The IR called them react-message and django-message, so a FastAPI channel had
to declare a DjangoMessage. They are client-message and server-message now,
and the direction words hold wherever a channel is declared: Params /
ClientMessage / ServerMessage, with mizan-core deriving <Pascal>Params and
friends so no backend names a type itself. Django's ReactChannel and
FastAPI's ReactChannel are both Channel.
mizan-fastapi never registered a channels extension, so build_ir() emitted no
channel at all and every payload type was invisible to codegen. It registers
one now. RegistryExtension is an ABC requiring all(), which is what the IR
reads — an extension that cannot enumerate its registrations no longer exists.
The gate that should have caught the rename could not: tests/afi registered no
channel because mizan-rust had no channel registry to register one in, so a
five-package rename of the wire contract passed byte-parity without a channel
byte crossing it. mizan-rust grows ChannelSlotKind, a CHANNELS slice, a
#[mizan::channel] macro, and KDL emission whose wire_to_pascal matches Python's
split; the AFI fixture now carries a channel with every slot and one with a
single slot, so all three backends prove the contract byte for byte.
MizanChannel held three Option<String> beside three has_*() predicates and
unwrapped them with defaults; it holds an ordered slot vector, so an absent
slot is absent rather than defaulted. The channels target emitted a React
hooks file that a stage1-only consumer could not compile — react emits that
now. The codegen's parity tests byte-compared emitted source against baselines
without ever compiling it: they compile the generated crate and run its tests,
import the generated Python package and call every method, and typecheck each
TypeScript target against a consumer.
Also fixed at source: app_visitor printed its import diagnostic to stdout, the
stream export_mizan_ir writes KDL to, so a failed import silently corrupted the
IR; the apps root was hardcoded to "apps"; _default_literal crashed build_ir on
any non-JSON-serializable field default; Django and mizan-core derived Pascal
names two different ways, disagreeing on every dotted channel name.
ir.py builds a document and renders templates/ir/document.kdl.j2 rather than
appending KDL strings with hand-tracked indentation, and named types resolve to
a fixed point — a model reachable only through a union branch was referenced by
a ref that no type block ever defined.
The rest is the write-gate's own classifiers run over the standing tree:
relative imports, silent swallows, Protocol contracts that should be ABCs,
emitters hand-rendering target source, catch-all arms over closed enums, and
comments narrating the project rather than the code.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Replaces the transitional OpenAPI 3.0 + `x-mizan-*` extensions
substrate with the canonical Mizan IR as KDL, per docs/AFI_ARCHITECTURE.md:
"KDL is the contract; everything else (REST envelopes, OpenAPI
documents, framework idioms) is sediment around it."
End-to-end cutover. No transitional path left on main.
Forward direction:
cores/mizan-python/src/mizan_core/ir.py
build_ir() walks mizan_core.registry, introspects Pydantic
models directly (no JSON-Schema indirection), and emits the
Mizan IR document. The KDL grammar is locked in this file's
module docstring.
Backends emit KDL:
backends/mizan-fastapi/src/mizan_fastapi/ir.py
`python -m mizan_fastapi.ir <module>` — CLI entry point.
backends/mizan-django/.../management/commands/export_mizan_ir.py
`manage.py export_mizan_ir` — Django mgmt command.
Codegen consumes KDL:
protocol/mizan-codegen/Cargo.toml: + kdl = "6"
protocol/mizan-codegen/src/ir.rs: NamedType { Struct/List/Enum/Alias }
+ TypeShape { Primitive/Ref/List/Optional/Enum/Union } sum types,
replacing the JsonSchema sprawl. KDL parser walks the
`kdl::KdlDocument` tree into typed Rust structs.
protocol/mizan-codegen/src/fetch.rs: subprocess command switches
to the new IR-export entry points.
All emit modules (stage1 / react / python / rust / vue / svelte /
channels) port their type-walkers from JsonSchema to the new
sum types — case analysis collapses substantially.
Substrate-honesty wins beyond the moat closure:
- `int | bool` multi-arm unions land as `TypeShape::Union` (was
silently coerced to "string" before).
- `<CamelName>Output = list[T]` returns emit as named alias
types instead of struct-shaped wrappers, so consumer code
`.map()` works directly on the type.
- Pydantic field defaults flow through to `default` properties
in KDL, then back to non-optional shape in every target.
Deleted:
- backends/mizan-fastapi/src/mizan_fastapi/{cli,schema}.py
- backends/mizan-django/.../export_mizan_schema.py
- openapi-bearing half of mizan/export/__init__.py (edge
manifest generator preserved — separate concern).
- tests/afi/schema_normalizer.py
- tests/fixtures/{afi_schema.json, channels_schema.json}
- tests/fixtures/js_* baseline directories.
Verification:
- 20 mizan-codegen unit tests green (IR deserialization,
byte-equivalence parity across stage1/rust/python/react/vue/svelte
against fresh KDL-driven baselines, channels structural).
- tests/rust/run_wire_parity.py: 12/12 probes green driving
the binary end-to-end through KDL.
- Blazr studio-ui typechecks against the regenerated React client.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Mizan codegen substrate moves off JavaScript template-literal emission
onto a compiled Rust binary that consumes the same OpenAPI + x-mizan-* IR
the JS substrate consumed. Three structural wins fall out of one move:
1. Moat closes. The codegen logic (how `affects` becomes auto-invalidation,
how named contexts collapse onto bundled fetches, how the registry-to-
Provider mapping is shaped) ships compiled instead of as source bytes
in every consumer's node_modules.
2. Pattern F (lines.push append-walls) becomes structurally unauthorable.
The emit substrate is askama templates in templates/<target>/*.j2 —
actual target-language files with {{ ... }} substitution markers,
syntax-highlighted natively, type-checked against the render context
structs at compile time. The Rust emit modules build typed render
contexts and call .render(); no string-builder surface exists.
3. OpenAPI `default`-bearing fields now emit as non-optional in TS / Python
/ Rust — the server always populates them, so consumer code reads them
without nullable checks. Surfaced by Blazr's typecheck on regeneration.
Layout:
frontends/mizan-rust/ — Rust port of @mizan/base; #[cfg(feature="pyo3")]
exposes PyMizanClient for the Python target.
protocol/mizan-codegen/ — codegen binary source + askama templates.
protocol/mizan-generate/ — npm-package shim. bin/launcher.mjs dispatches
to the platform-appropriate prebuilt binary.
Old generator/ JS tree deleted.
tests/rust/ — wire-parity drivers. drive_kernel exercises
raw client.call() / fetch_context(); drive_emitted
exercises the typed crate the codegen emits.
tests/afi/afi_codegen_app.py — codegen entrypoint module (imports + registers).
backends/mizan-fastapi/.../schema.py — adds outputNullable so the Rust
codegen can wrap T | None responses in Option<T>.
Verification:
- 20 mizan-codegen tests green (IR deserialization, byte-equivalent
parity vs JS baseline for stage1/rust/python/react/vue/svelte,
structural test for channels).
- tests/rust/run_wire_parity.py — 12/12 probes green via the Rust binary
driving the FastAPI fixture end-to-end.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The codegen consumes a schema from any backend and emits typed client
code for any frontend — it doesn't belong inside a backend adapter.
That placement was historical sediment from when there was only a
Django backend; it predates the AFI generalization.
New top-level slot: `protocol/` for protocol-level tooling. Tree is
now:
backends/ server protocol adapters
frontends/ client kernel + per-framework adapters
cores/ shared language-level primitives
protocol/ protocol-level tooling
workers/ runtime workers / bridges
Codegen moves to `protocol/mizan-generate/`. Same file layout under
`generator/` (cli.mjs, lib/), preserved via git mv.
Package metadata cleaned up:
- name: "generate" (placeholder) → "mizan-generate"
- description filled in
- type: module (cli.mjs is .mjs ESM, was previously declared "commonjs")
- bin entry added so `npx mizan-generate --config <config.mjs>` works
once the package is published, instead of `node path/to/cli.mjs`.
Path-reference fixups:
- backends/mizan-django/README.md: `node path/to/...` → `npx mizan-generate`
- backends/mizan-fastapi/README.md: same
- ISSUES.md: file paths in three issue entries
- CLAUDE.md: codegen description + Package Layout section refreshed
(added protocol/, mizan-fastapi entry, mizan-python entry)
- docs/AFI_ARCHITECTURE.md: Package Layout refreshed identically
Verified codegen runs from new location: regenerated the FastAPI
example harness's api/ output, identical to pre-move.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
