Files
ThothII/backend/src/routes/catalog-schema.ts
T

303 lines
14 KiB
TypeScript

import type { FastifyInstance, FastifyReply, FastifyRequest } from "fastify";
import { setTimeout as delay } from "node:timers/promises";
import { z } from "zod";
import { isPrincipalContext, requirePermission } from "../auth/authorization.js";
import type { CatalogSyncWorker } from "../catalog/sync-worker.js";
import type { CatalogOperationCoordinator } from "../catalog/operation-coordinator.js";
import {
CatalogConflictError,
CatalogConnectorError,
CatalogOperationInProgressError,
CatalogUnavailableError,
type CatalogRepository,
type CatalogSyncRun,
} from "../catalog/types.js";
const idSchema = z.uuid();
const metadataSchema = z.object({
version: z.number().int().positive(),
description: z.string().max(20_000).nullable().optional(),
generatedDescription: z.string().max(20_000).nullable().optional(),
sensitive: z.boolean().optional(),
sensitivityReason: z.string().max(2_000).nullable().optional(),
}).strict().refine((value) => (
"description" in value
|| "generatedDescription" in value
|| "sensitive" in value
|| "sensitivityReason" in value
)).refine((value) => (
value.sensitivityReason == null || value.sensitive === true
));
const createRunSchema = z.object({
version: z.number().int().positive(),
scope: z.enum(["tables", "columns", "relationships", "all"]),
tableIds: z.array(idSchema).max(10_000).default([]),
}).strict();
const databaseCleanupSchema = z.object({
databaseIds: z.array(idSchema).min(1).max(1_000),
target: z.enum(["tables", "relationships"]),
}).strict();
const tableCleanupSchema = z.object({
tableIds: z.array(idSchema).min(1).max(10_000),
target: z.enum(["columns", "relationships"]),
}).strict();
const confirmationSchema = z.object({ confirmationToken: z.string().uuid() }).strict();
const eventQuerySchema = z.object({ after: z.coerce.number().int().nonnegative().default(0) });
function manage(request: FastifyRequest, reply: FastifyReply) {
return isPrincipalContext(requirePermission(request, reply, "database.manage"));
}
function safeError(reply: FastifyReply, error: unknown) {
if (error instanceof CatalogUnavailableError) {
return reply.code(503).send({ code: "catalog_unavailable", message: "Database catalog is unavailable." });
}
if (error instanceof CatalogOperationInProgressError) {
return reply.code(409).send({
code: "database_operation_in_progress",
message: "A database operation is already in progress.",
});
}
if (error instanceof CatalogConflictError) {
return reply.code(409).send({ code: "schema_sync_conflict", message: error.message });
}
if (error instanceof CatalogConnectorError) {
return reply.code(502).send({
code: "schema_introspection_failed",
message: "The database schema could not be read. Check the connection and credentials, then try again.",
});
}
if (error instanceof z.ZodError) {
return reply.code(400).send({ code: "schema_request_invalid", message: "Schema request is invalid." });
}
return reply.code(500).send({ code: "schema_operation_failed", message: "Schema operation failed." });
}
function normalized(value: string | null): string | null {
return value?.trim() || null;
}
function publicRun(run: CatalogSyncRun) {
const {
observedSnapshot: _snapshot,
leaseOwner: _leaseOwner,
leaseExpiresAt: _leaseExpiresAt,
...result
} = run;
return result;
}
export function catalogSchemaRoutes(
app: FastifyInstance,
deps: {
repository: CatalogRepository;
worker: CatalogSyncWorker;
operations?: CatalogOperationCoordinator;
},
): void {
app.get("/catalog/databases/:databaseId/tables/:tableId/columns", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const params = request.params as { databaseId?: unknown; tableId?: unknown };
const databaseId = idSchema.parse(params.databaseId);
const tableId = idSchema.parse(params.tableId);
if (!(await deps.repository.getTable(databaseId, tableId))) {
return reply.code(404).send({ code: "table_not_found", message: "Catalog table was not found." });
}
return await deps.repository.listColumns(databaseId, tableId);
} catch (error) { return safeError(reply, error); }
});
app.patch("/catalog/databases/:databaseId/tables/:tableId/columns/:columnId", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const params = request.params as { databaseId?: unknown; tableId?: unknown; columnId?: unknown };
const databaseId = idSchema.parse(params.databaseId);
const tableId = idSchema.parse(params.tableId);
const columnId = idSchema.parse(params.columnId);
const input = metadataSchema.parse(request.body);
const operation = async () => {
const current = await deps.repository.getColumn(databaseId, tableId, columnId);
if (!current) return reply.code(404).send({ code: "column_not_found", message: "Catalog column was not found." });
if (current.version !== input.version) {
return reply.code(409).send({ code: "column_stale", message: "Column metadata changed. Reload and try again." });
}
const nextSensitive = input.sensitive ?? current.sensitive;
const nextSensitivityReason = nextSensitive
? ("sensitivityReason" in input
? normalized(input.sensitivityReason ?? null)
: current.sensitivityReason)
: null;
const updated = await deps.repository.updateColumnMetadata(
databaseId,
tableId,
columnId,
input.version,
"description" in input ? normalized(input.description ?? null) : current.description,
"generatedDescription" in input
? normalized(input.generatedDescription ?? null)
: current.generatedDescription,
input.sensitive,
nextSensitivityReason,
);
if (!updated) return reply.code(409).send({ code: "column_stale", message: "Column metadata changed. Reload and try again." });
return updated;
};
return deps.operations ? await deps.operations.run(databaseId, operation) : await operation();
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/databases/metadata-cleanup", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const input = databaseCleanupSchema.parse(request.body);
const databaseIds = [...new Set(input.databaseIds)];
const operation = async () => await deps.repository.deleteDatabaseMetadata(databaseIds, input.target);
const deleted = deps.operations
? await deps.operations.runMany(databaseIds, operation)
: await operation();
if (!deleted) {
return reply.code(404).send({
code: "database_not_found",
message: "One or more database configurations were not found.",
});
}
return deleted;
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/databases/:databaseId/tables/metadata-cleanup", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const databaseId = idSchema.parse((request.params as { databaseId?: unknown }).databaseId);
const input = tableCleanupSchema.parse(request.body);
const tableIds = [...new Set(input.tableIds)];
const operation = async () => await deps.repository.deleteTableMetadata(databaseId, tableIds, input.target);
const deleted = deps.operations
? await deps.operations.run(databaseId, operation)
: await operation();
if (!deleted) {
return reply.code(404).send({
code: "table_not_found",
message: "One or more catalog tables were not found.",
});
}
return deleted;
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/databases/:databaseId/sync-runs", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const databaseId = idSchema.parse((request.params as { databaseId?: unknown }).databaseId);
const input = createRunSchema.parse(request.body);
const database = await deps.repository.get(databaseId);
if (!database) return reply.code(404).send({ code: "database_not_found", message: "Database configuration was not found." });
if (database.version !== input.version) {
return reply.code(409).send({ code: "database_stale", message: "Database configuration changed. Reload and try again." });
}
return reply.code(202).send(publicRun(await deps.worker.start(database, input.scope, input.tableIds)));
} catch (error) { return safeError(reply, error); }
});
app.get("/catalog/databases/:databaseId/sync-runs", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const databaseId = idSchema.parse((request.params as { databaseId?: unknown }).databaseId);
if (!(await deps.repository.get(databaseId))) {
return reply.code(404).send({ code: "database_not_found", message: "Database configuration was not found." });
}
return (await deps.repository.listSyncRuns(databaseId)).map(publicRun);
} catch (error) { return safeError(reply, error); }
});
app.get("/catalog/sync-runs/:runId", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
const run = await deps.repository.getSyncRun(runId);
return run ? publicRun(run) : reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/sync-runs/:runId/confirm", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
const { confirmationToken } = confirmationSchema.parse(request.body);
const run = await deps.worker.confirm(runId, confirmationToken);
return run ? publicRun(run) : reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/sync-runs/:runId/cancel", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
const run = await deps.worker.cancel(runId);
return run ? publicRun(run) : reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
} catch (error) { return safeError(reply, error); }
});
app.post("/catalog/sync-runs/:runId/retry", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
const run = await deps.worker.retry(runId);
return run ? reply.code(202).send(publicRun(run)) : reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
} catch (error) { return safeError(reply, error); }
});
app.get("/catalog/sync-runs/:runId/events", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
let after = eventQuerySchema.parse(request.query).after;
const headerCursor = Number(request.headers["last-event-id"]);
if (Number.isInteger(headerCursor) && headerCursor >= 0) after = Math.max(after, headerCursor);
if (!(await deps.repository.getSyncRun(runId))) {
return reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
}
reply.hijack();
reply.raw.writeHead(200, {
"content-type": "text/event-stream; charset=utf-8",
"cache-control": "no-cache, no-transform",
connection: "keep-alive",
"x-accel-buffering": "no",
});
const controller = new AbortController();
request.raw.once("close", () => controller.abort());
let lastRunUpdate = "";
while (!controller.signal.aborted) {
const events = await deps.repository.listSyncEvents(runId, after);
for (const event of events) {
after = event.sequence;
reply.raw.write(`id: ${event.sequence}\nevent: log\ndata: ${JSON.stringify(event)}\n\n`);
}
const run = await deps.repository.getSyncRun(runId);
if (!run) break;
if (run.updatedAt !== lastRunUpdate) {
lastRunUpdate = run.updatedAt;
reply.raw.write(`event: run\ndata: ${JSON.stringify(publicRun(run))}\n\n`);
}
if (["succeeded", "failed", "cancelled", "interrupted"].includes(run.state)) break;
await delay(500, undefined, { signal: controller.signal }).catch(() => undefined);
}
reply.raw.end();
return reply;
} catch (error) { return safeError(reply, error); }
});
app.get("/catalog/sync-runs/:runId/events-list", async (request, reply) => {
if (!manage(request, reply)) return reply;
try {
const runId = idSchema.parse((request.params as { runId?: unknown }).runId);
const after = eventQuerySchema.parse(request.query).after;
if (!(await deps.repository.getSyncRun(runId))) {
return reply.code(404).send({ code: "sync_run_not_found", message: "Synchronization run was not found." });
}
return await deps.repository.listSyncEvents(runId, after);
} catch (error) { return safeError(reply, error); }
});
}