A command line utility that provides a lightweight setup for organizing and backing up files
SFS stores files from a variety of sources, aka collections, that may include directories and removable media, as symbolic links to the source files. It also stores the metadata of the source files so that files can later be queried without having to plug in the source media.
An SFS is a managed directory which is initialized with the command: sfs init
.
All commands to be executed in the context of an individual SFS must be run from within the SFS directory tree.
Files are added using the command sfs add-col my_collection path/to/source
(add collection).
SFS Files are symlinks to source files in added collections.
Foreign links and other files can also exist in an SFS but they are not managed by it and are mostly ignored.
-
Organizing Data Across Discs
SFS was built with the motivation of being able to have a combined view of data stored across multiple discs, organize the data in the view and reflect changes back to source discs. This is an effortless way of organizing content across discs which is otherwise painfully slow and limited as inter disc transfers are very slow and we can operate on a limited number of discs simultaneously. Since all operations in an SFS are performed within the same disc and on symlinks instead of heavy files, they are much faster
Note: To view the content of a file we obviously do need the source to be available. So, if there is a need of viewing file content while organizing them, the source needs to be plugged in which might or might not be appropriate for all use cases. However, SFS makes it easy to query the source of an SFS File when it is needed to be accessed
-
Backing up Files
While there are lots of ways to make direct backups of directories, an SFS allows you to organize the content while backing them up and potentially saving them to multiple destinations with a single command. For example, you might have an SFS in which you add local files, like multimedia and documents, organize them in hierarchies resembling your storage hierarchies, then map the top-level SFS directories to backup destinations and perform the backup with a single save command. Periodically, you will have to synchronize the SFS, sort the newly added local files and rerun the backup
-
Decouple Data Storage and View
Data often needs to be stored in a certain way which might not be similar to the hierarchy in which you want to view it. SFS allows you to create a virtual hierarchy for viewing content. For example, consider that you have data saved in a number of discs or directories, organized as music, documents, projects, etc. Your options are either to keep a copy of the important files locally, which we commonly do, or to plugin all the media one by one and search for the files you need, which hopefully no one does. You can instead create an SFS instance, dump all your discs into it, create a directory in the SFS for local files and copy all needed files to your local system. You can search for files in all your discs locally and, periodically, you can update what files to be kept in you local system
-
SFS Operations
init Initialize a new SFS in the current directory is-sfs [path] Check whether a path is inside an SFS
-
Collection Operations
add-col path/to/source Add a named collection to the current SFS -n, --name Collection name (defaults to source root directory name) is-col Check whether a path is inside any collection added to the current SFS list-cols List all collections in the current SFS del-col Delete a collection and associated symllinks from the current SFS sync-col Synchronize any changes made to a collection (addition, modification and deletion of files)
-
Querying SFS files
query [path] Query metadata of a file or directory in an SFS
-
Deduplication
find-dups target Check for duplicate files (by name and size) recursively in a target directory and save dulicates to a JSON file in the target directory -o, --override Override the generated JSON file if it exists -d, --del-duplicates Mark duplicates (all but first in a list of duplicate files) for deletion dedup target Use the JSON file (after manually choosing which files to keep) to delete duolicates in a target directory -d, --del-json Delete the generated JSON file after a successful de-deuplication
-
Merge
merge target source Merge two non-nested directories in an SFS. In case of merge conlicts, the process terminates after saving conflicting files to a JSON. The file can be edited and used for completing the merge operation -k, --on-conflict Conflict resoution can be one of keep-target, keep-source or keep-both -c, --continue Use specified or default conflict rsolution without saving conflicts JSON -j, --json Use the generated JSON file for handling conflicts -o, --override Override the generated JSON file if it exists -d, --del-json Delete the generated JSON file after a successful merge -s, --del-source Delete the source directory after a successful merge
Install with pip
pip install symbolic-file-system
Or clone this repo and run setup directly
python3 setup.py install
Access all SFS commands through the installed script named sfs
mkdir my-first-sfs
cd my-first-sfs
sfs init
sfs add-col my-hdd /media/hdd
You can run tests with nose
nosetests
-
Saving Changes Made Back to Source
Any changes made to the organization of links in an SFS, like deletion, renaming or relocation will be reflected back to the source discs or directories. There will be a number of modes of saving changes:
- Copy: Files will be copied to an actual directory or drive with the same file hierarchy as some directory in an SFS, the copied files being actual source files from various collections
- Move: The source files will be moved to a new destination as specified by the SFS file hierarchy and save mapping
- Delete: Files deleted in an SFS will be reflected back to a collection source or a part of it
- Save: In this mode, an exhaustive mapping of SFS directories and collection sources will be specified and changes will be reflected in all collections, internally executing Move and Delete on all of them
-
Filtering Files
SFS will add the ability to filter SFS Files and directories, a feature missing from most file systems. The following filters will be available:
- Filter by file size
- Filter by file type
- Filter by any custom properties
-
Adding Properties to Files and Directories
It will be possible to add properties to files and directories in an SFS an look them up which can be useful for simply tagging them and can even be used while applying filters
-
Freezing Directories
Freeze will be a special property that can be applied to directories in an SFS to prevent them from being manipulated by SFS commands like Merge, Filters and De-duplication. This can be useful for hierarchies like project and application directories which must remain intact
- Though SFS is all about symlinks and your source files are always safe, it is recommended to backup the SFS root directory before doing anything adventurous. Backing up is as simple is making a copying of the SFS root directory