Artifact [a9c2805cd2]

Artifact a9c2805cd2479253dd01b86130dfef92f44d642da9b314c5a21d1922322fe238:

• File internal/blob/doc.go — part of check-in [0fc4d176c3] at 2026-10-28 22:52:43 on branch feat/chunked-blobs — Record chunked length in blob.content and verify parts on read

  Review of the chunked-blob change found a missing last part read back short with no error, and a chunked write outside a transaction could leave a torn row that looked held.

  A chunked row now keeps its stored length in blob.content as an INTEGER. Reads check the parts run unbroken from seq 0, are full-sized but for the last, and add up to that length; anything else is ErrPartsCorrupt, and a NULL content is no longer mistaken for a chunked one. StoredLenSQL reads the length off the row and needs no schema probe.

  Chunked writes, and the phantom fill around them, run in a savepoint of the caller's transaction or a transaction of their own, so a failed part leaves nothing behind. A value needing more than 4096 parts is an error, not a panic, since stored values can come from peers. The cone and sparse local payload reads use ReadStoredExact, which loads a value only at the length already budgeted for.

  MaxValueBytes is documented as process-wide, and unversioned content as not chunked. (user: [email protected], size: 1535) [annotate] [blame] [check-ins using]

// Package blob handles content-addressed blob storage in Fossil
// repository databases.
//
// Fossil's blob format is a 4-byte big-endian uncompressed-size prefix
// followed by zlib-compressed data. [Compress] and [Decompress] handle
// this encoding transparently.
//
// [Store] compresses content, computes its SHA1 hash, and inserts it
// into the blob table. [Load] retrieves and decompresses a blob by RID.
// [StoreDeltaRaw] stores content that arrived over the wire already
// delta-encoded, without expanding it. [EncodeForStorage] is the shared
// decision point for whether a receive path can skip [Compress] entirely
// and write already-encoded wire bytes straight to blob.content.
//
// A row's stored bytes are read and written only through [ReadStored],
// [ReadStoredExact], [WriteStored], [InsertStored], [StoredHead] and
// [DropStored] (and the scan expressions [StoredLenSQL] and
// [StoredPrefixSQL]), because a value over db.DriverConfig.MaxValueBytes
// lives in blobpart rows, with only its length in blob.content.
//
// This package does not decide what gets delta-encoded. Deltifying an
// artifact that is already stored means rewriting an existing row, not
// inserting a new one, and the decision needs to expand content, which
// this package cannot do without importing internal/content. Both live
// in content.Deltify, which is the only place the policy is stated.
// [StoreDelta] inserts a new row as a delta and has no callers outside
// tests; it is not the commit path's primitive.
package blob