feat(dedup): CDC sub-file deduplication with FastCDC + parallel chunk storage + dedup skip

- Replace whole-file SHA-256 dedup with FastCDC 2020 content-defined chunking
  (min 64KB, avg 256KB, max 1MB) + BLAKE3 hashing
- Add chunk_manifests table (file_hash → chunk_hashes[] + chunk_sizes[])
- Add put_blob_from_bytes to BlobStorageBackend trait (all 7 backends)
- 3-phase store_chunks pipeline:
  Phase 0: batch-check existing chunks (single PG query)
  Phase 1: selective disk read (skip existing chunks entirely)
  Phase 2: parallel upload with buffer_unordered(8)
- CDC-aware read_blob_stream and read_blob_range_stream with legacy fallback
- Transactional manifest + chunk ref-count cascade on remove_reference
- 12 CDC tests (determinism, reassembly, contiguity, sub-file dedup, etc.)
- Update deduplication.md to reflect new architecture

migrations/20260414000000_chunk_manifests.sql

@@ -0,0 +1,24 @@
+-- Content-Defined Chunking (CDC) manifests for sub-file deduplication.
+--
+-- Each file uploaded via CDC is split into variable-size chunks (FastCDC).
+-- The manifest records the ordered list of chunk hashes that compose the file.
+-- Individual chunks are stored in storage.blobs (shared across manifests).
+--
+-- Legacy whole-file blobs (pre-CDC) remain in storage.blobs and are
+-- accessed directly when no matching manifest row exists.
+
+CREATE TABLE IF NOT EXISTS storage.chunk_manifests (
+    file_hash    VARCHAR(64) PRIMARY KEY,
+    chunk_hashes TEXT[]      NOT NULL,
+    chunk_sizes  BIGINT[]    NOT NULL,
+    total_size   BIGINT      NOT NULL,
+    chunk_count  INTEGER     NOT NULL,
+    content_type TEXT,
+    ref_count    INTEGER     NOT NULL DEFAULT 1 CHECK (ref_count >= 0),
+    created_at   TIMESTAMPTZ NOT NULL DEFAULT NOW()
+);
+
+-- Index for GC: find manifests with no references.
+CREATE INDEX IF NOT EXISTS idx_chunk_manifests_ref_count_zero
+    ON storage.chunk_manifests (file_hash)
+    WHERE ref_count = 0;