Files
ThothII/backend/test/workspaces-schema.test.ts
T

234 lines
8.5 KiB
TypeScript

import { expect, test } from "vitest";
import * as workspaceSchema from "../src/workspaces/schema.js";
import { parseWorkspaceYaml, serializeWorkspaceYaml, validateCanonicalWorkspace } from "../src/workspaces/schema.js";
export const validYaml = `workspace:
schema_version: 2
id: psd-clinical
name: Policlinico San Donato
description: Clinical data warehouse workspace
language: it
dwh:
engine: postgres
database: postgres
schema: datawarehouse
port: 5432
timeout_ms: 5000
supported_transports:
- postgres_direct
- rest_api
- ssh_tunnel
semantic_index:
vector_store:
engine: pgvector
database: postgres
schema: vectors
collection: clinical_documents
dimensions: 768
distance: cosine
port: 5432
timeout_ms: 5000
supported_transports:
- pgvector_direct
- rest_api
- ssh_tunnel
embedding:
provider: ollama_compatible
model: nomic-embed-text-v2-moe
dimensions: 768
timeout_ms: 5000
llm_policy:
default: zai/glm-5.2
allowed:
- zai/glm-5.2
- openai/gpt-5
`;
test("rejects a workspace whose embedding dimensions differ from its collection", () => {
expect(() => parseWorkspaceYaml(validYaml.replace("dimensions: 768", "dimensions: 1536")))
.toThrow(/dimensions/i);
});
test("rejects an LLM default outside its allowlist", () => {
expect(() => parseWorkspaceYaml(validYaml.replace("- zai/glm-5.2", "- openai/gpt-5")))
.toThrow(/allowlist/i);
});
test("rejects unknown keys and invalid immutable IDs", () => {
expect(() => parseWorkspaceYaml(validYaml.replace(" language: it", " language: it\n label: PSD")))
.toThrow(/unrecognized key/i);
expect(() => parseWorkspaceYaml(validYaml.replace("id: psd-clinical", "id: PSD")))
.toThrow(/id/i);
});
test("rejects executable or otherwise custom YAML tags in canonical descriptors", () => {
expect(() => parseWorkspaceYaml(validYaml.replace(
"name: Policlinico San Donato",
"name: !command echo-never-execute",
))).toThrow(/tag|yaml/i);
});
test("accepts optional connection ports and timeouts but rejects unsafe values", () => {
expect(parseWorkspaceYaml(validYaml).dwh.port).toBe(5432);
expect(() => parseWorkspaceYaml(validYaml.replace("port: 5432", "port: 0")))
.toThrow(/port/i);
expect(() => parseWorkspaceYaml(validYaml.replace("port: 5432", "port: 65536")))
.toThrow(/port/i);
expect(() => parseWorkspaceYaml(validYaml.replace("timeout_ms: 5000", "timeout_ms: 0")))
.toThrow(/timeout/i);
});
test("requires explicit vector database and schema identities with strict diagnostic declarations", () => {
const diagnosticWorkspace = validYaml.replace(
" engine: pgvector\n database: postgres",
" engine: pgvector\n database: vector_database",
).replace(
"llm_policy:\n",
"diagnostics:\n"
+ " dwh_rest:\n"
+ " method: POST\n"
+ " path: /rpc/ping\n"
+ " auth: bearer\n"
+ " response:\n"
+ " database: database\n"
+ " schema: schema\n"
+ " vector_rest:\n"
+ " metadata:\n"
+ " method: GET\n"
+ " path: /metadata\n"
+ " auth: bearer\n"
+ " response:\n"
+ " collection: collection\n"
+ " dimensions: dimensions\n"
+ " distance: distance\n"
+ " reversible_probe:\n"
+ " method: POST\n"
+ " path: /rpc/diagnostic_vector_probe\n"
+ " auth: bearer\n"
+ " response: { operation: operation }\n"
+ " embedding:\n"
+ " method: GET\n"
+ " path: /models\n"
+ " auth: none\n"
+ " response:\n"
+ " model: model\n"
+ " dimensions: dimensions\n"
+ "llm_policy:\n",
);
expect(parseWorkspaceYaml(diagnosticWorkspace).semantic_index.vector_store).toMatchObject({
database: "vector_database",
schema: "vectors",
});
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace("database: vector_database", 'database: " "')))
.toThrow(/database/i);
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace("schema: vectors", 'schema: " "')))
.toThrow(/schema/i);
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace("method: POST", "method: PATCH")))
.toThrow(/method/i);
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace(" distance: distance", " distance: distance\n extra: ignored")))
.toThrow(/unrecognized key/i);
for (const unsafePath of [
"//diagnostic.invalid/rpc", "'/\\\\diagnostic'", "'/rpc\\\\diagnostic'", "'/rpc/%5Cdiagnostic'", "'/rpc/\u0001'",
]) {
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace("path: /rpc/ping", `path: ${unsafePath}`)))
.toThrow(/origin-relative|path/i);
}
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace("auth: bearer", "auth: basic")))
.toThrow(/auth/i);
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace(" schema: schema", " schema: schema\n status: status")))
.toThrow(/unrecognized key/i);
expect(() => parseWorkspaceYaml(diagnosticWorkspace.replace(" schema: schema", " schema: bad field")))
.toThrow(/response field/i);
});
test("requires a reversible writer probe to declare the response operation it verifies", () => {
const writerProbe = validYaml.replace("llm_policy:\n", `diagnostics:
vector_rest:
metadata:
method: GET
path: /metadata
auth: bearer
response: { collection: collection, dimensions: dimensions, distance: distance }
reversible_probe:
method: POST
path: /diagnostic-probe
auth: bearer
llm_policy:
`);
expect(() => parseWorkspaceYaml(writerProbe)).toThrow(/response|operation/i);
});
test("keeps v1 descriptors readable but requires explicit migration before v2 operations", () => {
const v1WithoutVectorIdentity = validYaml.replace("schema_version: 2", "schema_version: 1").replace(
" database: postgres\n schema: vectors\n", "",
);
expect(() => parseWorkspaceYaml(v1WithoutVectorIdentity)).not.toThrow();
expect(() => parseWorkspaceYaml(validYaml)).not.toThrow();
const v1 = parseWorkspaceYaml(v1WithoutVectorIdentity);
const v2 = parseWorkspaceYaml(validYaml);
expect(v1.workspace.schema_version).toBe(1);
expect(() => validateCanonicalWorkspace(v1)).toThrow(/migrat/i);
expect(v2.workspace.schema_version).toBe(2);
const migrate = (workspaceSchema as { migrateWorkspaceV1ToV2?: unknown }).migrateWorkspaceV1ToV2;
expect(migrate).toBeTypeOf("function");
const migrated = (migrate as (workspace: typeof v1, identity: { database: string; schema: string }) => unknown)(v1, {
database: "vector_database",
schema: "vectors",
});
expect(validateCanonicalWorkspace(migrated)).toMatchObject({
workspace: { schema_version: 2 },
semantic_index: { vector_store: { database: "vector_database", schema: "vectors" } },
});
});
test("constructs diagnostic URLs only when the resolved URL remains on the service origin", () => {
const resolveDiagnosticUrl = (workspaceSchema as { resolveDiagnosticUrl?: unknown }).resolveDiagnosticUrl;
expect(resolveDiagnosticUrl).toBeTypeOf("function");
expect((resolveDiagnosticUrl as (baseUrl: string, path: string) => URL)("https://service.example/base", "/rpc/ping"))
.toMatchObject({ href: "https://service.example/rpc/ping" });
expect(() => (resolveDiagnosticUrl as (baseUrl: string, path: string) => URL)(
"https://service.example/base", "//diagnostic.invalid/rpc",
)).toThrow(/origin/i);
});
test("rejects REST diagnostic declarations without their matching connector transport", () => {
const diagnostics = `diagnostics:
dwh_rest:
method: POST
path: /rpc/ping
auth: bearer
response:
database: database
schema: schema
vector_rest:
metadata:
method: GET
path: /metadata
auth: bearer
response:
collection: collection
dimensions: dimensions
distance: distance
llm_policy:
`;
const declared = validYaml.replace("llm_policy:\n", diagnostics);
expect(() => parseWorkspaceYaml(declared.replace(" - rest_api\n", ""))).toThrow(/dwh_rest/i);
expect(() => parseWorkspaceYaml(declared.replace(" - rest_api\n", " - rest_api\n", 1).replace(
" - rest_api\n", "",
))).toThrow(/vector_rest/i);
});
test("serializes canonical YAML that parses back to the same workspace", () => {
const workspace = parseWorkspaceYaml(validYaml);
const serialized = serializeWorkspaceYaml(workspace);
expect(serializeWorkspaceYaml(parseWorkspaceYaml(serialized))).toBe(serialized);
expect(parseWorkspaceYaml(serialized)).toEqual(workspace);
});