Each root JSON key ('info', 'pushing', 'print') now gets its own category
folder under request/ and report/, so a category can hold multiple command
schemas without flattening them into one directory - v1/request.schema.json
and v1/report.schema.json route into these via oneOf per category (also
future-proofs for a category ever needing more than one shape).
Split pushall's ams/xcam fields the same way module_info was already
split: ams.schema.json/ams_unit.schema.json/ams_tray.schema.json/
xcam.schema.json/print.schema.json, each referencing the next by relative
$ref. ams_tray.schema.json uses oneOf for the empty-vs-loaded tray shapes
- verified (and initially got wrong, then fixed) that LoadedTray needs
every field required, not just id, or a bare {"id":"0"} validates
against both branches and oneOf fails, since oneOf demands exactly one
match.
Real correction: the report/ category for what a pushall request
triggers is "print", not "pushing" - re-reading the original examples,
pushall's REQUEST goes to "pushing", but the resulting state-push stream
arrives under "print", a different root key. report/pushing/ (my
original mistake) is removed; report/print/ holds the corrected content.
request/pushing/ is untouched - that side was correct.
Also: v2/ removed (found as broken copy-pasted stubs still pointing at
v1 $ids - no real V2-specific content yet, better to have nothing than
broken duplicates); all $ids switched to real, verified Gitea raw-file
URLs (raw/branch/main/<path> - confirmed this is the correct pattern,
not raw/<path>) instead of a placeholder domain; fixed a broken relative
$ref in the request-side files left over from an earlier move (still
pointing 2 levels up instead of 3 after nesting one level deeper).
Re-validated everything against real examples after restructuring,
including the full ams/xcam print report and two negative cases -
see schemas/bambulab/README.md for the validation script.
5.3 KiB
bambulab/ JSON Schemas
Documentation/validation schemas for Bambu's raw MQTT wire format, mirroring
continuum-proxy's src/printer/bambu_commands/ — not used for codegen (the
Rust structs there are hand-written; see that module's doc comment for why),
but genuinely useful on their own: language-agnostic documentation of the
wire format, and usable to validate a captured MQTT payload without writing
any Rust to do it.
Layout
bambulab/
envelope.schema.json the base — sequence_id + command, shared by
every command in every generation, so it lives
here unversioned, not duplicated under v1/v2
v1/
request.schema.json BambuRequest dispatcher — one key per category
("info", "pushing", ...), each a oneOf (a
category can have more than one request shape)
report.schema.json BambuReport dispatcher, same idea
request/
info/
get_version.schema.json
pushing/
pushall.schema.json
report/
info/
get_version.schema.json
module_info.schema.json
print/
print.schema.json the ongoing print-state push
ams.schema.json
ams_unit.schema.json
ams_tray.schema.json
xcam.schema.json
v2/ add the same split once V2's wire format is
known to differ from V1's — removed for now
rather than leave broken copy-pasted stubs
One real correction baked into this layout: the report category for
whatever a pushall request triggers is report/print/, not
report/pushing/. Sending pushall (which does go to the pushing key
on the request side) makes the printer start/refresh an ongoing state-push
stream, and that stream arrives under the print key — a different root
key from the request that triggered it. A category folder here means
"what root key does this arrive under", not "what request caused it" — the
first version of this schema got that wrong (see git history) before a
closer read of the real capture caught it.
The "inheritance" pattern
Every command/response schema extends envelope.schema.json with
allOf + $ref:
{
"allOf": [
{ "$ref": "../../../envelope.schema.json" },
{ "type": "object", "properties": { "command": { "const": "get_version" } } }
]
}
This is genuinely closer to real inheritance than anything else in this
project's schema stack (proto, Rust): the instance must satisfy the base
schema and the extension schema simultaneously, and — verified with
negative test cases, not just asserted — envelope.schema.json's
required: ["sequence_id", "command"] is actually enforced on every
schema that extends it, not just copy-pasted as documentation.
v1/report/print/ams_tray.schema.json is the other JSON-Schema-specific
tool worth knowing: oneOf, for "this is exactly one of several
genuinely different shapes" (an AMS tray slot is either empty — just
id — or loaded — the full field set — never something in between).
oneOf requires exactly one branch to match; watch out for a base-style
branch (like EmptyTray) accidentally also matching a more specific
branch's instance if the specific branch doesn't require enough fields —
verified and fixed once already here, see that file's description.
One gotcha if you add a new base-style schema: don't set
"additionalProperties": false on something meant to be allOf-extended.
allOf validates every sub-schema against the whole instance
independently — it doesn't merge object schemas — so a closed base schema
would reject every field the extending schema adds. Leave
additionalProperties unset (defaults to allowed) on anything used as a
base; enforce closedness only on the top-level dispatcher schemas
(v1/request.schema.json/v1/report.schema.json do this correctly —
check their additionalProperties: false).
$id convention
Every schema's $id is its real Gitea raw-file URL:
https://git.octoturge.com/Continuum/continuum-schemas/raw/branch/main/schemas/bambulab/...
— a genuinely dereferenceable URI (verified: raw/branch/main/<path> is
the correct Gitea pattern, not raw/<path> or raw/<branch>/<path>),
rather than a made-up placeholder domain. $refs between files stay
relative paths; $id is what a validator uses as the base URI to resolve
those relative refs, and it's also what lets a schema be fetched
standalone by anything that wants to dereference it directly.
Validating something
$ref resolution here relies on each schema's $id as the base URI for
its own relative refs (standard JSON Schema behavior) — register schemas
by $id, not by filename, or you'll hit collisions (get_version.schema.json
exists under both request/info/ and report/info/):
from referencing import Registry, Resource
from jsonschema import Draft202012Validator
import json, glob
resources = [
(json.load(open(p))["$id"], Resource.from_contents(json.load(open(p))))
for p in glob.glob("**/*.schema.json", recursive=True)
]
registry = Registry().with_resources(resources)
schema = json.load(open("v1/request.schema.json"))
Draft202012Validator(schema, registry=registry).validate(
{"info": {"sequence_id": "0", "command": "get_version"}}
)