Tianyi Cui
bf29247fad
V3 permits external event, source and content identities and additional request-tool fields. Copying those values into V4 can give an old plugin named tool a core source identity, turn old tool-addition content into a registry instruction, or activate formerly ignored deferLoading and delivery metadata. Rejecting such collisions would instead refuse valid source logs. Keep historical extensions in the existing extensible JSON representation. Prefix unknown content and ignorable event tags with plugin:, preserving all other content fields and opaque event payloads. Prefix external source identities with plugin:plugin: or plugin:source: so wrapper and direct kinds cannot collide even when original names already contain a category prefix. Known first-party mappings remain explicit and unchanged. Prefix extra request-tool field names and block-start field names with plugin:, retaining exact values and existing-prefix doubling. Core schema parameters, arguments, replay state and nested plugin data are not traversed. Keep future-generation delivery records inactive after header promotion. Pin the V3 event vocabulary independently of the current writer and provide an offline verifier against an explicitly selected historical source commit. Attachment authorization and export inspect declared event content only. Team checkpoints preserve all own keys of unknown JSON blocks while still rejecting malformed known content and retired core tool-result blocks. Exercise physical read/write/reopen, prefix injectivity, own __proto__ keys, provider opacity and retained source-identity snapshots. No generic extension block, metadata wrapper or new schema-tooling type is introduced. V4 type declarations, persistence schema, checkpoint and frozen V0-V3 codecs remain unchanged. Canonical tool-result lifting and incomplete child-catalog policy remain separate concerns. Adapted from the extension-preservation work in 7ba7e174dd1 by Tianyi Cui. Co-authored-by: Tianyi Cui <[email protected]>
description, kind
| description | kind |
|---|---|
| Build-static first-party Session format codec and adjacent migration assembly for persistence readers. | package-library |
@deepseek-ai/dsh-session-format-catalog
English | 中文
Summary
dsh-session-format-catalog gives persistence one deterministic Session format reader without consulting mounted plugins. It assembles codecs and adjacent edges from the earliest supported format through the current writer format, checks the complete gap-free chain at module initialization, and exposes physical dispatch, header-only classification, single-pass row restoration, and current record encoding through sessionFormatCatalog.
Table of Contents
- Use this package
- Understand the implementation
- Further Exploration
- Model Experience
- Known Limitations and Deferred Work
- Dev Note
Use this package
When to use it
Import this library from persistence and test-support readers that need the complete first-party released-format inventory before any feature plugin mounts. Feature compositions do not register or reorder its entries. No runtime invariant companion is published because construction rejects an invalid static inventory and each completed restore validates its result; mutable row-decoder state belongs to one caller-owned streaming restore.
Entry point
const descriptor = sessionFormatCatalog.readHeader(physicalHeader)
const catalog = createSessionFormatCatalogWithChildren(childFacts)
const restore = catalog.createRestore(physicalHeader, { recovery: 'recoverable', validation: 'transformed' })
for (const row of physicalRows) restore.decodeRow(row)
const current = restore.finish()
const headerRecord = sessionFormatCatalog.encodeCurrentHeader(current.header, current.inheritedEventCount)
const eventRecords = current.events.map(sessionFormatCatalog.encodeCurrentEvent)
Import sessionFormatCatalog from the package root. JSONL and fixture readers create one restore, push each parsed physical row through decodeRow(), and call finish() once. Writers serialize the returned current artifact through encodeCurrentHeader() and encodeCurrentEvent(). Listing calls readHeader() and never opens event bodies.
Production historical reads select { recovery: 'recoverable', validation: 'transformed' }. Worker and fixture verification select { recovery: 'strict', validation: 'current' }. Transformed validation runs the released-current rules after migration but deliberately skips installed semantic validation for input that is already current.
The catalog contains all supported historical readers directly. A profile cannot add, remove, or reorder an edge by mounting a feature plugin. Its peer dependency on dsh-session supplies the installed current event vocabulary and current restoration rules, while historical edge validators remain frozen. The browser-safe ./message-projections export assembles current plugin-owned interpreters for detached constructors and surface folds; it does not mount recovery listeners.
createSessionFormatCatalogWithChildren(childFacts) binds explicit child evidence to V3→V4 during assembly; see the catalog-completion specification. historicalSessionFormatCatalog restores V0–V3 using the fixed released-V3 event vocabulary to collect child prerequisites without recursively completing their catalogs. An ignorable V3 extension remains opaque even when the installed writer knows its name. Isolated transcript replay explicitly supplies an empty array; persistence must collect the complete available direct-child set. Keep the supplied evidence unchanged for the catalog’s lifetime. Each restore owns independent stage state. The static sessionFormatCatalog supports header and native current-format reads; historical body reads require the child-bound catalog.
Understand the implementation
Implementation internals — click to expand
src/generated.ts is the static owner of codec and edge ordering. src/current.ts delegates final header, envelope, message, surface, seed, and current request-header validation to the installed Session semantics. The low-level constructor rejects duplicate codecs, duplicate edges, gaps, and entries beyond the current version before any Session read can begin.
Further Exploration
- Migration machinery — catalog construction and dispatch behavior.
- Released v0 to v1 edge — codec and validator ownership.
- Released v1 to v2 edge — Assistant stream embedding and cardinality-changing reference remapping.
- Released V2 to V3 specification — transformations, preservation, and refusal.
- V3 to V4 specification — identity conversion and delivery-generation validation.
- JSONL persistence — immutable generation naming and exclusive publication.
Model Experience
Catalog dispatch
What the model sees
Nothing directly. The catalog only restores the SessionEvent history consumed by request reconstruction.
Token effect
Zero direct tokens.
KV Cache effect
No direct effect; restored history determines cache identity in its consumer.
Known Limitations and Deferred Work
- First-party build inventory only — external migration ownership and distribution are not supported.
- Generated ordering is closed — runtime plugin registration cannot supply a missing historical edge.
Dev Note
Working context for maintainers — click to expand
None.