* feat(s3): Add support for S3 object storage classes
Introduces a new 'storage_class' configuration option for S3 providers. Users can now specify the desired storage class (e.g., Standard, GLACIER, DEEP_ARCHIVE) for objects uploaded to S3-compatible services like AWS S3 and Tencent COS.
The input storage class string is normalized to match AWS SDK constants, supporting various common aliases. If an unknown storage class is provided, it will be used as a raw value with a warning. This enhancement provides greater control over storage costs and data access patterns.
* feat(storage): Support for displaying file storage classes
Adds storage class information to file metadata and API responses.
This change introduces the ability to store file storage classes in file metadata and display them in API responses. This allows users to view a file's storage tier (e.g., S3 Standard, Glacier), enhancing data management capabilities.
Implementation details include:
- Introducing the StorageClassProvider interface and the ObjWrapStorageClass structure to uniformly handle and communicate object storage class information.
- Updated file metadata structures (e.g., ArchiveObj, FileInfo, RespFile) to include a StorageClass field.
- Modified relevant API response functions (e.g., GetFileInfo, GetFileList) to populate and return storage classes.
- Integrated functionality for retrieving object storage classes from underlying storage systems (e.g., S3) and wrapping them in lists.
* feat(driver/s3): Added the "Other" interface and implemented it by the S3 driver.
A new `driver.Other` interface has been added and defined in the `other.go` file.
The S3 driver has been updated to implement this new interface, extending its functionality.
* feat(s3): Add S3 object archive and thaw task management
This commit introduces comprehensive support for S3 object archive and thaw operations, managed asynchronously through a new task system.
- **S3 Transition Task System**:
- Adds a new `S3Transition` task configuration, including workers, max retries, and persistence options.
- Initializes `S3TransitionTaskManager` to handle asynchronous S3 archive/thaw requests.
- Registers dedicated API routes for monitoring S3 transition tasks.
- **Integrate S3 Archive/Thaw with Other API**:
- Modifies the `Other` API handler to intercept `archive` and `thaw` methods for S3 storage drivers.
- Dispatches these operations as `S3TransitionTask` instances to the task manager for background processing.
- Returns a task ID to the client for tracking the status of the dispatched operation.
- **Refactor `other` package for improved API consistency**:
- Exports previously internal structs such as `archiveRequest`, `thawRequest`, `objectDescriptor`, `archiveResponse`, `thawResponse`, and `restoreStatus` by making their names public.
- Makes helper functions like `decodeOtherArgs`, `normalizeStorageClass`, and `normalizeRestoreTier` public.
- Introduces new constants for various S3 `Other` API methods.
* 标签管理
* pr检查优化
* feat(role): Implement role management functionality
- Add role management routes in `server/router.go` for listing, getting, creating, updating, and deleting roles
- Introduce `initRoles()` in `internal/bootstrap/data/data.go` for initializing roles during bootstrap
- Create `internal/op/role.go` to handle role operations including caching and singleflight
- Implement role handler functions in `server/handles/role.go` for API responses
- Define database operations for roles in `internal/db/role.go`
- Extend `internal/db/db.go` for role model auto-migration
- Design `internal/model/role.go` to represent role structure with ID, name, description, base path, and permissions
- Initialize default roles (`admin` and `guest`) in `internal/bootstrap/data/role.go` during startup
* refactor(user roles): Support multiple roles for users
- Change the `Role` field type from `int` to `[]int` in `drivers/alist_v3/types.go` and `drivers/quqi/types.go`.
- Update the `Role` field in `internal/model/user.go` to use a new `Roles` type with JSON and database support.
- Modify `IsGuest` and `IsAdmin` methods to check for roles using `Contains` method.
- Update `GetUserByRole` method in `internal/db/user.go` to handle multiple roles.
- Add `roles.go` to define a new `Roles` type with JSON marshalling and scanning capabilities.
- Adjust code in `server/handles/user.go` to compare roles with `utils.SliceEqual`.
- Change role initialization for users in `internal/bootstrap/data/dev.go` and `internal/bootstrap/data/user.go`.
- Update `Role` handling in `server/handles/task.go`, `server/handles/ssologin.go`, and `server/handles/ldap_login.go`.
* feat(user/role): Add path limit check for user and role permissions
- Add new permission bit for checking path limits in `user.go`
- Implement `CheckPathLimit` method in `User` struct to validate path access
- Modify `JoinPath` method in `User` to enforce path limit checks
- Update `role.go` to include path limit logic in `Role` struct
- Document new permission bit in `Role` and `User` comments for clarity
* feat(permission): Add role-based permission handling
- Introduce `role_perm.go` for managing user permissions based on roles.
- Implement `HasPermission` and `MergeRolePermissions` functions.
- Update `webdav.go` to utilize role-based permissions instead of direct user checks.
- Modify `fsup.go` to integrate `CanAccessWithRoles` function.
- Refactor `fsread.go` to use `common.HasPermission` for permission validation.
- Adjust `fsmanage.go` for role-based access control checks.
- Enhance `ftp.go` and `sftp.go` to manage FTP access via roles.
- Update `fsbatch.go` to employ `MergeRolePermissions` for batch operations.
- Replace direct user permission checks with role-based permission handling across various modules.
* refactor(user): Replace integer role values with role IDs
- Change `GetAdmin()` and `GetGuest()` functions to retrieve role by name and use role ID.
- Add patch for version `v3.45.2` to convert legacy integer roles to role IDs.
- Update `dev.go` and `user.go` to use role IDs instead of integer values for roles.
- Remove redundant code in `role.go` related to guest role creation.
- Modify `ssologin.go` and `ldap_login.go` to set user roles to nil instead of using integer roles.
- Introduce `convert_roles.go` to handle conversion of legacy roles and ensure role existence in the database.
* feat(role_perm): implement support for multiple base paths for roles
- Modify role permission checks to support multiple base paths
- Update role creation and update functions to handle multiple base paths
- Add migration script to convert old base_path to base_paths
- Define new Paths type for handling multiple paths in the model
- Adjust role model to replace BasePath with BasePaths
- Update existing patches to handle roles with multiple base paths
- Update bootstrap data to reflect the new base_paths field
* feat(role): Restrict modifications to default roles (admin and guest)
- Add validation to prevent changes to "admin" and "guest" roles in `UpdateRole` and `DeleteRole` functions.
- Introduce `ErrChangeDefaultRole` error in `internal/errs/role.go` to standardize error messaging.
- Update role-related API handlers in `server/handles/role.go` to enforce the new restriction.
- Enhance comments in `internal/bootstrap/data/role.go` to clarify the significance of default roles.
- Ensure consistent error responses for unauthorized role modifications across the application.
* 🔄 **refactor(role): Enhance role permission handling**
- Replaced `BasePaths` with `PermissionPaths` in `Role` struct for better permission granularity.
- Introduced JSON serialization for `PermissionPaths` using `RawPermission` field in `Role` struct.
- Implemented `BeforeSave` and `AfterFind` GORM hooks for handling `PermissionPaths` serialization.
- Refactored permission calculation logic in `role_perm.go` to work with `PermissionPaths`.
- Updated role creation logic to initialize `PermissionPaths` for `admin` and `guest` roles.
- Removed deprecated `CheckPathLimit` method from `Role` struct.
* fix(model/user/role): update permission settings for admin and role
- Change `RawPermission` field in `role.go` to hide JSON representation
- Update `Permission` field in `user.go` to `0xFFFF` for full access
- Modify `PermissionScopes` in `role.go` to `0xFFFF` for enhanced permissions
* 🔒 feat(role-permissions): Enhance role-based access control
- Introduce `canReadPathByRole` function in `role_perm.go` to verify path access based on user roles
- Modify `CanAccessWithRoles` to include role-based path read check
- Add `RoleNames` and `Permissions` to `UserResp` struct in `auth.go` for enhanced user role and permission details
- Implement role details aggregation in `auth.go` to populate `RoleNames` and `Permissions`
- Update `User` struct in `user.go` to include `RolesDetail` for more detailed role information
- Enhance middleware in `auth.go` to load and verify detailed role information for users
- Move `guest` user initialization logic in `user.go` to improve code organization and avoid repetition
* 🔒 fix(permissions): Add permission checks for archive operations
- Add `MergeRolePermissions` and `HasPermission` checks to validate user access for reading archives
- Ensure users have `PermReadArchives` before proceeding with `GetNearestMeta` in specific archive paths
- Implement permission checks for decompress operations, requiring `PermDecompress` for source paths
- Return `PermissionDenied` errors with 403 status if user lacks necessary permissions
* 🔒 fix(server): Add permission check for offline download
- Add permission merging logic for user roles
- Check user has permission for offline download addition
- Return error response with "permission denied" if check fails
* ✨ feat(role-permission): Implement path-based role permission checks
- Add `CheckPathLimitWithRoles` function to validate access based on `PermPathLimit` permission.
- Integrate `CheckPathLimitWithRoles` in `offline_download` to enforce path-based access control.
- Apply `CheckPathLimitWithRoles` across file system management operations (e.g., creation, movement, deletion).
- Ensure `CheckPathLimitWithRoles` is invoked for batch operations and archive-related actions.
- Update error handling to return `PermissionDenied` if the path validation fails.
- Import `errs` package in `offline_download` for consistent error responses.
* ✨ feat(role-permission): Implement path-based role permission checks
- Add `CheckPathLimitWithRoles` function to validate access based on `PermPathLimit` permission.
- Integrate `CheckPathLimitWithRoles` in `offline_download` to enforce path-based access control.
- Apply `CheckPathLimitWithRoles` across file system management operations (e.g., creation, movement, deletion).
- Ensure `CheckPathLimitWithRoles` is invoked for batch operations and archive-related actions.
- Update error handling to return `PermissionDenied` if the path validation fails.
- Import `errs` package in `offline_download` for consistent error responses.
* ♻️ refactor(access-control): Update access control logic to use role-based checks
- Remove deprecated logic from `CanAccess` function in `check.go`, replacing it with `CanAccessWithRoles` for improved role-based access control.
- Modify calls in `search.go` to use `CanAccessWithRoles` for more precise handling of permissions.
- Update `fsread.go` to utilize `CanAccessWithRoles`, ensuring accurate access validation based on user roles.
- Simplify import statements in `check.go` by removing unused packages to clean up the codebase.
* ✨ feat(fs): Improve visibility logic for hidden files
- Import `server/common` package to handle permissions more robustly
- Update `whetherHide` function to use `MergeRolePermissions` for user-specific path permissions
- Replace direct user checks with `HasPermission` for `PermSeeHides`
- Enhance logic to ensure `nil` user cases are handled explicitly
* 标签管理
* feat(db/auth/user): Enhance role handling and clean permission paths
- Comment out role modification checks in `server/handles/user.go` to allow flexible role changes.
- Improve permission path handling in `server/handles/auth.go` by normalizing and deduplicating paths.
- Introduce `addedPaths` map in `CurrentUser` to prevent duplicate permissions.
* feat(storage/db): Implement role permissions path prefix update
- Add `UpdateRolePermissionsPathPrefix` function in `role.go` to update role permissions paths.
- Modify `storage.go` to call the new function when the mount path is renamed.
- Introduce path cleaning and prefix matching logic for accurate path updates.
- Ensure roles are updated only if their permission scopes are modified.
- Handle potential errors with informative messages during database operations.
* feat(role-migration): Implement role conversion and introduce NEWGENERAL role
- Add `NEWGENERAL` to the roles enumeration in `user.go`
- Create new file `convert_role.go` for migrating legacy roles to new model
- Implement `ConvertLegacyRoles` function to handle role conversion with permission scopes
- Add `convert_role.go` patch to `all.go` under version `v3.46.0`
* feat(role/auth): Add role retrieval by user ID and update path prefixes
- Add `GetRolesByUserID` function for efficient role retrieval by user ID
- Implement `UpdateUserBasePathPrefix` to update user base paths
- Modify `UpdateRolePermissionsPathPrefix` to return modified role IDs
- Update `auth.go` middleware to use the new role retrieval function
- Refresh role and user caches upon path prefix updates to maintain consistency
---------
Co-authored-by: Leslie-Xy <540049476@qq.com>
* feat: support general users view and cancel own tasks
Add a creator attribute to the upload, copy and offline download
tasks, so that a GENERAL task creator can view and cancel them.
BREAKING CHANGE:
1. A new internal package `task` including the struct `TaskWithCreator`
which embeds `tache.Base` is created, and the past dependence on
`tache.Task` will all be transferred to dependence on this package.
2. The API `/admin/task` can now also be accessed via `/task`, and the
old endpoint is retained to ensure compatibility with legacy
automation scripts.
Closes#7398
* fix(deps): update github.com/xhofe/tache to v0.1.3
general: add createTime/updateTime support in webdav and some drivers
general: add hash support in some drivers
general: cross-storage rapid-upload support
general: enhance upload to avoid local temp file if possible
general: replace readseekcloser with File interface to speed upstream operations
feat(aliyun_open): same as above
feat(crypt): add hack for 139cloud
Close#4934Close#4819
baidu_netdisk needs to improve the upload code to support rapid-upload
this PR has several enhancements, fixes, and features:
- [x] Crypt: a transparent encryption driver. Anyone can easily, and safely store encrypted data on the remote storage provider. Consider your data is safely stored in the safe, and the storage provider can only see the safe, but not your data.
- [x] Optional: compatible with [Rclone Crypt](https://rclone.org/crypt/). More ways to manipulate the encrypted data.
- [x] directory and filename encryption
- [x] server-side encryption mode (server encrypts & decrypts all data, all data flows thru the server)
- [x] obfuscate sensitive information internally
- [x] introduced a server memory-cached multi-thread downloader.
- [x] Driver: **Quark** enabled this feature, faster load in any single thread scenario. e.g. media player directly playing from the link, now it's faster.
- [x] general improvement on HTTP/WebDAV stream processing & header handling & response handling
- [x] Driver: **Mega** driver support ranged http header
- [x] Driver: **Quark** fix bug of not closing HTTP request to Quark server while user end has closed connection to alist
## Crypt, a transparent Encrypt/Decrypt Driver. (Rclone Crypt compatible)
e.g.
Crypt mount path -> /vault
Crypt remote path -> /ali/encrypted
Aliyun mount paht -> /ali
when the user uploads a.jpg to /vault, the data will be encrypted and saved to /ali/encrypted/xxxxx. And when the user wants to access a.jpg, it's automatically decrypted, and the user can do anything with it.
Since it's Rclone Crypt compatible, users can download /ali/encrypted/xxxxx and decrypt it with rclone crypt tool. Or the user can mount this folder using rclone, then mount the decrypted folder in Linux...
NB. Some breaking changes is made to make it follow global standard, e.g. processing the HTTP header properly.
close#4679close#4827
Co-authored-by: Sean He <866155+seanhe26@users.noreply.github.com>
Co-authored-by: Andy Hsu <i@nn.ci>
* refactor:separate the setting method from the db package to the op package and add the cache
* refactor:separate the meta method from the db package to the op package
* fix:setting not load database data
* refactor:separate the user method from the db package to the op package
* refactor:remove user JoinPath error
* fix:op package user cache
* refactor:fs package list method
* fix:tile virtual paths (close#2743)
* Revert "refactor:remove user JoinPath error"
This reverts commit 4e20daaf9e.
* clean path directly may lead to unknown behavior
* fix: The path of the meta passed in must be prefix of reqPath
* chore: rename all virtualPath to mountPath
* fix: `getStoragesByPath` and `GetStorageVirtualFilesByPath`
is_sub_path:
/a/b isn't subpath of /a/bc
* fix: don't save setting if hook error
Co-authored-by: Noah Hsu <i@nn.ci>
* refactor:Prepare to remove the get interface
* feat:add obj Unwarp interface
* refactor:obj name mapping and program internal path processing
* chore: fix typo
* feat: unwrap get
* fix: no use op.Get to get parent id
* fix: set the path uniformly
Co-authored-by: Noah Hsu <i@nn.ci>
* feat: ignore AList storage on indexing
* fix: remove unused err in `walkFn`
* chore(ci): fix auto_lang trigger and run it
* feat: batch index
* feat: quit index & init index
* feat: set DocType for bleve data
* fix: build index cleanup check origin err
