffice/communityserver/appsettings.production.json`,
-    `providerName: MySql.Data.MySqlClient`.
-  - Таблицы: `files_file`, `files_folder`, `files_folder_tree`,
-    `files_security`, тенанты — `tenants_tenants` (не `tenants`).
-- PostgreSQL 16 в той же VM — **наш** контур (`edw_docs`, роли `edw`/`edw_ro`,
-  office-assistant), к OnlyOffice отношения не имеет. `files_file` в PG нет.
-- Портал хранит файлы в **S3/MinIO** (DiscStorage только для мелочи).
-  Бакет `office`, объект — по ключу (см. ниже).
-
-Вывод: бэкенд назван по issue «PostgreSQL», но живой источник — MySQL.
-`database/sql` + драйвер по DSN: `mysql` для MySQL, `pgx` для PostgreSQL.
-`Name()` возвращает фактический движок (`mysql` или `postgres`).
-
-## Схема
-
-`files_file` — одна строка **на версию** (PK `tenant_id, id, version`):
-
-| поле | смысл |
-|------|-------|
-| `id` | id файла (тот же, что в REST/ES) |
-| `version` | номер версии этой строки |
-| `version_group` | номер версии |
-| `current_version` | `1` = текущая версия, `0` = старая |
-| `folder_id` | id родительской папки |
-| `title` | имя файла с расширением |
-| `content_length` | размер в байтах |
-| `create_on`, `modified_on` | даты (UTC, без зоны) |
-| `tenant_id` | тенант (портал) |
-
-`files_folder`: `id`, `parent_id`, `title`, `create_on`, `modified_on`,
-`tenant_id`. `files_folder_tree`: `folder_id`, `parent_id`, `level` — готовое
-дерево, пока не используется.
-
-Текущую строку файла берём по `current_version = 1`.
-
-## Доступ (SSH-туннель)
-
-MySQL слушает только `127.0.0.1:3306` внутри VM. Снаружи — SSH-туннель
-(SSH в VM открыт как `127.0.0.1:32`):
-
-```bash
-ssh -f -N -o ControlMaster=no -o ControlPath=none \
-    -p 32 -i ~/.ssh/id_ed25519 \
-    -L 13306:127.0.0.1:3306 root@127.0.0.1
-
-# MySQL DSN затем:
-# root:<pw>@tcp(127.0.0.1:13306)/onlyoffice?parseTime=true
-```
-
-`-o ControlMaster=no -o ControlPath=none` обязательны: иначе forward уходит в
-persistent master из `~/.ssh/config`.
-
-## Переменные
-
-| env | default | смысл |
-|-----|---------|-------|
-| `ONLYOFFICE_DSN` | — | DSN драйвера (MySQL `...@tcp(...)/...` или `postgres://...`) |
-| `ONLYOFFICE_PG_DRIVER` | авто | `postgres` или `mysql`; иначе по форме DSN |
-| `ONLYOFFICE_PG_TENANT` | `ONLYOFFICE_TENANT` | фильтр `tenant_id` (пусто = все) |
-| `ONLYOFFICE_PG_HOST/PORT/USER/PASSWORD/DBNAME/SSLMODE` | — | собрать PG DSN, если `ONLYOFFICE_DSN` пуст |
-
-Имена — в [`.env.example`](../.env.example). Секретов нет.
-
-## Download (MinIO)
-
-`Download` не ходит в REST. Ключ объекта собирается из строки `files_file`:
-
-```
-00/00/<tenant>/files/folder_<shard>/file_<id>/v<version>/content.<ext>
-shard = (id/1000 + 1) * 1000
-```
-
-`shard` — не `folder_id`, а следующая тысяча над `id` (файл 3727 →
-`folder_4000`). Проверено live по бакету `office`.
-
-Стриминг переиспользует `downloadMinioObject` из `storage_fallback.go`
-(та же подпись SigV4 и `MINIO_*`), без дублирования.
-
-## Тесты
-
-```bash
-go test ./...                              # unit: rebind, csObjectKey, маппинг
-go test -race ./...
-
-# integration (нужен DSN; skip без него)
-ONLYOFFICE_DSN='root:<pw>@tcp(127.0.0.1:13306)/onlyoffice?parseTime=true' \
-ONLYOFFICE_PG_TENANT=1 \
-ONLYOFFICE_PG_TEST_FILE_ID=22484 \
-ONLYOFFICE_PG_TEST_FOLDER_ID=649 \
-MINIO_ACCESS_KEY=... MINIO_SECRET_KEY=... \
-  go test -tags=integration -run TestIntegrationPGStore -v ./...
-```
-
-Integration сверяет `Stat`/`List`/`Download` с REST (`c.Files()`) и проверяет,
-что write-методы дают `ErrReadOnly`.
-
-## Грабли
-
-- MySQL хранит `datetime` без зоны; `parseTime=true` (ставится автоматически)
-  читает их как UTC. REST отдаёт `+02:00` — сравнивать моменты, не строки.
-- `GetFile` (REST) не отдаёт `contentLength` — размер сверять с `Stat` SQL.
-- Один файл = много строк `files_file` (по версиям). Без `current_version = 1`
-  получите дубликаты.
-- `folder_id` не входит в ключ MinIO; ключ считает `shard` от `id`.
-- Searcher SQL ищет только по имени (`LIKE`). Контент — Elasticsearch
-  ([elasticsearch.md](elasticsearch.md)).

docs/elasticsearch.md

@@ -174,22 +174,6 @@ OnlyOffice PDF лежит только по имени.
 - CLI: `oo index folder|files` наполняет индекс; `oo search --backend own`
   ищет по нему.
 
-### Встроенные вложения PDF
-
-Оцифрованные PDF несут вложения (`<doc>.md` — текст/таблицы скана,
-`<doc>.yaml`/`.json` — метаданные, `.xml` — EN 16931 CII eRechnung,
-`factur-x.xml` у ZUGFeRD; см. `office-assistant/docs/reference/document-metadata.md`).
-`TextIndexer` обходит их: `pdfdetach -list` перечисляет, `-save` сохраняет,
-каждое вложение проходит штатный `docpipe.ToMarkdown` (PDF/картинки → OCR,
-`.md`/`.txt` — как есть). Форматы, которые docpipe не конвертирует
-(`.xml`/`.html` — снимаются теги; `.json`/`.csv` — как текст), извлекаются
-текстом; нечитаемые — пропускаются.
-
-Текст склеивается: тело, затем по секции на вложение с маркером
-`[attachment: <имя>]` (функция `docpipe.JoinWithAttachments`). Индекс — тот же
-`file_id`, upsert идемпотентен. Нет вложений или pdfdetach/формат нечитаем —
-индексируется тело (без падения).
-
 Поля `oo_docs_text`:
 
 | поле | тип | смысл |
@@ -233,12 +217,8 @@ ONLYOFFICE_ES_URL=http://127.0.0.1:9200 \
 ```
 
 Интеграционный тест создаёт временный индекс, наполняет, ищет по контенту,
-проверяет фильтры и удаление, затем удаляет индекс;
+проверяет фильтры и удаление, затем удаляет индекс. Unit-тесты используют
-`TestIntegrationESTextIndexPDFAttachment` индексирует
+fake-store/fake-extractor и не требуют pdftotext/OCR.
-`testdata/pdf-with-attachment.pdf` реальным конвейером (pdfdetach + pdftotext)
-и ищет токен, лежащий только во вложении. Unit-тесты используют
-fake-store/fake-extractor и не требуют pdftotext/OCR (парсер списка, склейка
-`JoinWithAttachments`, снятие тегов `xmlToText` — чистые).
 
 ## Грабли
 
@@ -249,7 +229,3 @@ fake-store/fake-extractor и не требуют pdftotext/OCR (парсер с
 - `folder` фильтруется как id папки, а не как путь.
 - Дубликаты (напр. `S1055.pdf` и `2026-08-20-S1055-…`) дадут несколько строк —
   это ожидаемо, дедуп — на стороне потребителя.
-- Вложения: нужен `pdfdetach` (poppler); если его нет — индексируется только
-  тело. Вложенный PDF/картинка с плохим текстовым слоем проходит OCR, это
-  медленно. `.json`-метаданные (CuraSoft) индексируются как текст и могут
-  добавить шумовых токенов.

file_core.go

@@ -52,17 +52,8 @@ type Entry struct {
 	MIME     string
 	Created  time.Time
 	Modified time.Time
-	// Updated is the backend-native timestamp string, when the backend exposes
-	// one. It lets list output round-trip the API value; Modified is the
-	// parsed form for logic.
-	Updated  string
 	Version  int
 	Provider string
-
-	// Folder-only counters. Zero for files and for backends that do not
-	// report them.
-	FilesCount   int
-	FoldersCount int
 }
 
 // FileStore is the operation surface every file backend implements.
@@ -106,8 +97,8 @@ type Searcher interface {
 }
 
 // FileStore returns the adapter for a backend name: ProviderREST (default) or
-// ProviderDAV. Unknown or empty names select the REST backend. The composed
+// ProviderDAV. Unknown or empty names select the REST backend. The full facade
-// facade (backend selection/fallback) lives on FileClient in file_facade.go.
+// (backend composition) is deliberately left to a later change.
 func (c *Client) FileStore(backend string) FileStore {
 	switch strings.ToLower(strings.TrimSpace(backend)) {
 	case ProviderDAV, "webdav":
@@ -117,9 +108,8 @@ func (c *Client) FileStore(backend string) FileStore {
 	}
 }
 
-// Files returns the composed file facade. The returned *FileClient implements
+// Files returns the default (REST) file store.
-// FileStore, so callers that used Files() as the plain REST store keep working.
+func (c *Client) Files() FileStore { return c.FileStore(ProviderREST) }
-func (c *Client) Files() *FileClient { return c.newFileClient() }
 
 // retryStoreOp runs one store operation under the shared deterministic
 // transient-error policy (429/502/503/504).
@@ -150,7 +140,6 @@ func FileEntryToEntry(f *FileEntry, provider string) Entry {
 	e.MIME = mimeForTitle(e.Title, exst)
 	if f.Updated != nil {
 		e.Modified = *f.Updated
-		e.Updated = f.Updated.Format(time.RFC3339)
 	}
 	return e
 }
@@ -164,7 +153,6 @@ func DavFileToEntry(f DavFile, provider string) Entry {
 		Size:     f.Size,
 		MIME:     mimeForTitle(f.Title, ""),
 		Modified: f.ModTime(),
-		Updated:  f.Updated,
 		Provider: provider,
 	}
 }
@@ -177,10 +165,7 @@ func DavFolderToEntry(f DavFolder, provider string) Entry {
 		Title:    f.Title,
 		Kind:     Folder,
 		Modified: f.ModTime(),
-		Updated:      f.Updated,
 		Provider: provider,
-		FilesCount:   f.FilesCount,
-		FoldersCount: f.FoldersCount,
 	}
 }
 

file_es_text_integration_test.go

@@ -9,8 +9,6 @@ import (
 	"strings"
 	"testing"
 	"time"
-
-	"github.com/eslider/go-onlyoffice/internal/docpipe"
 )
 
 // TestIntegrationESTextIndex verifies the own full-text index end to end
@@ -92,67 +90,3 @@ func TestIntegrationESTextIndex(t *testing.T) {
 		t.Errorf("after delete search returned %d hits, want 0", len(hits))
 	}
 }
-
-// TestIntegrationESTextIndexPDFAttachment indexes testdata/pdf-with-attachment.pdf
-// through the real pipeline (TextIndexer + docpipe: pdfdetach + pdftotext) and
-// verifies that text living only in the embedded attachment is searchable.
-//
-// Requires ONLYOFFICE_ES_URL plus poppler (pdfdetach/pdftotext). No OnlyOffice
-// credentials are needed: a fixture FileStore serves the PDF bytes.
-func TestIntegrationESTextIndexPDFAttachment(t *testing.T) {
-	esURL := strings.TrimSpace(os.Getenv("ONLYOFFICE_ES_URL"))
-	if esURL == "" {
-		t.Skip("ONLYOFFICE_ES_URL not set — skipping Elasticsearch integration test")
-	}
-	if docpipe.LookPath().PDFDetach == "" {
-		t.Skip("pdfdetach not on PATH — skipping PDF attachment integration test")
-	}
-	pdf, err := os.ReadFile("testdata/pdf-with-attachment.pdf")
-	if err != nil {
-		t.Fatalf("read fixture: %v", err)
-	}
-
-	stamp := time.Now().UTC().Format("20060102150405")
-	idx, err := NewESTextIndex(ESTextConfig{URL: esURL, Index: "oo_docs_text_it_att_" + stamp})
-	if err != nil {
-		t.Fatalf("NewESTextIndex: %v", err)
-	}
-	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
-	defer cancel()
-	t.Cleanup(func() {
-		cleanupCtx, done := context.WithTimeout(context.Background(), 30*time.Second)
-		defer done()
-		_, _, _ = idx.do(cleanupCtx, http.MethodDelete, "/"+idx.Index(), nil, "")
-	})
-	if err := idx.Ensure(ctx); err != nil {
-		t.Fatalf("Ensure: %v", err)
-	}
-
-	store := &textFakeStore{files: map[string][]byte{"9001": pdf}}
-	ix := NewTextIndexer(store, idx)
-	res, err := ix.IndexEntries(ctx, []Entry{{ID: "9001", Title: "scan.pdf", ParentID: "777", Kind: File}}, IndexOptions{MinChars: 1})
-	if err != nil {
-		t.Fatalf("IndexEntries: %v", err)
-	}
-	if res.Indexed != 1 || res.Failed != 0 {
-		t.Fatalf("result = %+v, want one indexed doc", res)
-	}
-
-	// Token appears only inside the embedded goo-note.txt attachment.
-	hits, err := idx.Search(ctx, SearchQuery{Text: "gooattachmenttoken"})
-	if err != nil {
-		t.Fatalf("Search attachment token: %v", err)
-	}
-	if len(hits) != 1 || hits[0].ID != "9001" {
-		t.Fatalf("attachment-token hits = %+v, want doc 9001", hits)
-	}
-	if !strings.Contains(hits[0].Highlight, "gooattachmenttoken") {
-		t.Errorf("highlight = %q, want attachment token", hits[0].Highlight)
-	}
-	// Body text is indexed as before.
-	if hits, err := idx.Search(ctx, SearchQuery{Text: "goobodytoken"}); err != nil {
-		t.Fatalf("Search body token: %v", err)
-	} else if len(hits) != 1 {
-		t.Errorf("body-token hits = %d, want 1", len(hits))
-	}
-}

file_es_text_test.go

@@ -135,12 +135,12 @@ func TestESTextConfigFromEnvIndexDefault(t *testing.T) {
 }
 
 func TestTextIndexerIndexEntries(t *testing.T) {
-	store := &textFakeStore{
+	store := &fakeStore{
 		files: map[string][]byte{"1": []byte("PDFBYTES")},
 	}
-	idx := &textFakeIndex{}
+	idx := &fakeIndex{}
 	ix := NewTextIndexer(store, idx)
-	ix.Extractor = textFakeExtractor{prefix: "TEXT "}
+	ix.Extractor = fakeExtractor{prefix: "TEXT "}
 
 	res, err := ix.IndexEntries(context.Background(), []Entry{
 		{ID: "1", Title: "Rechnung.PDF", ParentID: "649", Kind: File},
@@ -164,10 +164,10 @@ func TestTextIndexerIndexEntries(t *testing.T) {
 }
 
 func TestTextIndexerRecordsExtractionFailure(t *testing.T) {
-	store := &textFakeStore{files: map[string][]byte{"1": []byte("x")}}
+	store := &fakeStore{files: map[string][]byte{"1": []byte("x")}}
-	idx := &textFakeIndex{}
+	idx := &fakeIndex{}
 	ix := NewTextIndexer(store, idx)
-	ix.Extractor = textFailingExtractor{}
+	ix.Extractor = failingExtractor{}
 
 	res, err := ix.IndexEntries(context.Background(), []Entry{{ID: "1", Title: "a.pdf", Kind: File}}, IndexOptions{})
 	if err != nil {
@@ -179,7 +179,7 @@ func TestTextIndexerRecordsExtractionFailure(t *testing.T) {
 }
 
 func TestTextIndexerPlanFolder(t *testing.T) {
-	store := &textFakeStore{dirs: map[string][]Entry{
+	store := &fakeStore{dirs: map[string][]Entry{
 		"root": {
 			{ID: "10", Title: "a.pdf", Kind: File},
 			{ID: "11", Title: "sub", Kind: Folder},
@@ -189,7 +189,7 @@ func TestTextIndexerPlanFolder(t *testing.T) {
 			{ID: "13", Title: "c.xlsx", Kind: File},
 		},
 	}}
-	ix := NewTextIndexer(store, &textFakeIndex{})
+	ix := NewTextIndexer(store, &fakeIndex{})
 
 	flat, err := ix.PlanFolder(context.Background(), "root", IndexOptions{})
 	if err != nil {
@@ -209,26 +209,26 @@ func TestTextIndexerPlanFolder(t *testing.T) {
 
 // --- fakes -----------------------------------------------------------------
 
-type textFakeStore struct {
+type fakeStore struct {
 	dirs  map[string][]Entry
 	files map[string][]byte
 	stat  map[string]Entry
 }
 
-func (f *textFakeStore) Name() string { return "fake" }
+func (f *fakeStore) Name() string { return "fake" }
 
-func (f *textFakeStore) List(_ context.Context, parentID string) ([]Entry, error) {
+func (f *fakeStore) List(_ context.Context, parentID string) ([]Entry, error) {
 	return f.dirs[parentID], nil
 }
 
-func (f *textFakeStore) Stat(_ context.Context, id string) (Entry, error) {
+func (f *fakeStore) Stat(_ context.Context, id string) (Entry, error) {
 	if e, ok := f.stat[id]; ok {
 		return e, nil
 	}
 	return Entry{}, fmt.Errorf("not found: %s", id)
 }
 
-func (f *textFakeStore) Download(_ context.Context, id string, w io.Writer) (int64, error) {
+func (f *fakeStore) Download(_ context.Context, id string, w io.Writer) (int64, error) {
 	b, ok := f.files[id]
 	if !ok {
 		return 0, fmt.Errorf("no bytes for %s", id)
@@ -237,30 +237,30 @@ func (f *textFakeStore) Download(_ context.Context, id string, w io.Writer) (int
 	return int64(n), err
 }
 
-func (f *textFakeStore) CreateFolder(context.Context, string, string) (Entry, error) {
+func (f *fakeStore) CreateFolder(context.Context, string, string) (Entry, error) {
 	return Entry{}, nil
 }
-func (f *textFakeStore) Upload(context.Context, string, string, io.Reader) (Entry, error) {
+func (f *fakeStore) Upload(context.Context, string, string, io.Reader) (Entry, error) {
 	return Entry{}, nil
 }
-func (f *textFakeStore) Move(context.Context, []string, string) error { return nil }
+func (f *fakeStore) Move(context.Context, []string, string) error { return nil }
-func (f *textFakeStore) Copy(context.Context, []string, string) error { return nil }
+func (f *fakeStore) Copy(context.Context, []string, string) error { return nil }
-func (f *textFakeStore) Rename(context.Context, string, string) error { return nil }
+func (f *fakeStore) Rename(context.Context, string, string) error { return nil }
-func (f *textFakeStore) Delete(context.Context, []string) error       { return nil }
+func (f *fakeStore) Delete(context.Context, []string) error       { return nil }
 
-type textFakeIndex struct{ docs []TextDoc }
+type fakeIndex struct{ docs []TextDoc }
 
-func (f *textFakeIndex) Put(_ context.Context, docs []TextDoc) error {
+func (f *fakeIndex) Put(_ context.Context, docs []TextDoc) error {
 	f.docs = append(f.docs, docs...)
 	return nil
 }
-func (f *textFakeIndex) Delete(context.Context, []string) error                   { return nil }
+func (f *fakeIndex) Delete(context.Context, []string) error                   { return nil }
-func (f *textFakeIndex) Search(context.Context, SearchQuery) ([]SearchHit, error) { return nil, nil }
+func (f *fakeIndex) Search(context.Context, SearchQuery) ([]SearchHit, error) { return nil, nil }
-func (f *textFakeIndex) Name() string                                             { return "fake" }
+func (f *fakeIndex) Name() string                                             { return "fake" }
 
-type textFakeExtractor struct{ prefix string }
+type fakeExtractor struct{ prefix string }
 
-func (f textFakeExtractor) Extract(path, _, _ string, _ int) (string, error) {
+func (f fakeExtractor) Extract(path, _, _ string, _ int) (string, error) {
 	b, err := os.ReadFile(path)
 	if err != nil {
 		return "", err
@@ -268,8 +268,8 @@ func (f textFakeExtractor) Extract(path, _, _ string, _ int) (string, error) {
 	return f.prefix + string(b), nil
 }
 
-type textFailingExtractor struct{}
+type failingExtractor struct{}
 
-func (textFailingExtractor) Extract(string, string, string, int) (string, error) {
+func (failingExtractor) Extract(string, string, string, int) (string, error) {
 	return "", fmt.Errorf("boom")
 }

file_facade.go

@@ -1,249 +0,0 @@
-package onlyoffice
-
-// Single file client (epic #34, F4 #38). FileClient composes the registered
-// FileStore and Searcher backends and picks one per operation: REST/DAV for
-// writes, PostgreSQL (when registered) for fast reads, Elasticsearch for name
-// and content search. Client.Files returns the facade; it also implements
-// FileStore, so existing callers keep compiling.
-
-import (
-	"context"
-	"errors"
-	"io"
-	"strings"
-)
-
-// ProviderES is the composed Elasticsearch searcher. The SQL store owns
-// ProviderPG/ProviderMySQL (file_pg.go); the facade references ProviderPG in
-// readOrder.
-const ProviderES = "elasticsearch"
-
-var (
-	errNoReadBackend  = errors.New("onlyoffice: no file backend registered for reads")
-	errNoWriteBackend = errors.New("onlyoffice: no file backend registered for writes")
-	errNoSearcher     = errors.New("onlyoffice: no search backend registered (set ONLYOFFICE_ES_URL)")
-)
-
-// FileClient is the single entry point for file operations. It holds the
-// registered backends and the order in which each operation tries them.
-type FileClient struct {
-	stores    map[string]FileStore
-	searchers map[string]Searcher
-
-	readOrder   []string
-	writeOrder  []string
-	searchOrder []string
-}
-
-// newFileClient builds the facade over the built-in REST and DAV stores. The
-// Elasticsearch searcher is registered when ONLYOFFICE_ES_URL is set; the
-// missing-credential case is left to Search so read-only commands still work.
-func (c *Client) newFileClient() *FileClient {
-	f := &FileClient{
-		stores: map[string]FileStore{
-			ProviderREST: &restStore{c: c},
-			ProviderDAV:  &davStore{c: c},
-		},
-		searchers:   map[string]Searcher{},
-		readOrder:   []string{ProviderPG, ProviderREST, ProviderDAV},
-		writeOrder:  []string{ProviderREST, ProviderDAV},
-		searchOrder: []string{ProviderES},
-	}
-	if cfg := ESConfigFromEnv(); cfg.URL != "" {
-		if es, err := NewESSearcher(cfg); err == nil {
-			f.searchers[ProviderES] = es
-		}
-	}
-	return f
-}
-
-// RegisterStore adds or replaces a named backend (for example the PostgreSQL
-// read store). The name is matched case-insensitively.
-func (f *FileClient) RegisterStore(name string, s FileStore) {
-	if f == nil || s == nil {
-		return
-	}
-	name = normalizeProvider(name)
-	if name == "" {
-		return
-	}
-	if f.stores == nil {
-		f.stores = map[string]FileStore{}
-	}
-	f.stores[name] = s
-}
-
-// RegisterSearcher adds or replaces a named search backend.
-func (f *FileClient) RegisterSearcher(name string, s Searcher) {
-	if f == nil || s == nil {
-		return
-	}
-	name = normalizeProvider(name)
-	if name == "" {
-		return
-	}
-	if f.searchers == nil {
-		f.searchers = map[string]Searcher{}
-	}
-	f.searchers[name] = s
-}
-
-// Read returns the preferred backend for reads: PostgreSQL when registered,
-// then REST, then WebDAV.
-func (f *FileClient) Read() FileStore { return f.firstStore(f.readOrder) }
-
-// Write returns the preferred backend for writes: REST, then WebDAV.
-func (f *FileClient) Write() FileStore { return f.firstStore(f.writeOrder) }
-
-// Search returns the preferred name/content searcher (Elasticsearch), or an
-// error when no search backend is configured.
-func (f *FileClient) Search() (Searcher, error) {
-	if f == nil {
-		return nil, errNoSearcher
-	}
-	for _, name := range f.searchOrder {
-		if s := f.searchers[normalizeProvider(name)]; s != nil {
-			return s, nil
-		}
-	}
-	return nil, errNoSearcher
-}
-
-// firstStore returns the first registered store in the order.
-func (f *FileClient) firstStore(order []string) FileStore {
-	if f == nil {
-		return nil
-	}
-	for _, name := range order {
-		if s := f.stores[normalizeProvider(name)]; s != nil {
-			return s
-		}
-	}
-	return nil
-}
-
-// orderedStores returns the registered stores in the order.
-func (f *FileClient) orderedStores(order []string) []FileStore {
-	if f == nil {
-		return nil
-	}
-	out := make([]FileStore, 0, len(order))
-	for _, name := range order {
-		if s := f.stores[normalizeProvider(name)]; s != nil {
-			out = append(out, s)
-		}
-	}
-	return out
-}
-
-func normalizeProvider(name string) string {
-	return strings.ToLower(strings.TrimSpace(name))
-}
-
-// Name implements FileStore and reports the preferred read backend.
-func (f *FileClient) Name() string {
-	if s := f.Read(); s != nil {
-		return s.Name()
-	}
-	return ""
-}
-
-// List reads from the preferred backend, falling back to the next read backend
-// only on a transient error (429/502/503/504).
-func (f *FileClient) List(ctx context.Context, parentID string) ([]Entry, error) {
-	return fallbackRead(ctx, f.orderedStores(f.readOrder), func(s FileStore) ([]Entry, error) {
-		return s.List(ctx, parentID)
-	})
-}
-
-// Stat reads from the preferred backend, with the same transient fallback.
-func (f *FileClient) Stat(ctx context.Context, id string) (Entry, error) {
-	return fallbackRead(ctx, f.orderedStores(f.readOrder), func(s FileStore) (Entry, error) {
-		return s.Stat(ctx, id)
-	})
-}
-
-// Download streams file bytes. It does not fall back: a failed attempt may have
-// already written partial bytes into w, so a second backend would append.
-func (f *FileClient) Download(ctx context.Context, id string, w io.Writer) (int64, error) {
-	s := f.Read()
-	if s == nil {
-		return 0, errNoReadBackend
-	}
-	return s.Download(ctx, id, w)
-}
-
-// CreateFolder writes to the preferred write backend.
-func (f *FileClient) CreateFolder(ctx context.Context, parentID, title string) (Entry, error) {
-	s := f.Write()
-	if s == nil {
-		return Entry{}, errNoWriteBackend
-	}
-	return s.CreateFolder(ctx, parentID, title)
-}
-
-// Upload writes to the preferred write backend.
-func (f *FileClient) Upload(ctx context.Context, parentID, title string, r io.Reader) (Entry, error) {
-	s := f.Write()
-	if s == nil {
-		return Entry{}, errNoWriteBackend
-	}
-	return s.Upload(ctx, parentID, title, r)
-}
-
-// Move writes to the preferred write backend.
-func (f *FileClient) Move(ctx context.Context, ids []string, parentID string) error {
-	s := f.Write()
-	if s == nil {
-		return errNoWriteBackend
-	}
-	return s.Move(ctx, ids, parentID)
-}
-
-// Copy writes to the preferred write backend.
-func (f *FileClient) Copy(ctx context.Context, ids []string, parentID string) error {
-	s := f.Write()
-	if s == nil {
-		return errNoWriteBackend
-	}
-	return s.Copy(ctx, ids, parentID)
-}
-
-// Rename writes to the preferred write backend.
-func (f *FileClient) Rename(ctx context.Context, id, title string) error {
-	s := f.Write()
-	if s == nil {
-		return errNoWriteBackend
-	}
-	return s.Rename(ctx, id, title)
-}
-
-// Delete writes to the preferred write backend.
-func (f *FileClient) Delete(ctx context.Context, ids []string) error {
-	s := f.Write()
-	if s == nil {
-		return errNoWriteBackend
-	}
-	return s.Delete(ctx, ids)
-}
-
-// fallbackRead runs op against each store in order, moving on only when the
-// error is transient. Non-transient errors (not found, forbidden) are final.
-func fallbackRead[T any](ctx context.Context, stores []FileStore, op func(FileStore) (T, error)) (T, error) {
-	var zero T
-	if len(stores) == 0 {
-		return zero, errNoReadBackend
-	}
-	var err error
-	for i, s := range stores {
-		var v T
-		v, err = op(s)
-		if err == nil {
-			return v, nil
-		}
-		if i == len(stores)-1 || !Transient(err) {
-			return zero, err
-		}
-	}
-	return zero, err
-}

file_facade_test.go

@@ -1,230 +0,0 @@
-package onlyoffice
-
-import (
-	"context"
-	"errors"
-	"fmt"
-	"io"
-	"strings"
-	"testing"
-)
-
-// fakeStore is a FileStore test double; it records which backend served a call
-// and returns a canned result or error.
-type fakeStore struct {
-	name    string
-	entries []Entry
-	err     error
-	calls   *[]string
-}
-
-func (f *fakeStore) record(op string) {
-	if f.calls != nil {
-		*f.calls = append(*f.calls, op+":"+f.name)
-	}
-}
-
-func (f *fakeStore) Name() string { return f.name }
-
-func (f *fakeStore) List(_ context.Context, _ string) ([]Entry, error) {
-	f.record("list")
-	if f.err != nil {
-		return nil, f.err
-	}
-	return f.entries, nil
-}
-
-func (f *fakeStore) Stat(_ context.Context, id string) (Entry, error) {
-	f.record("stat")
-	if f.err != nil {
-		return Entry{}, f.err
-	}
-	return Entry{ID: id, Title: "t-" + f.name, Provider: f.name}, nil
-}
-
-func (f *fakeStore) CreateFolder(_ context.Context, _, title string) (Entry, error) {
-	f.record("mkdir")
-	if f.err != nil {
-		return Entry{}, f.err
-	}
-	return Entry{ID: "new", Title: title, Provider: f.name}, nil
-}
-
-func (f *fakeStore) Upload(_ context.Context, _, title string, _ io.Reader) (Entry, error) {
-	f.record("upload")
-	return Entry{ID: "up", Title: title, Provider: f.name}, f.err
-}
-
-func (f *fakeStore) Download(_ context.Context, _ string, _ io.Writer) (int64, error) {
-	f.record("download")
-	return 0, f.err
-}
-
-func (f *fakeStore) Move(_ context.Context, _ []string, _ string) error {
-	f.record("move")
-	return f.err
-}
-
-func (f *fakeStore) Copy(_ context.Context, _ []string, _ string) error {
-	f.record("copy")
-	return f.err
-}
-
-func (f *fakeStore) Rename(_ context.Context, _, _ string) error {
-	f.record("rename")
-	return f.err
-}
-
-func (f *fakeStore) Delete(_ context.Context, _ []string) error {
-	f.record("delete")
-	return f.err
-}
-
-type fakeSearcher struct{ name string }
-
-func (s *fakeSearcher) Name() string { return s.name }
-
-func (s *fakeSearcher) Search(_ context.Context, _ SearchQuery) ([]SearchHit, error) {
-	return []SearchHit{{Entry: Entry{Title: s.name}}}, nil
-}
-
-func newFacadeTestClient(stores map[string]FileStore, read, write []string) *FileClient {
-	return &FileClient{
-		stores:     stores,
-		searchers:  map[string]Searcher{},
-		readOrder:  read,
-		writeOrder: write,
-	}
-}
-
-// TestFileClientIsFileStore guarantees the facade can stand in for the
-// interface anywhere a plain FileStore is expected.
-func TestFileClientIsFileStore(t *testing.T) {
-	var _ FileStore = (*FileClient)(nil)
-}
-
-func TestClientFilesPrefersRESTForReadsAndWrites(t *testing.T) {
-	c := NewClient(Credentials{})
-	f := c.Files()
-	if got := f.Read().Name(); got != ProviderREST {
-		t.Errorf("Read().Name() = %q, want %q", got, ProviderREST)
-	}
-	if got := f.Write().Name(); got != ProviderREST {
-		t.Errorf("Write().Name() = %q, want %q", got, ProviderREST)
-	}
-	if got := f.Name(); got != ProviderREST {
-		t.Errorf("Name() = %q, want %q", got, ProviderREST)
-	}
-}
-
-func TestFileClientPostgresTakesReadPriority(t *testing.T) {
-	pg := &fakeStore{name: ProviderPG}
-	f := newFacadeTestClient(
-		map[string]FileStore{ProviderREST: &fakeStore{name: ProviderREST}, ProviderPG: pg},
-		[]string{ProviderPG, ProviderREST},
-		[]string{ProviderREST},
-	)
-	if got := f.Read().Name(); got != ProviderPG {
-		t.Errorf("Read().Name() = %q, want %q", got, ProviderPG)
-	}
-	if got := f.Write().Name(); got != ProviderREST {
-		t.Errorf("Write().Name() = %q, want %q (PG is read-only)", got, ProviderREST)
-	}
-}
-
-func TestFileClientRegisterStoreNormalizesName(t *testing.T) {
-	pg := &fakeStore{name: "pg"}
-	f := newFacadeTestClient(map[string]FileStore{}, []string{ProviderPG}, nil)
-	f.RegisterStore("  POSTGRES  ", pg)
-	if got := f.Read(); got != pg {
-		t.Fatalf("Read() = %v, want registered postgres store", got)
-	}
-	f.RegisterStore("", pg)
-	f.RegisterStore("pg", nil)
-}
-
-func TestFileClientReadFallsBackOnlyOnTransient(t *testing.T) {
-	var calls []string
-	primary := &fakeStore{name: "primary", err: fmt.Errorf("onlyoffice: list: 503 unavailable"), calls: &calls}
-	secondary := &fakeStore{name: "secondary", entries: []Entry{{ID: "1"}}, calls: &calls}
-	f := newFacadeTestClient(
-		map[string]FileStore{"primary": primary, "secondary": secondary},
-		[]string{"primary", "secondary"},
-		nil,
-	)
-	got, err := f.List(context.Background(), "root")
-	if err != nil {
-		t.Fatalf("List: %v", err)
-	}
-	if len(got) != 1 || got[0].ID != "1" {
-		t.Fatalf("List() = %+v, want secondary entry", got)
-	}
-	want := []string{"list:primary", "list:secondary"}
-	if fmt.Sprint(calls) != fmt.Sprint(want) {
-		t.Fatalf("call order = %v, want %v", calls, want)
-	}
-}
-
-func TestFileClientReadStopsOnPermanentError(t *testing.T) {
-	var calls []string
-	primary := &fakeStore{name: "primary", err: errors.New("onlyoffice: not found"), calls: &calls}
-	secondary := &fakeStore{name: "secondary", entries: []Entry{{ID: "1"}}, calls: &calls}
-	f := newFacadeTestClient(
-		map[string]FileStore{"primary": primary, "secondary": secondary},
-		[]string{"primary", "secondary"},
-		nil,
-	)
-	if _, err := f.List(context.Background(), "root"); err == nil {
-		t.Fatal("expected permanent error to be returned")
-	}
-	if len(calls) != 1 || calls[0] != "list:primary" {
-		t.Fatalf("secondary backend must not run on a permanent error: %v", calls)
-	}
-}
-
-func TestFileClientWriteUsesWriteBackend(t *testing.T) {
-	var calls []string
-	rest := &fakeStore{name: ProviderREST, calls: &calls}
-	dav := &fakeStore{name: ProviderDAV, calls: &calls}
-	f := newFacadeTestClient(
-		map[string]FileStore{ProviderREST: rest, ProviderDAV: dav},
-		[]string{ProviderREST},
-		[]string{ProviderREST, ProviderDAV},
-	)
-	if _, err := f.CreateFolder(context.Background(), "p", "t"); err != nil {
-		t.Fatalf("CreateFolder: %v", err)
-	}
-	if _, err := f.Upload(context.Background(), "p", "t", nil); err != nil {
-		t.Fatalf("Upload: %v", err)
-	}
-	if len(calls) != 2 || calls[0] != "mkdir:rest" || calls[1] != "upload:rest" {
-		t.Fatalf("write calls = %v, want REST", calls)
-	}
-}
-
-func TestFileClientWriteWithoutBackend(t *testing.T) {
-	f := newFacadeTestClient(map[string]FileStore{}, nil, nil)
-	if err := f.Delete(context.Background(), []string{"1"}); !errors.Is(err, errNoWriteBackend) {
-		t.Fatalf("Delete err = %v, want errNoWriteBackend", err)
-	}
-	if _, err := f.List(context.Background(), "root"); !errors.Is(err, errNoReadBackend) {
-		t.Fatalf("List err = %v, want errNoReadBackend", err)
-	}
-}
-
-func TestFileClientSearchSelection(t *testing.T) {
-	f := &FileClient{searchers: map[string]Searcher{}, searchOrder: []string{ProviderES}}
-	_, err := f.Search()
-	if err == nil || !strings.Contains(err.Error(), "ONLYOFFICE_ES_URL") {
-		t.Fatalf("Search without backend = %v, want ONLYOFFICE_ES_URL hint", err)
-	}
-	es := &fakeSearcher{name: "fake-es"}
-	f.RegisterSearcher(ProviderES, es)
-	got, err := f.Search()
-	if err != nil {
-		t.Fatalf("Search: %v", err)
-	}
-	if got.Name() != "fake-es" {
-		t.Fatalf("searcher = %q, want fake-es", got.Name())
-	}
-}

file_pg.go

@@ -1,516 +0,0 @@
-package onlyoffice
-
-// Read-only SQL backend of the unified file client (epic #34, F2 #36).
-//
-// The goal is to read files and folders straight from the Community Server
-// database, without the REST layer. Research on the live portal (VM
-// `onlyoffice-v2`) showed the server runs on **MySQL 8.0** (`files_file`,
-// `files_folder`, `files_folder_tree`, tenant `tenants_tenants`), not
-// PostgreSQL — see docs/community-server-db.md. The store below therefore
-// speaks `database/sql` and selects its driver from the DSN, so it works
-// against the live MySQL today and against PostgreSQL if the portal is ever
-// migrated. Every query is a SELECT; the write methods of FileStore return
-// ErrReadOnly.
-//
-// Downloads follow the portal's S3/MinIO object layout through the shared
-// MinIO helper in storage_fallback.go — no HTTP file endpoint is used.
-
-import (
-	"context"
-	"database/sql"
-	"errors"
-	"fmt"
-	"io"
-	"net/http"
-	"net/url"
-	"os"
-	"path/filepath"
-	"strconv"
-	"strings"
-	"time"
-
-	"github.com/go-sql-driver/mysql"
-	_ "github.com/jackc/pgx/v5/stdlib"
-)
-
-// Provider names for the SQL backend. ProviderPG is the value Name reports for
-// a PostgreSQL connection and ProviderMySQL for MySQL.
-const (
-	ProviderPG    = "postgres"
-	ProviderMySQL = "mysql"
-)
-
-// ErrReadOnly is returned by every FileStore write method of the SQL backend.
-var ErrReadOnly = errors.New("onlyoffice: sql file store is read-only")
-
-const (
-	pgConnectTimeout = 10 * time.Second
-	pgSearchLimit    = 50
-	pgSearchMaxLimit = 500
-)
-
-// PGConfig configures the read-only SQL store. DSN is a driver DSN:
-// `user:pass@tcp(host:port)/onlyoffice?parseTime=true` for MySQL or a
-// `postgres://` / libpq keyword string for PostgreSQL. Driver, when set,
-// forces the engine ("postgres" or "mysql"); otherwise it is detected from the
-// DSN. Tenant filters rows (empty means all tenants).
-type PGConfig struct {
-	DSN    string
-	Driver string
-	Tenant string
-}
-
-// PGConfigFromEnv reads ONLYOFFICE_DSN (or the ONLYOFFICE_PG_* parts),
-// ONLYOFFICE_PG_DRIVER and the tenant from ONLYOFFICE_PG_TENANT /
-// ONLYOFFICE_TENANT. The library never loads dotfiles — the CLI does that.
-func PGConfigFromEnv() PGConfig {
-	dsn := strings.TrimSpace(os.Getenv("ONLYOFFICE_DSN"))
-	if dsn == "" {
-		dsn = pgDSNFromParts()
-	}
-	return PGConfig{
-		DSN:    dsn,
-		Driver: strings.TrimSpace(os.Getenv("ONLYOFFICE_PG_DRIVER")),
-		Tenant: firstNonEmpty(os.Getenv("ONLYOFFICE_PG_TENANT"), os.Getenv("ONLYOFFICE_TENANT")),
-	}
-}
-
-// pgDSNFromParts builds a libpq keyword DSN from ONLYOFFICE_PG_* variables.
-// It returns "" unless a host is set, which keeps the MySQL path (ONLYOFFICE_DSN)
-// the default.
-func pgDSNFromParts() string {
-	host := strings.TrimSpace(os.Getenv("ONLYOFFICE_PG_HOST"))
-	if host == "" {
-		return ""
-	}
-	port := firstNonEmpty(os.Getenv("ONLYOFFICE_PG_PORT"), "5432")
-	dbname := firstNonEmpty(os.Getenv("ONLYOFFICE_PG_DBNAME"), "onlyoffice")
-	sslmode := firstNonEmpty(os.Getenv("ONLYOFFICE_PG_SSLMODE"), "disable")
-	return fmt.Sprintf("host=%s port=%s user=%s password=%s dbname=%s sslmode=%s",
-		host, port, os.Getenv("ONLYOFFICE_PG_USER"), os.Getenv("ONLYOFFICE_PG_PASSWORD"), dbname, sslmode)
-}
-
-// pgStore is a read-only FileStore/Searcher over the Community Server database.
-type pgStore struct {
-	db        *sql.DB
-	driver    string
-	tenantID  int64
-	hasTenant bool
-	http      *http.Client
-}
-
-var (
-	_ FileStore = (*pgStore)(nil)
-	_ Searcher  = (*pgStore)(nil)
-)
-
-// NewPGStore opens the database and verifies connectivity. It never writes.
-func NewPGStore(cfg PGConfig) (*pgStore, error) {
-	dsn := strings.TrimSpace(cfg.DSN)
-	if dsn == "" {
-		return nil, fmt.Errorf("onlyoffice: sql file store: empty DSN (set ONLYOFFICE_DSN)")
-	}
-	driver := pgDriver(dsn, cfg.Driver)
-	dsn, err := normalizeSQLDSN(driver, dsn)
-	if err != nil {
-		return nil, err
-	}
-	db, err := sql.Open(sqlDriverName(driver), dsn)
-	if err != nil {
-		return nil, fmt.Errorf("onlyoffice: sql file store: open %s: %w", driver, err)
-	}
-	ctx, cancel := context.WithTimeout(context.Background(), pgConnectTimeout)
-	defer cancel()
-	if err := db.PingContext(ctx); err != nil {
-		db.Close()
-		return nil, fmt.Errorf("onlyoffice: sql file store: ping %s: %w", driver, err)
-	}
-	s := &pgStore{db: db, driver: driver, http: &http.Client{}}
-	if t := strings.TrimSpace(cfg.Tenant); t != "" {
-		n, err := strconv.ParseInt(t, 10, 64)
-		if err != nil {
-			db.Close()
-			return nil, fmt.Errorf("onlyoffice: sql file store: non-numeric tenant %q", t)
-		}
-		s.tenantID, s.hasTenant = n, true
-	}
-	return s, nil
-}
-
-// Close releases the database handle.
-func (s *pgStore) Close() error { return s.db.Close() }
-
-// Name implements FileStore and Searcher.
-func (s *pgStore) Name() string { return s.driver }
-
-// pgDriver resolves the engine: the explicit value wins, otherwise the DSN
-// shape decides. A leading postgres:// scheme or a libpq keyword DSN (which
-// always carries '=') selects PostgreSQL; anything else is MySQL.
-func pgDriver(dsn, explicit string) string {
-	switch strings.ToLower(strings.TrimSpace(explicit)) {
-	case ProviderPG, "pg", "postgresql", "pgx":
-		return ProviderPG
-	case ProviderMySQL, "mariadb":
-		return ProviderMySQL
-	}
-	l := strings.ToLower(strings.TrimSpace(dsn))
-	switch {
-	case strings.HasPrefix(l, "postgres://"), strings.HasPrefix(l, "postgresql://"):
-		return ProviderPG
-	case strings.HasPrefix(l, "mysql://"), strings.Contains(l, "@tcp("), strings.Contains(l, "@unix("):
-		return ProviderMySQL
-	case strings.Contains(l, "="):
-		return ProviderPG
-	default:
-		return ProviderMySQL
-	}
-}
-
-// sqlDriverName maps the engine to its registered database/sql driver.
-func sqlDriverName(driver string) string {
-	if driver == ProviderPG {
-		return "pgx"
-	}
-	return "mysql"
-}
-
-// normalizeSQLDSN converts a mysql:// URL to the go-sql-driver form and forces
-// parseTime so datetime columns scan into time.Time. PostgreSQL DSNs pass
-// through untouched.
-func normalizeSQLDSN(driver, dsn string) (string, error) {
-	if driver != ProviderMySQL {
-		return dsn, nil
-	}
-	if strings.HasPrefix(strings.ToLower(dsn), "mysql://") {
-		converted, err := mysqlDSNFromURL(dsn)
-		if err != nil {
-			return "", err
-		}
-		dsn = converted
-	}
-	cfg, err := mysql.ParseDSN(dsn)
-	if err != nil {
-		return "", fmt.Errorf("onlyoffice: sql file store: parse mysql DSN: %w", err)
-	}
-	cfg.ParseTime = true
-	return cfg.FormatDSN(), nil
-}
-
-// mysqlDSNFromURL turns mysql://user:pass@host:port/db into the driver DSN.
-func mysqlDSNFromURL(raw string) (string, error) {
-	u, err := url.Parse(raw)
-	if err != nil || u.Host == "" {
-		return "", fmt.Errorf("onlyoffice: sql file store: bad mysql URL %q", raw)
-	}
-	user := ""
-	if u.User != nil {
-		user = u.User.Username()
-		if p, ok := u.User.Password(); ok {
-			user += ":" + p
-		}
-	}
-	q := u.Query()
-	q.Set("parseTime", "true")
-	return fmt.Sprintf("%s@tcp(%s)/%s?%s", user, u.Host, strings.TrimPrefix(u.Path, "/"), q.Encode()), nil
-}
-
-// rebind rewrites '?' placeholders to PostgreSQL's $1..$n. MySQL keeps them.
-func rebind(query, driver string) string {
-	if driver != ProviderPG {
-		return query
-	}
-	var b strings.Builder
-	b.Grow(len(query) + 8)
-	n := 0
-	for _, r := range query {
-		if r == '?' {
-			n++
-			b.WriteByte('$')
-			b.WriteString(strconv.Itoa(n))
-			continue
-		}
-		b.WriteRune(r)
-	}
-	return b.String()
-}
-
-// List returns the folders and files directly below parentID, folders first.
-func (s *pgStore) List(ctx context.Context, parentID string) ([]Entry, error) {
-	pid, err := parseEntryID(parentID)
-	if err != nil {
-		return nil, err
-	}
-	folders, err := s.queryFolders(ctx, "parent_id = ?", pid)
-	if err != nil {
-		return nil, err
-	}
-	files, err := s.queryFiles(ctx, "folder_id = ? AND current_version = 1", pid)
-	if err != nil {
-		return nil, err
-	}
-	out := make([]Entry, 0, len(folders)+len(files))
-	for _, f := range folders {
-		out = append(out, folderRowToEntry(f, s.Name()))
-	}
-	for _, f := range files {
-		out = append(out, fileRowToEntry(f, s.Name()))
-	}
-	return out, nil
-}
-
-// Stat resolves a folder or file id to an Entry. Folders win when both id
-// spaces overlap (they never do on a real portal, but the lookup is cheap).
-func (s *pgStore) Stat(ctx context.Context, id string) (Entry, error) {
-	n, err := parseEntryID(id)
-	if err != nil {
-		return Entry{}, err
-	}
-	folders, err := s.queryFolders(ctx, "id = ?", n)
-	if err != nil {
-		return Entry{}, err
-	}
-	if len(folders) > 0 {
-		return folderRowToEntry(folders[0], s.Name()), nil
-	}
-	files, err := s.queryFiles(ctx, "id = ? AND current_version = 1", n)
-	if err != nil {
-		return Entry{}, err
-	}
-	if len(files) == 0 {
-		return Entry{}, fmt.Errorf("onlyoffice: sql file store: id %s not found", id)
-	}
-	return fileRowToEntry(files[0], s.Name()), nil
-}
-
-// Download streams the file's current version from the portal's S3/MinIO store.
-// The object key is reconstructed from the file id and version; the parent
-// folder id is not part of the key.
-func (s *pgStore) Download(ctx context.Context, id string, w io.Writer) (int64, error) {
-	n, err := parseEntryID(id)
-	if err != nil {
-		return 0, err
-	}
-	files, err := s.queryFiles(ctx, "id = ? AND current_version = 1", n)
-	if err != nil {
-		return 0, err
-	}
-	if len(files) == 0 {
-		return 0, fmt.Errorf("onlyoffice: sql file store: file %s not found", id)
-	}
-	f := files[0]
-	key := csObjectKey(s.tenantID, f.id, f.version, filepath.Ext(f.title))
-	return downloadMinioObject(ctx, s.http, key, w)
-}
-
-// CreateFolder is unavailable: the SQL backend is read-only.
-func (s *pgStore) CreateFolder(context.Context, string, string) (Entry, error) {
-	return Entry{}, ErrReadOnly
-}
-
-// Upload is unavailable: the SQL backend is read-only.
-func (s *pgStore) Upload(context.Context, string, string, io.Reader) (Entry, error) {
-	return Entry{}, ErrReadOnly
-}
-
-// Move is unavailable: the SQL backend is read-only.
-func (s *pgStore) Move(context.Context, []string, string) error { return ErrReadOnly }
-
-// Copy is unavailable: the SQL backend is read-only.
-func (s *pgStore) Copy(context.Context, []string, string) error { return ErrReadOnly }
-
-// Rename is unavailable: the SQL backend is read-only.
-func (s *pgStore) Rename(context.Context, string, string) error { return ErrReadOnly }
-
-// Delete is unavailable: the SQL backend is read-only.
-func (s *pgStore) Delete(context.Context, []string) error { return ErrReadOnly }
-
-// Search matches file titles by substring. Content search lives in the
-// Elasticsearch backend; q.InContent is ignored here.
-func (s *pgStore) Search(ctx context.Context, q SearchQuery) ([]SearchHit, error) {
-	text := strings.TrimSpace(q.Text)
-	if text == "" {
-		return nil, fmt.Errorf("onlyoffice: empty search query")
-	}
-	limit := q.Limit
-	if limit <= 0 {
-		limit = pgSearchLimit
-	}
-	if limit > pgSearchMaxLimit {
-		limit = pgSearchMaxLimit
-	}
-
-	where := "title LIKE ? AND current_version = 1"
-	args := []any{"%" + text + "%"}
-	if s.hasTenant {
-		where += " AND tenant_id = ?"
-		args = append(args, s.tenantID)
-	}
-	if fid := strings.TrimSpace(q.FolderID); fid != "" {
-		n, err := parseEntryID(fid)
-		if err != nil {
-			return nil, err
-		}
-		where += " AND folder_id = ?"
-		args = append(args, n)
-	}
-	for _, ext := range normalizeExtensions(q.Extensions) {
-		where += " AND LOWER(title) LIKE ?"
-		args = append(args, "%."+ext)
-	}
-	query := rebind(`SELECT id, folder_id, title, content_length, version, create_on, modified_on
-        FROM files_file WHERE `+where+` ORDER BY modified_on DESC, id DESC LIMIT ?`, s.driver)
-	args = append(args, limit)
-
-	rows, err := s.db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, fmt.Errorf("onlyoffice: sql search: %w", err)
-	}
-	defer rows.Close()
-	var hits []SearchHit
-	for rows.Next() {
-		f, err := scanFileRow(rows)
-		if err != nil {
-			return nil, err
-		}
-		hits = append(hits, SearchHit{Entry: fileRowToEntry(f, s.Name())})
-	}
-	return hits, rows.Err()
-}
-
-// queryFolders runs a folder SELECT with the tenant filter applied.
-func (s *pgStore) queryFolders(ctx context.Context, where string, arg any) ([]pgFolderRow, error) {
-	args := []any{arg}
-	if s.hasTenant {
-		where += " AND tenant_id = ?"
-		args = append(args, s.tenantID)
-	}
-	query := rebind(`SELECT id, parent_id, title, create_on, modified_on
-        FROM files_folder WHERE `+where+` ORDER BY title, id`, s.driver)
-	rows, err := s.db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, fmt.Errorf("onlyoffice: sql list folders: %w", err)
-	}
-	defer rows.Close()
-	var out []pgFolderRow
-	for rows.Next() {
-		var r pgFolderRow
-		if err := rows.Scan(&r.id, &r.parentID, &r.title, &r.created, &r.modified); err != nil {
-			return nil, fmt.Errorf("onlyoffice: sql folder row: %w", err)
-		}
-		out = append(out, r)
-	}
-	return out, rows.Err()
-}
-
-// queryFiles runs a file SELECT for the current version with the tenant filter.
-func (s *pgStore) queryFiles(ctx context.Context, where string, arg any) ([]pgFileRow, error) {
-	args := []any{arg}
-	if s.hasTenant {
-		where += " AND tenant_id = ?"
-		args = append(args, s.tenantID)
-	}
-	query := rebind(`SELECT id, folder_id, title, content_length, version, create_on, modified_on
-        FROM files_file WHERE `+where+` ORDER BY title, id`, s.driver)
-	rows, err := s.db.QueryContext(ctx, query, args...)
-	if err != nil {
-		return nil, fmt.Errorf("onlyoffice: sql list files: %w", err)
-	}
-	defer rows.Close()
-	var out []pgFileRow
-	for rows.Next() {
-		f, err := scanFileRow(rows)
-		if err != nil {
-			return nil, err
-		}
-		out = append(out, f)
-	}
-	return out, rows.Err()
-}
-
-// pgFileRow is one current files_file row.
-type pgFileRow struct {
-	id       int64
-	folderID int64
-	title    string
-	size     int64
-	version  int
-	created  time.Time
-	modified time.Time
-}
-
-// pgFolderRow is one files_folder row.
-type pgFolderRow struct {
-	id       int64
-	parentID int64
-	title    string
-	created  time.Time
-	modified time.Time
-}
-
-// scanFileRow reads the canonical file column order.
-func scanFileRow(rows *sql.Rows) (pgFileRow, error) {
-	var f pgFileRow
-	if err := rows.Scan(&f.id, &f.folderID, &f.title, &f.size, &f.version, &f.created, &f.modified); err != nil {
-		return f, fmt.Errorf("onlyoffice: sql file row: %w", err)
-	}
-	return f, nil
-}
-
-// fileRowToEntry maps a files_file row to the canonical model.
-func fileRowToEntry(f pgFileRow, provider string) Entry {
-	return Entry{
-		ID:       strconv.FormatInt(f.id, 10),
-		ParentID: strconv.FormatInt(f.folderID, 10),
-		Title:    f.title,
-		Kind:     File,
-		Size:     f.size,
-		MIME:     mimeForTitle(f.title, ""),
-		Created:  f.created.UTC(),
-		Modified: f.modified.UTC(),
-		Version:  f.version,
-		Provider: provider,
-	}
-}
-
-// folderRowToEntry maps a files_folder row to the canonical model.
-func folderRowToEntry(f pgFolderRow, provider string) Entry {
-	return Entry{
-		ID:       strconv.FormatInt(f.id, 10),
-		ParentID: strconv.FormatInt(f.parentID, 10),
-		Title:    f.title,
-		Kind:     Folder,
-		Created:  f.created.UTC(),
-		Modified: f.modified.UTC(),
-		Provider: provider,
-	}
-}
-
-// csObjectKey reconstructs the object key the portal's S3 consumer uses:
-//
-//	00/00/<tenant>/files/folder_<shard>/file_<id>/v<version>/content.<ext>
-//
-// The shard is the next thousand above the file id (file 3727 -> folder_4000),
-// NOT the parent folder id — verified live against the MinIO bucket.
-func csObjectKey(tenant int64, fileID int64, version int, ext string) string {
-	shard := (fileID/1000 + 1) * 1000
-	ext = strings.TrimPrefix(strings.ToLower(strings.TrimSpace(ext)), ".")
-	if ext == "" {
-		ext = "bin"
-	}
-	if version < 1 {
-		version = 1
-	}
-	if tenant <= 0 {
-		tenant = 1
-	}
-	return fmt.Sprintf("00/00/%02d/files/folder_%d/file_%d/v%d/content.%s", tenant, shard, fileID, version, ext)
-}
-
-// parseEntryID parses a numeric OnlyOffice id or returns a store error.
-func parseEntryID(id string) (int64, error) {
-	n, err := strconv.ParseInt(strings.TrimSpace(id), 10, 64)
-	if err != nil {
-		return 0, fmt.Errorf("onlyoffice: sql file store: non-numeric id %q", id)
-	}
-	return n, nil
-}

file_pg_integration_test.go

@@ -1,139 +0,0 @@
-//go:build integration
-
-package onlyoffice
-
-import (
-	"bytes"
-	"context"
-	"errors"
-	"os"
-	"strings"
-	"testing"
-	"time"
-)
-
-// TestIntegrationPGStore exercises the read-only SQL backend against the live
-// Community Server database and cross-checks list/stat/download with the REST
-// FileStore. It needs ONLYOFFICE_DSN plus the usual ONLYOFFICE_URL/USER/PASS;
-// ONLYOFFICE_PG_TEST_FILE_ID / ONLYOFFICE_PG_TEST_FOLDER_ID pick a real file
-// (a file reachable over REST too). Download streams from MinIO, so it also
-// needs MINIO_ACCESS_KEY/MINIO_SECRET_KEY.
-//
-// The live Community Server runs on MySQL; PostgreSQL is supported by the same
-// code path when the DSN says so.
-func TestIntegrationPGStore(t *testing.T) {
-	cfg := PGConfigFromEnv()
-	if strings.TrimSpace(cfg.DSN) == "" {
-		t.Skip("ONLYOFFICE_DSN not set — skipping SQL store integration test")
-	}
-	store, err := NewPGStore(cfg)
-	if err != nil {
-		t.Fatalf("NewPGStore: %v", err)
-	}
-	t.Cleanup(func() { _ = store.Close() })
-	t.Logf("sql store backend: %s", store.Name())
-	ctx := context.Background()
-
-	if err := testPGStoreReadOnly(ctx, store); err != nil {
-		t.Fatal(err)
-	}
-
-	fileID := strings.TrimSpace(os.Getenv("ONLYOFFICE_PG_TEST_FILE_ID"))
-	folderID := strings.TrimSpace(os.Getenv("ONLYOFFICE_PG_TEST_FOLDER_ID"))
-	if fileID == "" || folderID == "" {
-		t.Skip("ONLYOFFICE_PG_TEST_FILE_ID / ONLYOFFICE_PG_TEST_FOLDER_ID not set — skipping live comparison")
-	}
-
-	c := liveClient(t)
-	rest := c.Files()
-
-	dbFile, err := store.Stat(ctx, fileID)
-	if err != nil {
-		t.Fatalf("sql Stat(%s): %v", fileID, err)
-	}
-	restFile, err := rest.Stat(ctx, fileID)
-	if err != nil {
-		t.Fatalf("rest Stat(%s): %v", fileID, err)
-	}
-	if dbFile.Kind != File {
-		t.Errorf("sql kind = %v, want file", dbFile.Kind)
-	}
-	if dbFile.ID != restFile.ID || dbFile.Title != restFile.Title || dbFile.ParentID != restFile.ParentID {
-		t.Errorf("stat mismatch sql=%+v rest=%+v", dbFile, restFile)
-	}
-	// GetFile omits contentLength, so size is only comparable when REST has it.
-	if restFile.Size > 0 && dbFile.Size != restFile.Size {
-		t.Errorf("size sql=%d rest=%d", dbFile.Size, restFile.Size)
-	}
-	if d := dbFile.Modified.Sub(restFile.Modified); d > 2*time.Minute || d < -2*time.Minute {
-		t.Errorf("modified sql=%v rest=%v", dbFile.Modified, restFile.Modified)
-	}
-
-	list, err := store.List(ctx, folderID)
-	if err != nil {
-		t.Fatalf("sql List(%s): %v", folderID, err)
-	}
-	if entryByID(list, fileID) == nil {
-		t.Errorf("file %s not in sql List(%s)", fileID, folderID)
-	}
-
-	// Every file the REST layer can see in the folder must be in the SQL list
-	// (the SQL store sees more, so only assert this direction).
-	restList, err := rest.List(ctx, folderID)
-	if err != nil {
-		t.Fatalf("rest List(%s): %v", folderID, err)
-	}
-	dbIDs := make(map[string]bool, len(list))
-	for _, e := range list {
-		dbIDs[e.ID] = true
-	}
-	for _, e := range restList {
-		if e.Kind == File && !dbIDs[e.ID] {
-			t.Errorf("rest file %s (%q) missing from sql list", e.ID, e.Title)
-		}
-	}
-
-	var buf bytes.Buffer
-	n, err := store.Download(ctx, fileID, &buf)
-	if err != nil {
-		t.Fatalf("sql Download(%s): %v", fileID, err)
-	}
-	if n == 0 || n != dbFile.Size {
-		t.Errorf("sql Download = %d bytes, stat says %d", n, dbFile.Size)
-	}
-	if os.Getenv("MINIO_ACCESS_KEY") != "" && os.Getenv("MINIO_SECRET_KEY") != "" {
-		var restBuf bytes.Buffer
-		rn, err := rest.Download(ctx, fileID, &restBuf)
-		if err != nil {
-			t.Fatalf("rest Download(%s): %v", fileID, err)
-		}
-		if rn != n || !bytes.Equal(restBuf.Bytes(), buf.Bytes()) {
-			t.Errorf("download mismatch sql=%d rest=%d bytes", n, rn)
-		}
-	} else {
-		t.Log("MINIO_ACCESS_KEY/MINIO_SECRET_KEY not set — REST download cross-check skipped")
-	}
-}
-
-// testPGStoreReadOnly asserts that every write method returns ErrReadOnly.
-func testPGStoreReadOnly(ctx context.Context, s *pgStore) error {
-	if _, err := s.CreateFolder(ctx, "1", "x"); !errors.Is(err, ErrReadOnly) {
-		return errors.New("CreateFolder did not return ErrReadOnly")
-	}
-	if _, err := s.Upload(ctx, "1", "x", strings.NewReader("x")); !errors.Is(err, ErrReadOnly) {
-		return errors.New("Upload did not return ErrReadOnly")
-	}
-	if err := s.Move(ctx, []string{"1"}, "2"); !errors.Is(err, ErrReadOnly) {
-		return errors.New("Move did not return ErrReadOnly")
-	}
-	if err := s.Copy(ctx, []string{"1"}, "2"); !errors.Is(err, ErrReadOnly) {
-		return errors.New("Copy did not return ErrReadOnly")
-	}
-	if err := s.Rename(ctx, "1", "x"); !errors.Is(err, ErrReadOnly) {
-		return errors.New("Rename did not return ErrReadOnly")
-	}
-	if err := s.Delete(ctx, []string{"1"}); !errors.Is(err, ErrReadOnly) {
-		return errors.New("Delete did not return ErrReadOnly")
-	}
-	return nil
-}

file_pg_test.go

@@ -1,165 +0,0 @@
-package onlyoffice
-
-import (
-	"context"
-	"errors"
-	"testing"
-	"time"
-)
-
-func TestRebind(t *testing.T) {
-	mysqlQuery := "SELECT id FROM files_file WHERE folder_id = ? AND title = ? LIMIT ?"
-	if got := rebind(mysqlQuery, ProviderMySQL); got != mysqlQuery {
-		t.Errorf("mysql query changed: %q", got)
-	}
-	want := "SELECT id FROM files_file WHERE folder_id = $1 AND title = $2 LIMIT $3"
-	if got := rebind(mysqlQuery, ProviderPG); got != want {
-		t.Errorf("rebind = %q, want %q", got, want)
-	}
-}
-
-func TestCSPObjectKey(t *testing.T) {
-	cases := []struct {
-		tenant  int64
-		fileID  int64
-		version int
-		ext     string
-		want    string
-	}{
-		{1, 2, 1, ".docx", "00/00/01/files/folder_1000/file_2/v1/content.docx"},
-		{1, 999, 1, ".pdf", "00/00/01/files/folder_1000/file_999/v1/content.pdf"},
-		{1, 1000, 1, ".xlsx", "00/00/01/files/folder_2000/file_1000/v1/content.xlsx"},
-		{1, 3727, 1, ".pdf", "00/00/01/files/folder_4000/file_3727/v1/content.pdf"},
-		{1, 22484, 1, ".PDF", "00/00/01/files/folder_23000/file_22484/v1/content.pdf"},
-		{1, 4, 6, "xlsx", "00/00/01/files/folder_1000/file_4/v6/content.xlsx"},
-		{0, 7, 0, "", "00/00/01/files/folder_1000/file_7/v1/content.bin"},
-		{2, 11, 3, ".doc", "00/00/02/files/folder_1000/file_11/v3/content.doc"},
-	}
-	for _, tc := range cases {
-		if got := csObjectKey(tc.tenant, tc.fileID, tc.version, tc.ext); got != tc.want {
-			t.Errorf("csObjectKey(%d,%d,%d,%q) = %q, want %q", tc.tenant, tc.fileID, tc.version, tc.ext, got, tc.want)
-		}
-	}
-}
-
-func TestPGDriverDetection(t *testing.T) {
-	cases := []struct {
-		dsn, explicit, want string
-	}{
-		{"postgres://u:p@h:5432/onlyoffice", "", ProviderPG},
-		{"postgresql://u:p@h/db", "", ProviderPG},
-		{"host=h user=u password=p dbname=onlyoffice sslmode=disable", "", ProviderPG},
-		{"root:secret@tcp(127.0.0.1:3306)/onlyoffice?parseTime=true", "", ProviderMySQL},
-		{"mysql://root:secret@127.0.0.1:3306/onlyoffice", "", ProviderMySQL},
-		{"root:secret@tcp(h:3306)/db", "postgres", ProviderPG},
-		{"postgres://u:p@h/db", "mysql", ProviderMySQL},
-	}
-	for _, tc := range cases {
-		if got := pgDriver(tc.dsn, tc.explicit); got != tc.want {
-			t.Errorf("pgDriver(%q, %q) = %q, want %q", tc.dsn, tc.explicit, got, tc.want)
-		}
-	}
-}
-
-func TestNormalizeSQLDSNMySQL(t *testing.T) {
-	got, err := normalizeSQLDSN(ProviderMySQL, "mysql://root:secret@127.0.0.1:3306/onlyoffice")
-	if err != nil {
-		t.Fatalf("normalizeSQLDSN: %v", err)
-	}
-	want := "root:secret@tcp(127.0.0.1:3306)/onlyoffice?parseTime=true"
-	if got != want {
-		t.Errorf("normalize = %q, want %q", got, want)
-	}
-
-	// A driver DSN keeps parseTime and gains it when missing.
-	got, err = normalizeSQLDSN(ProviderMySQL, "root:secret@tcp(127.0.0.1:3306)/onlyoffice")
-	if err != nil {
-		t.Fatalf("normalizeSQLDSN: %v", err)
-	}
-	if got != want {
-		t.Errorf("normalize = %q, want %q", got, want)
-	}
-}
-
-func TestFileRowToEntry(t *testing.T) {
-	created := time.Date(2026, 9, 12, 18, 0, 37, 0, time.UTC)
-	modified := time.Date(2026, 9, 13, 13, 50, 36, 0, time.UTC)
-	e := fileRowToEntry(pgFileRow{
-		id: 22484, folderID: 649, title: "Rechnung.pdf",
-		size: 123433, version: 2, created: created, modified: modified,
-	}, ProviderMySQL)
-	if e.ID != "22484" || e.ParentID != "649" {
-		t.Errorf("ids = %q/%q", e.ID, e.ParentID)
-	}
-	if e.Title != "Rechnung.pdf" || e.Kind != File {
-		t.Errorf("title/kind = %q/%v", e.Title, e.Kind)
-	}
-	if e.Size != 123433 || e.Version != 2 {
-		t.Errorf("size/version = %d/%d", e.Size, e.Version)
-	}
-	if e.MIME != "application/pdf" {
-		t.Errorf("mime = %q", e.MIME)
-	}
-	if !e.Created.Equal(created) || !e.Modified.Equal(modified) {
-		t.Errorf("times = %v/%v", e.Created, e.Modified)
-	}
-	if e.Provider != ProviderMySQL {
-		t.Errorf("provider = %q", e.Provider)
-	}
-}
-
-func TestFolderRowToEntry(t *testing.T) {
-	modified := time.Date(2026, 8, 1, 10, 30, 0, 0, time.UTC)
-	e := folderRowToEntry(pgFolderRow{id: 649, parentID: 647, title: "2025", modified: modified}, ProviderMySQL)
-	if e.ID != "649" || e.ParentID != "647" || e.Title != "2025" {
-		t.Errorf("folder = %+v", e)
-	}
-	if e.Kind != Folder {
-		t.Errorf("kind = %v, want folder", e.Kind)
-	}
-	if e.Size != 0 || e.MIME != "" {
-		t.Errorf("folder size/mime = %d/%q", e.Size, e.MIME)
-	}
-	if !e.Modified.Equal(modified) {
-		t.Errorf("modified = %v", e.Modified)
-	}
-}
-
-func TestPGStoreWriteMethodsReadOnly(t *testing.T) {
-	s := &pgStore{driver: ProviderPG}
-	ctx := context.Background()
-	if _, err := s.CreateFolder(ctx, "1", "x"); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("CreateFolder err = %v", err)
-	}
-	if _, err := s.Upload(ctx, "1", "x", nil); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("Upload err = %v", err)
-	}
-	if err := s.Move(ctx, nil, "1"); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("Move err = %v", err)
-	}
-	if err := s.Copy(ctx, nil, "1"); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("Copy err = %v", err)
-	}
-	if err := s.Rename(ctx, "1", "x"); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("Rename err = %v", err)
-	}
-	if err := s.Delete(ctx, nil); !errors.Is(err, ErrReadOnly) {
-		t.Errorf("Delete err = %v", err)
-	}
-}
-
-func TestPGStoreName(t *testing.T) {
-	if got := (&pgStore{driver: ProviderPG}).Name(); got != ProviderPG {
-		t.Errorf("Name = %q, want %q", got, ProviderPG)
-	}
-	if got := (&pgStore{driver: ProviderMySQL}).Name(); got != ProviderMySQL {
-		t.Errorf("Name = %q, want %q", got, ProviderMySQL)
-	}
-}
-
-func TestPGStoreStatRejectsNonNumeric(t *testing.T) {
-	s := &pgStore{driver: ProviderPG}
-	if _, err := s.Stat(context.Background(), "not-a-number"); err == nil {
-		t.Error("Stat accepted a non-numeric id")
-	}
-}

file_text_index.go

@@ -3,10 +3,9 @@ package onlyoffice
 // Text extraction pipeline for the own full-text index (epic #34, F6 #42).
 //
 // TextIndexer downloads stored documents, extracts text through docpipe
-// (pdftotext; OCR for scans) and writes the result to a TextIndex. For PDFs it
+// (pdftotext; OCR for scans) and writes the result to a TextIndex. It is the
-// also indexes the text of embedded attachments (pdfdetach), so a scan filed
+// write side of ESTextIndex and never touches the OnlyOffice server's own ES
-// as an attachment is searchable too. It is the write side of ESTextIndex and
+// index.
-// never touches the OnlyOffice server's own ES index.
 
 import (
 	"context"
@@ -39,14 +38,13 @@ type TextExtractor interface {
 type docpipeExtractor struct{ tools docpipe.Tools }
 
 // Extract renders the file as Markdown, OCRing PDFs/images with a weak text
-// layer first and appending the text of embedded PDF attachments
+// layer first (docpipe.ToMarkdown).
-// (docpipe.ToMarkdownWithAttachments).
 func (d docpipeExtractor) Extract(path, workDir, lang string, minChars int) (string, error) {
-	text, err := d.tools.ToMarkdownWithAttachments(path, workDir, lang, minChars)
+	res, err := d.tools.ToMarkdown(path, workDir, lang, minChars)
 	if err != nil {
 		return "", err
 	}
-	return strings.TrimSpace(text), nil
+	return strings.TrimSpace(res.Markdown), nil
 }
 
 // IndexOptions controls a TextIndexer run.

files.go

@@ -247,8 +247,6 @@ func (c *Client) UploadProjectFileReplacing(ctx context.Context, projectID, loca
 }
 
 // GetFile returns file metadata including viewUrl for download.
-//
-// Deprecated: use FileStore.Stat via Client.Files()/Client.FileStore.
 func (c *Client) GetFile(ctx context.Context, fileID string) (*FileEntry, error) {
 	if fileID == "" {
 		return nil, fmt.Errorf("file id is required")
@@ -262,8 +260,6 @@ func (c *Client) GetFile(ctx context.Context, fileID string) (*FileEntry, error)
 }
 
 // RenameFile sets a new title (including extension) for the file.
-//
-// Deprecated: use FileStore.Rename via Client.Files()/Client.FileStore.
 func (c *Client) RenameFile(ctx context.Context, fileID, newTitle string) (*FileEntry, error) {
 	if fileID == "" || newTitle == "" {
 		return nil, fmt.Errorf("file id and new title are required")
@@ -279,8 +275,6 @@ func (c *Client) RenameFile(ctx context.Context, fileID, newTitle string) (*File
 // DeleteFiles permanently deletes files by numeric id (Documents module).
 // Uses per-file DELETE (DeleteDavItems); fileops/delete returns 200 on some
 // portals (e.g. produktor.io) without removing the file.
-//
-// Deprecated: use FileStore.Delete via Client.Files()/Client.FileStore.
 func (c *Client) DeleteFiles(ctx context.Context, fileIDs []int) error {
 	if len(fileIDs) == 0 {
 		return fmt.Errorf("no file ids to delete")
@@ -294,8 +288,6 @@ func (c *Client) DeleteFiles(ctx context.Context, fileIDs []int) error {
 
 // ListFolder returns the Documents module listing for a folder id
 // (GET /api/2.0/files/{folderId}).
-//
-// Deprecated: use FileStore.List via Client.Files()/Client.FileStore.
 func (c *Client) ListFolder(ctx context.Context, folderID string) (map[string]any, error) {
 	if folderID == "" {
 		return nil, fmt.Errorf("folder id is required")
@@ -321,8 +313,6 @@ func (c *Client) CreateFolder(ctx context.Context, parentFolderID, title string)
 }
 
 // MoveFiles moves file ids into destFolderID (Documents fileops/move).
-//
-// Deprecated: use FileStore.Move via Client.Files()/Client.FileStore.
 func (c *Client) MoveFiles(ctx context.Context, destFolderID int, fileIDs []int) (map[string]any, error) {
 	if destFolderID == 0 || len(fileIDs) == 0 {
 		return nil, fmt.Errorf("dest folder and file ids are required")
@@ -353,8 +343,6 @@ func (c *Client) MoveFiles(ctx context.Context, destFolderID int, fileIDs []int)
 }
 
 // UploadToFolder uploads a local file into an arbitrary Documents folder id.
-//
-// Deprecated: use FileStore.Upload via Client.Files()/Client.FileStore.
 func (c *Client) UploadToFolder(ctx context.Context, folderID, localPath string) (*FileEntry, error) {
 	if folderID == "" || localPath == "" {
 		return nil, fmt.Errorf("folder id and local path are required")
@@ -412,8 +400,6 @@ func FileFolderID(f *FileEntry) string {
 // as API calls. Writes into dst. When the portal serves the file from its stale
 // AWS S3 consumer, the bytes are fetched from the local MinIO store instead
 // (see storage_fallback.go).
-//
-// Deprecated: use FileStore.Download via Client.Files()/Client.FileStore.
 func (c *Client) DownloadFile(ctx context.Context, fileID string, dst io.Writer) (int64, error) {
 	f, err := c.GetFile(ctx, fileID)
 	if err != nil {

files_webdav.go

@@ -52,8 +52,6 @@ type DavListing struct {
 
 // ListDavFolder returns the contents of a folder by id, which may be a
 // symbolic root such as "@my". For "@root" use ListDavSections.
-//
-// Deprecated: use FileStore.List via Client.Files()/Client.FileStore.
 func (c *Client) ListDavFolder(ctx context.Context, id string) (*DavListing, error) {
 	raw, err := c.getJSON(ctx, "/api/2.0/files/"+url.PathEscape(id))
 	if err != nil {
@@ -113,8 +111,6 @@ func (c *Client) ListDavSections(ctx context.Context) ([]DavFolder, error) {
 }
 
 // CreateDavFolder creates a folder titled title inside parentID.
-//
-// Deprecated: use FileStore.CreateFolder via Client.Files()/Client.FileStore.
 func (c *Client) CreateDavFolder(ctx context.Context, parentID, title string) (*DavFolder, error) {
 	raw, err := c.postJSON(ctx, "/api/2.0/files/folder/"+url.PathEscape(parentID),
 		map[string]string{"title": title})
@@ -134,8 +130,6 @@ func (c *Client) CreateDavFolder(ctx context.Context, parentID, title string) (*
 }
 
 // RenameDavFolder renames a folder.
-//
-// Deprecated: use FileStore.Rename via Client.Files()/Client.FileStore.
 func (c *Client) RenameDavFolder(ctx context.Context, id, title string) error {
 	_, err := c.putJSON(ctx, "/api/2.0/files/folder/"+url.PathEscape(id),
 		map[string]string{"title": title})
@@ -143,8 +137,6 @@ func (c *Client) RenameDavFolder(ctx context.Context, id, title string) error {
 }
 
 // RenameDavFile renames a file (title includes the extension).
-//
-// Deprecated: use FileStore.Rename via Client.Files()/Client.FileStore.
 func (c *Client) RenameDavFile(ctx context.Context, id, title string) error {
 	_, err := c.putJSON(ctx, "/api/2.0/files/file/"+url.PathEscape(id),
 		map[string]string{"title": title})
@@ -155,8 +147,6 @@ func (c *Client) RenameDavFile(ctx context.Context, id, title string) error {
 // The fileops API answers 200 with per-operation error strings even when
 // nothing moves (e.g. missing permission), so the response is parsed and the
 // first operation error is returned instead of a silent nil.
-//
-// Deprecated: use FileStore.Move via Client.Files()/Client.FileStore.
 func (c *Client) MoveDavItems(ctx context.Context, folderIDs, fileIDs []string, destFolderID string) error {
 	raw, err := c.putJSON(ctx, "/api/2.0/files/fileops/move", map[string]any{
 		"folderIds":    nums(folderIDs),
@@ -173,8 +163,6 @@ func (c *Client) MoveDavItems(ctx context.Context, folderIDs, fileIDs []string,
 
 // CopyDavItems copies the given folders and/or files into destFolderID.
 // Per-operation errors are surfaced like in MoveDavItems.
-//
-// Deprecated: use FileStore.Copy via Client.Files()/Client.FileStore.
 func (c *Client) CopyDavItems(ctx context.Context, folderIDs, fileIDs []string, destFolderID string) error {
 	raw, err := c.putJSON(ctx, "/api/2.0/files/fileops/copy", map[string]any{
 		"folderIds":           nums(folderIDs),
@@ -238,8 +226,6 @@ func fileopsError(raw json.RawMessage) error {
 }
 
 // DeleteDavItems deletes the given folders and/or files.
-//
-// Deprecated: use FileStore.Delete via Client.Files()/Client.FileStore.
 func (c *Client) DeleteDavItems(ctx context.Context, folderIDs, fileIDs []string) error {
 	body := map[string]any{"DeleteAfter": true, "Immediately": true}
 	for _, id := range folderIDs {
@@ -256,8 +242,6 @@ func (c *Client) DeleteDavItems(ctx context.Context, folderIDs, fileIDs []string
 }
 
 // UploadDavFile uploads src (fileName) into folderID, streaming from src.
-//
-// Deprecated: use FileStore.Upload via Client.Files()/Client.FileStore.
 func (c *Client) UploadDavFile(ctx context.Context, folderID, fileName string, src io.Reader) (*DavFile, error) {
 	raw, err := c.uploadReader(ctx, "/api/2.0/files/"+url.PathEscape(folderID)+"/upload", "file", fileName, src)
 	if err != nil {
@@ -277,8 +261,6 @@ func (c *Client) UploadDavFile(ctx context.Context, folderID, fileName string, s
 
 // DownloadDavFile streams the file identified by id to w, returning bytes
 // copied. It shares the MinIO stale-S3 fallback with DownloadFile.
-//
-// Deprecated: use FileStore.Download via Client.Files()/Client.FileStore.
 func (c *Client) DownloadDavFile(ctx context.Context, id string, w io.Writer) (int64, error) {
 	return c.DownloadFile(ctx, id, w)
 }

go.mod

@@ -13,9 +13,7 @@ require (
 	github.com/emersion/go-vcard v0.0.0-20260618161152-d854b7e0e2d3
 	github.com/eslider/go-hocr v0.2.2-0.20260827163626-8ff01582b002
 	github.com/eslider/go-xls/v2 v2.1.0
-	github.com/go-sql-driver/mysql v1.10.1
 	github.com/google/go-querystring v1.2.0
-	github.com/jackc/pgx/v5 v5.11.0
 	github.com/joho/godotenv v1.5.1
 	github.com/mattn/go-runewidth v0.0.15
 	github.com/muesli/termenv v0.16.0
@@ -26,7 +24,6 @@ require (
 )
 
 require (
-	filippo.io/edwards25519 v1.2.0 // indirect
 	github.com/JohannesKaufmann/dom v0.3.1 // indirect
 	github.com/alecthomas/chroma/v2 v2.14.0 // indirect
 	github.com/atotto/clipboard v0.1.4 // indirect
@@ -39,10 +36,6 @@ require (
 	github.com/google/uuid v1.6.0 // indirect
 	github.com/gorilla/css v1.0.1 // indirect
 	github.com/inconshreveable/mousetrap v1.1.0 // indirect
-	github.com/jackc/pgpassfile v1.0.0 // indirect
-	github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761 // indirect
-	github.com/jackc/puddle/v2 v2.2.2 // indirect
-	github.com/kr/text v0.2.0 // indirect
 	github.com/lucasb-eyer/go-colorful v1.4.0 // indirect
 	github.com/mattn/go-isatty v0.0.24 // indirect
 	github.com/mattn/go-localereader v0.0.1 // indirect
@@ -55,7 +48,6 @@ require (
 	github.com/richardlehane/mscfb v1.0.7 // indirect
 	github.com/richardlehane/msoleps v1.0.6 // indirect
 	github.com/rivo/uniseg v0.4.7 // indirect
-	github.com/rogpeppe/go-internal v1.16.0 // indirect
 	github.com/spf13/pflag v1.0.9 // indirect
 	github.com/tiendc/go-deepcopy v1.7.2 // indirect
 	github.com/xuri/efp v0.0.1 // indirect

go.sum

@@ -1,5 +1,3 @@
-filippo.io/edwards25519 v1.2.0 h1:crnVqOiS4jqYleHd9vaKZ+HKtHfllngJIiOpNpoJsjo=
-filippo.io/edwards25519 v1.2.0/go.mod h1:xzAOLCNug/yB62zG1bQ8uziwrIqIuxhctzJT18Q77mc=
 github.com/JohannesKaufmann/dom v0.3.1 h1:J16l9JAHWgkFPR3VIPbQ1gvS0cWab6laK1q7PFL3qh0=
 github.com/JohannesKaufmann/dom v0.3.1/go.mod h1:BZPkf8ZeYrBgABjwJn9iiKt8aiCtkxpHkevms+Yp2DE=
 github.com/JohannesKaufmann/html-to-markdown/v2 v2.5.2 h1:XFJZFWESIWlUEHHjzBuv8RvrtCWnSGlimEX17ysSDb8=
@@ -37,8 +35,6 @@ github.com/charmbracelet/x/exp/golden v0.0.0-20240715153702-9ba8adf781c4/go.mod
 github.com/containerd/console v1.0.4-0.20230313162750-1ae8d489ac81 h1:q2hJAaP1k2wIvVRd/hEHD7lacgqrCPS+k8g1MndzfWY=
 github.com/containerd/console v1.0.4-0.20230313162750-1ae8d489ac81/go.mod h1:YynlIjWYF8myEu6sdkwKIvGQq+cOckRm6So2avqoYAk=
 github.com/cpuguy83/go-md2man/v2 v2.0.6/go.mod h1:oOW0eioCTA6cOiMLiUPZOpcVxMig6NIQQ7OS05n1F4g=
-github.com/creack/pty v1.1.9/go.mod h1:oKZEueFk5CKHvIhNR5MUki03XCEU+Q6VDXinZuGJ33E=
-github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/dlclark/regexp2 v1.11.0 h1:G/nrcoOa7ZXlpoa/91N3X7mM3r8eIlMBBJZvsz/mxKI=
@@ -51,8 +47,6 @@ github.com/eslider/go-hocr v0.2.2-0.20260827163626-8ff01582b002 h1:LOFxQG4mxvlH7
 github.com/eslider/go-hocr v0.2.2-0.20260827163626-8ff01582b002/go.mod h1:fIgfH/E1j3rU8du4X4+7mxTD0GPtPQibTzytgitdJWU=
 github.com/eslider/go-xls/v2 v2.1.0 h1:HszWKqYQbXxACmAXXWdMsfNl1NDBfGVBnJUPtyUHQ7A=
 github.com/eslider/go-xls/v2 v2.1.0/go.mod h1:xgxO6JrfuBr9jGUB+0z5l/yDmFFZ5diGk0ATGihxlMU=
-github.com/go-sql-driver/mysql v1.10.1 h1:arlSnNLq6a5yxGxV7qg9lF4j0C+KwD6NbQyKr9QL6ME=
-github.com/go-sql-driver/mysql v1.10.1/go.mod h1:M+cqaI7+xxXGG9swrdeUIoPG3Y3KCkF0pZej+SK+nWk=
 github.com/google/go-cmp v0.6.0 h1:ofyhxvXcZhMsU5ulbFiLKl/XBFqE1GSq7atu8tAmTRI=
 github.com/google/go-cmp v0.6.0/go.mod h1:17dUlkBOakJ0+DkrSSNjCkIjxS6bF9zb3elmeNGIjoY=
 github.com/google/go-querystring v1.2.0 h1:yhqkPbu2/OH+V9BfpCVPZkNmUXhb2gBxJArfhIxNtP0=
@@ -69,20 +63,8 @@ github.com/hexops/gotextdiff v1.0.3 h1:gitA9+qJrrTCsiCl7+kh75nPqQt1cx4ZkudSTLoUq
 github.com/hexops/gotextdiff v1.0.3/go.mod h1:pSWU5MAI3yDq+fZBTazCSJysOMbxWL1BSow5/V2vxeg=
 github.com/inconshreveable/mousetrap v1.1.0 h1:wN+x4NVGpMsO7ErUn/mUI3vEoE6Jt13X2s0bqwp9tc8=
 github.com/inconshreveable/mousetrap v1.1.0/go.mod h1:vpF70FUmC8bwa3OWnCshd2FqLfsEA9PFc4w1p2J65bw=
-github.com/jackc/pgpassfile v1.0.0 h1:/6Hmqy13Ss2zCq62VdNG8tM1wchn8zjSGOBJ6icpsIM=
-github.com/jackc/pgpassfile v1.0.0/go.mod h1:CEx0iS5ambNFdcRtxPj5JhEz+xB6uRky5eyVu/W2HEg=
-github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761 h1:iCEnooe7UlwOQYpKFhBabPMi4aNAfoODPEFNiAnClxo=
-github.com/jackc/pgservicefile v0.0.0-20240606120523-5a60cdf6a761/go.mod h1:5TJZWKEWniPve33vlWYSoGYefn3gLQRzjfDlhSJ9ZKM=
-github.com/jackc/pgx/v5 v5.11.0 h1:IzBBtyK9AHqf98cctWFifYSci2hgQR/cd56wB4p+ogg=
-github.com/jackc/pgx/v5 v5.11.0/go.mod h1:mal1tBGAFfLHvZzaYh77YS/eC6IX9OWbRV1QIIM0Jn4=
-github.com/jackc/puddle/v2 v2.2.2 h1:PR8nw+E/1w0GLuRFSmiioY6UooMp6KJv0/61nB7icHo=
-github.com/jackc/puddle/v2 v2.2.2/go.mod h1:vriiEXHvEE654aYKXXjOvZM39qJ0q+azkZFrfEOc3H4=
 github.com/joho/godotenv v1.5.1 h1:7eLL/+HRGLY0ldzfGMeQkb7vMd0as4CfYvUVzLqw0N0=
 github.com/joho/godotenv v1.5.1/go.mod h1:f4LDr5Voq0i2e/R5DDNOoa2zzDfwtkZa6DnEwAbqwq4=
-github.com/kr/pretty v0.3.0 h1:WgNl7dwNpEZ6jJ9k1snq4pZsg7DOEN8hP9Xw0Tsjwk0=
-github.com/kr/pretty v0.3.0/go.mod h1:640gp4NfQd8pI5XOwp5fnNeVWj67G7CFk/SaSQn7NBk=
-github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
-github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/lucasb-eyer/go-colorful v1.4.0 h1:UtrWVfLdarDgc44HcS7pYloGHJUjHV/4FwW4TvVgFr4=
 github.com/lucasb-eyer/go-colorful v1.4.0/go.mod h1:R4dSotOR9KMtayYi1e77YzuveK+i7ruzyGqttikkLy0=
 github.com/mattn/go-isatty v0.0.24 h1:tGZZoVgT/KiqK1c8ocVLeDS8BSWMRd47J3Lbz7vsReI=
@@ -116,8 +98,6 @@ github.com/rivo/uniseg v0.1.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJ
 github.com/rivo/uniseg v0.2.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
 github.com/rivo/uniseg v0.4.7 h1:WUdvkW8uEhrYfLC4ZzdpI2ztxP1I582+49Oc5Mq64VQ=
 github.com/rivo/uniseg v0.4.7/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
-github.com/rogpeppe/go-internal v1.16.0 h1:O9DK+vNMDVGLr2BeZqmpLeMjiMNkuXfcqntWbZV6S5g=
-github.com/rogpeppe/go-internal v1.16.0/go.mod h1:DrUVZyrJU+txYW5/1kwtXQSMFio52ZOxX7yM1VHvnxs=
 github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/sebdah/goldie/v2 v2.8.0 h1:dZb9wR8q5++oplmEiJT+U/5KyotVD+HNGCAc5gNr8rc=
 github.com/sebdah/goldie/v2 v2.8.0/go.mod h1:oZ9fp0+se1eapSRjfYbsV/0Hqhbuu3bJVvKI/NNtssI=
@@ -127,9 +107,6 @@ github.com/spf13/cobra v1.10.2 h1:DMTTonx5m65Ic0GOoRY2c16WCbHxOOw6xxezuLaBpcU=
 github.com/spf13/cobra v1.10.2/go.mod h1:7C1pvHqHw5A4vrJfjNwvOdzYu0Gml16OCs2GRiTUUS4=
 github.com/spf13/pflag v1.0.9 h1:9exaQaMOCwffKiiiYk6/BndUBv+iRViNW+4lEMi0PvY=
 github.com/spf13/pflag v1.0.9/go.mod h1:McXfInJRrz4CZXVZOBLb0bTZqETkiAhM9Iw0y3An2Bg=
-github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
-github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=
-github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/stretchr/testify v1.11.1 h1:7s2iGBzp5EwR7/aIZr8ao5+dra3wiQyKjjFuvgVKu7U=
 github.com/stretchr/testify v1.11.1/go.mod h1:wZwfW3scLgRK+23gO65QZefKpKQRnfz6sD981Nm4B6U=
 github.com/tiendc/go-deepcopy v1.7.2 h1:Ut2yYR7W9tWjTQitganoIue4UGxZwCcJy3orjrrIj44=
@@ -165,10 +142,8 @@ golang.org/x/text v0.38.0 h1:sXmwo9DwP3OK9EZ7PqAdaooSGozfl/3a6/xJcbzPRhE=
 golang.org/x/text v0.38.0/go.mod h1:YXZt3QhHUKYT53r2lLKFIVi6Ao1jdzrTR/KQ09qyxF4=
 golang.org/x/tools v0.47.0 h1:7Kn5x/d1svx/PzryTsqeoZN4TZwqeH5pGWjefhLi/1Q=
 golang.org/x/tools v0.47.0/go.mod h1:dFHnyTvFWY212G+h7ZY4Vsp/K3U4/7W9TyVaAul8uCA=
+gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
-gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
-gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c/go.mod h1:JHkPIbrfpd72SG/EVd6muEfDQjcINNoR0C8j2r3qZ4Q=
-gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
 gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
 gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
 modernc.org/cc/v4 v4.29.1 h1:MKgdCV3WykTSPqpVrnxdEDS0HEd2FHpKZDzxzU5LyeI=

internal/docpipe/docpipe.go

@@ -5,7 +5,6 @@
 //   - pandoc     — md↔docx
 //   - ocrmypdf   — OCR into a searchable PDF
 //   - pdftotext  — extract text layer
-//   - pdfdetach  — list/save embedded PDF attachments
 //   - tesseract  — OCR single images when ocrmypdf is unsuitable
 //   - ghostscript (gs) — PDF rewrite/optimize via PostScript (pdfwrite)
 package docpipe
@@ -27,7 +26,6 @@ type Tools struct {
 	Pandoc      string
 	OCRMyPDF    string
 	PDFToText   string
-	PDFDetach   string
 	Tesseract   string
 	Ghostscript string
 }
@@ -46,7 +44,6 @@ func LookPath() Tools {
 		Pandoc:      find("pandoc"),
 		OCRMyPDF:    find("ocrmypdf"),
 		PDFToText:   find("pdftotext"),
-		PDFDetach:   find("pdfdetach"),
 		Tesseract:   find("tesseract"),
 		Ghostscript: find("gs", "ghostscript"),
 	}

internal/docpipe/pdfattach.go

@@ -1,231 +0,0 @@
-package docpipe
-
-// Embedded PDF attachments (F6 #42). Digitised invoices often carry the
-// original scan as a PDF attachment; the searchable body may hold only a
-// summary. pdfdetach (poppler) lists/saves them; each saved attachment is run
-// through the normal docpipe extraction (pdftotext/OCR).
-
-import (
-	"bytes"
-	"encoding/xml"
-	"fmt"
-	"os"
-	"os/exec"
-	"path/filepath"
-	"strconv"
-	"strings"
-)
-
-// PDFAttachment is one embedded file in a PDF.
-type PDFAttachment struct {
-	Index int    // 1-based number as `pdfdetach -list` reports it
-	Name  string // embedded file name
-}
-
-// AttachmentText is the extracted text of one embedded attachment.
-type AttachmentText struct {
-	Name string
-	Text string
-}
-
-// parseAttachmentList parses `pdfdetach -list` output. The first line is a
-// count ("N embedded files"); every following line is "<index>: <name>".
-// Pure, so it is unit-tested.
-func parseAttachmentList(out string) []PDFAttachment {
-	var atts []PDFAttachment
-	for _, line := range strings.Split(out, "\n") {
-		line = strings.TrimSpace(line)
-		if line == "" {
-			continue
-		}
-		colon := strings.Index(line, ":")
-		if colon <= 0 {
-			continue
-		}
-		n, err := strconv.Atoi(strings.TrimSpace(line[:colon]))
-		if err != nil {
-			continue
-		}
-		name := strings.TrimSpace(line[colon+1:])
-		if name == "" {
-			continue
-		}
-		atts = append(atts, PDFAttachment{Index: n, Name: name})
-	}
-	return atts
-}
-
-// safeAttachmentName strips directories and leading dots so a hostile
-// attachment name cannot escape the extraction directory.
-func safeAttachmentName(name string) string {
-	name = strings.ReplaceAll(strings.TrimSpace(name), "\\", "/")
-	name = filepath.Base(name)
-	name = strings.TrimLeft(name, ".")
-	if name == "" || name == "." || name == "/" {
-		return ""
-	}
-	return name
-}
-
-// JoinWithAttachments appends attachment text to the document body, each
-// section preceded by an "[attachment: <name>]" marker so a search hit shows
-// its source. Empty attachments are skipped. Pure, so it is unit-tested.
-func JoinWithAttachments(body string, atts []AttachmentText) string {
-	var b strings.Builder
-	b.WriteString(strings.TrimRight(body, "\n"))
-	for _, a := range atts {
-		text := strings.TrimSpace(a.Text)
-		if text == "" {
-			continue
-		}
-		b.WriteString("\n\n[attachment: ")
-		b.WriteString(a.Name)
-		b.WriteString("]\n\n")
-		b.WriteString(text)
-	}
-	return b.String()
-}
-
-// ListAttachments returns the embedded files of a PDF. A PDF without
-// attachments yields an empty slice and no error.
-func (t Tools) ListAttachments(pdfPath string) ([]PDFAttachment, error) {
-	if t.PDFDetach == "" {
-		return nil, fmt.Errorf("pdfdetach not found on PATH")
-	}
-	cmd := exec.Command(t.PDFDetach, "-list", pdfPath)
-	var stderr bytes.Buffer
-	cmd.Stderr = &stderr
-	out, err := cmd.Output()
-	if err != nil {
-		return nil, fmt.Errorf("pdfdetach -list %s: %w (%s)", filepath.Base(pdfPath), err, strings.TrimSpace(stderr.String()))
-	}
-	return parseAttachmentList(string(out)), nil
-}
-
-// SaveAttachment writes the n-th embedded file (1-based) to outPath.
-func (t Tools) SaveAttachment(pdfPath string, index int, outPath string) error {
-	if t.PDFDetach == "" {
-		return fmt.Errorf("pdfdetach not found on PATH")
-	}
-	if strings.TrimSpace(outPath) == "" {
-		return fmt.Errorf("output path required")
-	}
-	if err := EnsureDir(outPath); err != nil {
-		return err
-	}
-	cmd := exec.Command(t.PDFDetach, "-save", strconv.Itoa(index), "-o", outPath, pdfPath)
-	var stderr bytes.Buffer
-	cmd.Stderr = &stderr
-	if err := cmd.Run(); err != nil {
-		return fmt.Errorf("pdfdetach -save %d: %w (%s)", index, err, strings.TrimSpace(stderr.String()))
-	}
-	return nil
-}
-
-// ToMarkdownWithAttachments extracts the file as ToMarkdown does, then — for
-// PDFs — appends the text of every embedded attachment under an
-// "[attachment: <name>]" marker. Attachment failures are non-fatal: the body
-// is returned unchanged.
-func (t Tools) ToMarkdownWithAttachments(path, workDir, lang string, minChars int) (string, error) {
-	res, err := t.ToMarkdown(path, workDir, lang, minChars)
-	if err != nil {
-		return "", err
-	}
-	if Ext(path) != ".pdf" {
-		return res.Markdown, nil
-	}
-	atts, err := t.attachmentTexts(path, workDir, lang, minChars)
-	if err != nil {
-		return res.Markdown, nil
-	}
-	return JoinWithAttachments(res.Markdown, atts), nil
-}
-
-// attachmentTexts saves and extracts every embedded attachment, skipping the
-// ones that cannot be read. It returns an error only when the attachment list
-// itself cannot be obtained.
-func (t Tools) attachmentTexts(pdfPath, workDir, lang string, minChars int) ([]AttachmentText, error) {
-	list, err := t.ListAttachments(pdfPath)
-	if err != nil || len(list) == 0 {
-		return nil, err
-	}
-	if workDir == "" {
-		workDir = os.TempDir()
-	}
-	dir := filepath.Join(workDir, "att-"+trimExt(filepath.Base(pdfPath)))
-	if err := os.MkdirAll(dir, 0o755); err != nil {
-		return nil, err
-	}
-	defer os.RemoveAll(dir)
-
-	out := make([]AttachmentText, 0, len(list))
-	for _, a := range list {
-		name := safeAttachmentName(a.Name)
-		if name == "" {
-			continue
-		}
-		saved := filepath.Join(dir, fmt.Sprintf("%d-%s", a.Index, name))
-		if err := t.SaveAttachment(pdfPath, a.Index, saved); err != nil {
-			continue
-		}
-		text, err := t.attachmentMarkdown(saved, dir, lang, minChars)
-		if err != nil {
-			continue
-		}
-		out = append(out, AttachmentText{Name: a.Name, Text: text})
-	}
-	return out, nil
-}
-
-// attachmentMarkdown extracts a saved attachment with the regular pipeline.
-// Structured attachments that docpipe does not convert (e-invoice XML,
-// CuraSoft JSON, CSV/HTML) fall back to their text content, so the embedded
-// original is still searchable. Other unreadable formats return an error and
-// the caller skips them.
-func (t Tools) attachmentMarkdown(path, workDir, lang string, minChars int) (string, error) {
-	if res, err := t.ToMarkdown(path, workDir, lang, minChars); err == nil {
-		return res.Markdown, nil
-	}
-	switch Ext(path) {
-	case ".xml", ".html", ".htm":
-		raw, err := os.ReadFile(path)
-		if err != nil {
-			return "", err
-		}
-		return xmlToText(raw), nil
-	case ".json", ".csv":
-		raw, err := os.ReadFile(path)
-		if err != nil {
-			return "", err
-		}
-		return string(raw), nil
-	default:
-		return "", fmt.Errorf("unsupported attachment type %q", Ext(path))
-	}
-}
-
-// xmlToText returns the character data of an XML/HTML document: element text
-// values with decoded entities, one per line. Used for invoice XML (EN 16931
-// CII / ZUGFeRD) and HTML attachments. Pure, so it is unit-tested.
-func xmlToText(raw []byte) string {
-	dec := xml.NewDecoder(bytes.NewReader(raw))
-	dec.Strict = false
-	var b strings.Builder
-	for {
-		tok, err := dec.Token()
-		if err != nil {
-			break
-		}
-		cd, ok := tok.(xml.CharData)
-		if !ok {
-			continue
-		}
-		s := strings.TrimSpace(string(cd))
-		if s == "" {
-			continue
-		}
-		b.WriteString(s)
-		b.WriteByte('\n')
-	}
-	return b.String()
-}

internal/docpipe/pdfattach_test.go

@@ -1,163 +0,0 @@
-package docpipe
-
-import (
-	"os"
-	"path/filepath"
-	"strings"
-	"testing"
-)
-
-func TestParseAttachmentList(t *testing.T) {
-	out := "2 embedded files\n1: original.pdf\n2: scan_001.png\n"
-	got := parseAttachmentList(out)
-	want := []PDFAttachment{{Index: 1, Name: "original.pdf"}, {Index: 2, Name: "scan_001.png"}}
-	if len(got) != len(want) {
-		t.Fatalf("got %+v, want %+v", got, want)
-	}
-	for i := range want {
-		if got[i] != want[i] {
-			t.Errorf("att[%d] = %+v, want %+v", i, got[i], want[i])
-		}
-	}
-}
-
-func TestParseAttachmentListEmptyAndMalformed(t *testing.T) {
-	for _, in := range []string{"", "0 embedded files\n", "garbage\n\n  \n"} {
-		if got := parseAttachmentList(in); len(got) != 0 {
-			t.Errorf("parseAttachmentList(%q) = %+v, want empty", in, got)
-		}
-	}
-}
-
-func TestSafeAttachmentName(t *testing.T) {
-	cases := map[string]string{
-		"note.txt":            "note.txt",
-		"../../evil.pdf":      "evil.pdf",
-		`..\..\evil.pdf`:      "evil.pdf",
-		"/abs/scan_001.pdf":   "scan_001.pdf",
-		".hidden":             "hidden",
-		"  spaced name.txt  ": "spaced name.txt",
-		"..":                  "",
-		"":                    "",
-	}
-	for in, want := range cases {
-		if got := safeAttachmentName(in); got != want {
-			t.Errorf("safeAttachmentName(%q) = %q, want %q", in, got, want)
-		}
-	}
-}
-
-func TestJoinWithAttachments(t *testing.T) {
-	body := "# scan.pdf\n\nbody token\n"
-	atts := []AttachmentText{
-		{Name: "original.pdf", Text: "  original token  "},
-		{Name: "empty.txt", Text: "   "},
-	}
-	got := JoinWithAttachments(body, atts)
-	if !strings.Contains(got, "body token") {
-		t.Errorf("body text lost: %q", got)
-	}
-	if !strings.Contains(got, "[attachment: original.pdf]") {
-		t.Errorf("marker missing: %q", got)
-	}
-	if !strings.Contains(got, "original token") {
-		t.Errorf("attachment text missing: %q", got)
-	}
-	if strings.Contains(got, "empty.txt") {
-		t.Errorf("empty attachment must be skipped: %q", got)
-	}
-}
-
-func TestJoinWithAttachmentsNoAttachments(t *testing.T) {
-	got := JoinWithAttachments("# a.pdf\n\ntext\n\n", nil)
-	if got != "# a.pdf\n\ntext" {
-		t.Errorf("got %q, want trimmed body only", got)
-	}
-}
-
-func TestListAttachmentsWithoutTool(t *testing.T) {
-	if _, err := (Tools{}).ListAttachments("x.pdf"); err == nil || !strings.Contains(err.Error(), "pdfdetach") {
-		t.Fatalf("want pdfdetach error, got %v", err)
-	}
-}
-
-// TestToMarkdownWithAttachmentsFixture exercises the real pdfdetach + pdftotext
-// pipeline on testdata/pdf-with-attachment.pdf (body token + embedded
-// goo-note.txt). Skips when poppler is not installed.
-func TestToMarkdownWithAttachmentsFixture(t *testing.T) {
-	tools := LookPath()
-	if tools.PDFDetach == "" || tools.PDFToText == "" {
-		t.Skip("pdfdetach/pdftotext not on PATH — skipping attachment extraction test")
-	}
-	fixture := filepath.Join("..", "..", "testdata", "pdf-with-attachment.pdf")
-	got, err := tools.ToMarkdownWithAttachments(fixture, t.TempDir(), "eng", 1)
-	if err != nil {
-		t.Fatalf("ToMarkdownWithAttachments: %v", err)
-	}
-	for _, want := range []string{"goobodytoken", "[attachment: goo-note.txt]", "gooattachmenttoken"} {
-		if !strings.Contains(got, want) {
-			t.Errorf("result missing %q:\n%s", want, got)
-		}
-	}
-}
-
-func TestXMLToText(t *testing.T) {
-	raw := []byte(`<?xml version="1.0" encoding="UTF-8"?>
-<rsm:CrossIndustryInvoice><rsm:ExchangedDocument>
-<ram:ID>S1063</ram:ID></rsm:ExchangedDocument>
-<ram:Name>Edelweiss &amp; Co</ram:Name><ram:GrandTotalAmount>42.00</ram:GrandTotalAmount>
-</rsm:CrossIndustryInvoice>`)
-	got := xmlToText(raw)
-	for _, want := range []string{"S1063", "Edelweiss & Co", "42.00"} {
-		if !strings.Contains(got, want) {
-			t.Errorf("xmlToText missing %q:\n%s", want, got)
-		}
-	}
-	if strings.ContainsAny(got, "<>") {
-		t.Errorf("xmlToText left markup: %q", got)
-	}
-}
-
-// TestAttachmentMarkdownFallback verifies structured attachments that docpipe
-// cannot convert are still reduced to searchable text, and unknown binary
-// formats error (so the caller skips them).
-func TestAttachmentMarkdownFallback(t *testing.T) {
-	dir := t.TempDir()
-	xmlPath := filepath.Join(dir, "factur-x.xml")
-	if err := os.WriteFile(xmlPath, []byte(`<Invoice><Number>S1063</Number></Invoice>`), 0o644); err != nil {
-		t.Fatal(err)
-	}
-	got, err := (Tools{}).attachmentMarkdown(xmlPath, dir, "", 0)
-	if err != nil {
-		t.Fatalf("attachmentMarkdown(xml): %v", err)
-	}
-	if !strings.Contains(got, "S1063") {
-		t.Errorf("xml attachment text = %q, want S1063", got)
-	}
-
-	binPath := filepath.Join(dir, "data.bin")
-	if err := os.WriteFile(binPath, []byte{0, 1, 2, 3}, 0o644); err != nil {
-		t.Fatal(err)
-	}
-	if _, err := (Tools{}).attachmentMarkdown(binPath, dir, "", 0); err == nil {
-		t.Error("unsupported attachment: want error, got nil")
-	}
-}
-
-// TestToMarkdownWithAttachmentsPlainPDF ensures a PDF without attachments
-// returns just the body (pdfdetach prints "0 embedded files").
-func TestToMarkdownWithAttachmentsPlainPDF(t *testing.T) {
-	tools := LookPath()
-	if tools.PDFDetach == "" || tools.PDFToText == "" {
-		t.Skip("pdfdetach/pdftotext not on PATH")
-	}
-	// The fixture itself is a PDF with one attachment; strip it by extracting
-	// the body only through ToMarkdown and compare JoinWithAttachments(nil).
-	res, err := tools.ToMarkdown(filepath.Join("..", "..", "testdata", "pdf-with-attachment.pdf"), t.TempDir(), "eng", 1)
-	if err != nil {
-		t.Fatalf("ToMarkdown: %v", err)
-	}
-	if strings.Contains(res.Markdown, "gooattachmenttoken") {
-		t.Fatalf("body must not contain attachment text: %q", res.Markdown)
-	}
-}

storage_fallback.go

@@ -157,13 +157,6 @@ func (c *Client) downloadFileEntry(ctx context.Context, f *FileEntry, dst io.Wri
 
 // downloadFromMinio streams objectKey from the configured MinIO bucket.
 func (c *Client) downloadFromMinio(ctx context.Context, objectKey string, dst io.Writer) (int64, error) {
-	return downloadMinioObject(ctx, c.client, objectKey, dst)
-}
-
-// downloadMinioObject streams objectKey from the configured MinIO bucket using
-// hc. It is the shared implementation behind (*Client).downloadFromMinio and
-// the read-only SQL store (file_pg.go), which has no *Client.
-func downloadMinioObject(ctx context.Context, hc *http.Client, objectKey string, dst io.Writer) (int64, error) {
 	if objectKey == "" {
 		return 0, fmt.Errorf("onlyoffice: minio fallback: empty object key")
 	}
@@ -184,7 +177,7 @@ func downloadMinioObject(ctx context.Context, hc *http.Client, objectKey string,
 	if err := signMinioRequest(ctx, cfg, req); err != nil {
 		return 0, err
 	}
-	resp, err := hc.Do(req)
+	resp, err := c.client.Do(req)
 	if err != nil {
 		return 0, fmt.Errorf("onlyoffice: minio fallback: %w", err)
 	}

testdata/pdf-with-attachment.pdf

@@ -1,114 +0,0 @@
-%PDF-1.7
-%Çì�¢
-%%Invocation: gs -q -dNOPAUSE -dBATCH -sDEVICE=pdfwrite ? ? ?
-5 0 obj
-<</Length 6 0 R/Filter /FlateDecode>>
-stream
-xœ-ŠA
-ƒ0÷ÿoÙnìO7q]èZ*ÿ�IˆmJ½½M
-³f7
-\¨6‘žfRÿŠ*ñºõª…8:g}‡f†Dºø”ÞiØsúØ	ÝôÝ;炱(Ùn.ly]ìUFz
-½~Aë øendstream
-endobj
-6 0 obj
-110
-endobj
-4 0 obj
-<</Type/Page/MediaBox [0 0 612 792]
-/Rotate 0/Parent 3 0 R
-/Resources<</ProcSet[/PDF /Text]
-/Font 8 0 R
->>
-/Contents 5 0 R
->>
-endobj
-3 0 obj
-<< /Type /Pages /Kids [
-4 0 R
-] /Count 1
->>
-endobj
-1 0 obj
-<</Type /Catalog /Pages 3 0 R
-/Metadata 9 0 R
->>
-endobj
-8 0 obj
-<</R7
-7 0 R>>
-endobj
-7 0 obj
-<</BaseFont/Helvetica/Type/Font
-/Subtype/Type1>>
-endobj
-9 0 obj
-<</Type/Metadata
-/Subtype/XML/Length 1173>>stream
-<?xpacket begin='' id='W5M0MpCehiHzreSzNTczkc9d'?>
-<?adobe-xap-filters esc="CRLF"?>
-<x:xmpmeta xmlns:x='adobe:ns:meta/' x:xmptk='XMP toolkit 2.9.1-13, framework 1.6'>
-<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#' xmlns:iX='http://ns.adobe.com/iX/1.0/'>
-<rdf:Description rdf:about="" xmlns:pdf='http://ns.adobe.com/pdf/1.3/' pdf:Producer='GPL Ghostscript 10.02.1'/>
-<rdf:Description rdf:about="" xmlns:xmp='http://ns.adobe.com/xap/1.0/'><xmp:ModifyDate>2026-09-16T17:29:12Z</xmp:ModifyDate>
-<xmp:CreateDate>2026-09-16T17:29:12Z</xmp:CreateDate>
-<xmp:CreatorTool>UnknownApplication</xmp:CreatorTool></rdf:Description>
-<rdf:Description rdf:about="" xmlns:xapMM='http://ns.adobe.com/xap/1.0/mm/' xapMM:DocumentID='uuid:ae8d7575-ea10-11fc-0000-3a98742e0b71'/>
-<rdf:Description rdf:about="" xmlns:dc='http://purl.org/dc/elements/1.1/' dc:format='application/pdf'><dc:title><rdf:Alt><rdf:li xml:lang='x-default'>Untitled</rdf:li></rdf:Alt></dc:title></rdf:Description>
-</rdf:RDF>
-</x:xmpmeta>
-                                                                        
-                                                                        
-<?xpacket end='w'?>
-endstream
-endobj
-2 0 obj
-<</Producer(GPL Ghostscript 10.02.1)
-/CreationDate(D:20260916172912Z00'00')
-/ModDate(D:20260916172912Z00'00')>>endobj
-xref
-0 10
-0000000000 65535 f 
-0000000476 00000 n 
-0000001882 00000 n 
-0000000417 00000 n 
-0000000276 00000 n 
-0000000077 00000 n 
-0000000257 00000 n 
-0000000569 00000 n 
-0000000540 00000 n 
-0000000633 00000 n 
-trailer
-<< /Size 10 /Root 1 0 R /Info 2 0 R
-/ID [<5A562476FBF11852133AA873452909F9><5A562476FBF11852133AA873452909F9>]
->>
-startxref
-2008
-%%EOF
-1 0 obj
-<</Type /Catalog /Pages 3 0 R /Metadata 9 0 R /Names <</EmbeddedFiles 12 0 R >> >> 
-endobj
-10 0 obj
-<</Length 44 /Params <</Size 44 >> >> stream
-gooattachmenttoken embedded attachment text
-
-endstream
-
-endobj
-11 0 obj
-<</Type /Filespec /UF <feff0067006f006f002d006e006f00740065002e007400780074> /EF <</F 10 0 R >> >> 
-endobj
-12 0 obj
-<</Names [<feff0067006f006f002d006e006f00740065002e007400780074> 11 0 R ] >> 
-endobj
-xref
-0 2
-0000000002 65535 f
-0000002361 00000 n
-10 3
-0000002463 00000 n
-0000002586 00000 n
-0000002705 00000 n
-trailer
-<</Size 13 /ID [(ZV$vûñR:¨sE\)	ù) (P…õrˆ‚ Åð¡Ï) ] /Root 1 0 R /Prev 2008 /Info 2 0 R >> 
-startxref
-2802