Reliable file and object workflows beyond ordinary File / Directory APIs.
It provides transactional multi-file writes, content-addressed storage, content-defined chunking, and deterministic directory comparison.
Creates or replaces multiple files as one logical operation:
- staging directory
- write-ahead journal
- per-file checksums
- atomic replace where the OS supports it (documented fallback otherwise)
- rollback and crash recovery with deterministic recovery reports
- cancellation without undocumented half-state
Streams blobs into a hash-addressed layout (SHA-256):
- deduplication by content hash
- integrity verification
- safe concurrent writers (temp + rename)
- orphan detection
- garbage-collection planning without automatic destructive deletion
Rabin-style rolling-hash chunking with configurable min / target / max sizes. Streams input with bounded memory so similar content tends to share chunk boundaries.
Builds deterministic directory snapshots and diffs:
- added / removed / modified / moved detection
- content hashing and metadata policies
- symlink policy and cycle protection
- cross-platform path normalization
- Installers, exporters, or config deployers that must not leave partial file sets
- Local caches or artifact stores with dedup
- Backup/dedup pipelines needing stable chunk boundaries
- Sync tools that need a deterministic tree diff
- Thin wrappers around
File.WriteAllTextorFileStream - A full distributed object store or cloud SDK
dotnet add package FunctionFoundry.StorageNone beyond the .NET shared framework.