H1-H10 优化OCR实时体验闭环

这个提交包含在:
2026-05-09 11:48:57 +08:00
父节点 549bd81e30
当前提交 c26cf9b954
共修改 34 个文件,包含 1217 行新增和 184 行删除
+4
查看文件
@@ -73,8 +73,10 @@ const travelAuthorizationsRoutes = require("./routes/travelAuthorizations");
const travelTransactionsRoutes = require("./routes/travelTransactions");
const travelReconciliationRoutes = require("./routes/travelReconciliation");
const aiRoutes = require("./routes/ai");
const eventsRoutes = require("./routes/events");
const { startJobs } = require("./jobs/jobSystem");
const { initMetrics } = require("./observability");
const { createAppEventBus } = require("./services/events");
const { AppError } = require("./utils/AppError");
const { buildDataEnvelopeSchema, buildErrorEnvelopeSchema } = require("./utils/openapi");
@@ -122,6 +124,7 @@ async function buildApp() {
});
app.decorate("config", config);
app.decorate("appEvents", createAppEventBus());
await fs.mkdir(config.storageDir, { recursive: true });
await app.register(cookie);
@@ -332,6 +335,7 @@ async function buildApp() {
await app.register(authRoutes, { prefix: "/api/auth" });
await app.register(wecomAuthRoutes, { prefix: "/api/auth" });
await app.register(eventsRoutes, { prefix: "/api" });
await app.register(meRoutes, { prefix: "/api" });
await app.register(reimbursementRoutes, { prefix: "/api" });
await app.register(invoicePoolRoutes, { prefix: "/api" });
+66 -10
查看文件
@@ -1,6 +1,7 @@
const cron = require("node-cron");
const { buildMissingRequiredOcrError, runOcrForInvoice } = require("../services/ocr");
const { publishInvoiceUpdated } = require("../services/events");
const { resolveInvoiceDedupBlocker } = require("../services/invoiceGovernance");
const { readFileBuffer } = require("../services/storage");
@@ -70,7 +71,30 @@ async function claimPendingInvoices(prisma, config, workerId) {
ocr_lock_owner = ${workerId}
FROM candidates
WHERE i.id = candidates.id
RETURNING i.id, i.owner_employee_id, i.file_sha256, i.ocr_provider, i.ocr_type, i.storage_path, i.mime_type, i.ocr_attempts
RETURNING i.id, i.owner_employee_id, i.uploaded_by_employee_id, i.file_sha256, i.ocr_provider, i.ocr_type, i.storage_path, i.mime_type, i.ocr_attempts
`;
}
async function claimPendingInvoiceById(prisma, config, workerId, invoiceId) {
const lockTtlSeconds = Math.max(60, Number(config.ocrLockTtlSeconds || 300));
return prisma.$queryRaw`
WITH candidates AS (
SELECT id
FROM "invoice_pool"
WHERE id = ${invoiceId}::uuid
AND status = 'pending_ocr'
AND (ocr_locked_at IS NULL OR ocr_locked_at < NOW() - make_interval(secs => ${lockTtlSeconds}))
ORDER BY created_at ASC
LIMIT 1
FOR UPDATE SKIP LOCKED
)
UPDATE "invoice_pool" AS i
SET ocr_locked_at = NOW(),
ocr_lock_owner = ${workerId}
FROM candidates
WHERE i.id = candidates.id
RETURNING i.id, i.owner_employee_id, i.uploaded_by_employee_id, i.file_sha256, i.ocr_provider, i.ocr_type, i.storage_path, i.mime_type, i.ocr_attempts
`;
}
@@ -226,11 +250,12 @@ async function markInvoiceAsFailed(prisma, claimedInvoice, errorMessage, config)
const now = new Date();
const nextAttempts = Number(claimedInvoice.ocr_attempts || 0) + 1;
const isTerminal = nextAttempts >= config.ocrMaxAttempts;
const nextStatus = isTerminal ? "ocr_failed" : "pending_ocr";
await prisma.invoicePool.update({
where: { id: claimedInvoice.id },
data: {
status: isTerminal ? "ocr_failed" : "pending_ocr",
status: nextStatus,
ocr_provider: getEffectiveOcrProvider(config),
ocr_attempts: nextAttempts,
ocr_last_error: errorMessage,
@@ -239,6 +264,8 @@ async function markInvoiceAsFailed(prisma, claimedInvoice, errorMessage, config)
ocr_lock_owner: null,
},
});
return nextStatus;
}
async function processClaimedInvoice(prisma, config, claimedInvoice, workerId, logger) {
@@ -276,7 +303,7 @@ async function processClaimedInvoice(prisma, config, claimedInvoice, workerId, l
},
"OCR invoice marked as failed for missing required fields",
);
return { outcome: "failed" };
return { outcome: "failed", status: "ocr_failed" };
}
const duplicate = await resolveInvoiceDedupBlocker(
@@ -302,14 +329,14 @@ async function processClaimedInvoice(prisma, config, claimedInvoice, workerId, l
},
"OCR invoice marked as duplicate",
);
return { outcome: "duplicate" };
return { outcome: "duplicate", status: "ocr_failed" };
}
await markInvoiceAsAvailable(prisma, config, claimedInvoice.id, normalized, ocrResultId);
return { outcome: "available" };
return { outcome: "available", status: "available" };
} catch (error) {
const message = trimErrorMessage(error);
await markInvoiceAsFailed(prisma, claimedInvoice, message, config);
const status = await markInvoiceAsFailed(prisma, claimedInvoice, message, config);
logger.error(
{
err: error,
@@ -320,12 +347,11 @@ async function processClaimedInvoice(prisma, config, claimedInvoice, workerId, l
},
"OCR processing failed",
);
return { outcome: "failed" };
return { outcome: "failed", status };
}
}
async function runOcrOnce(prisma, config, logger, workerId = `ocr:${process.pid}`) {
const claimed = await claimPendingInvoices(prisma, config, workerId);
async function processClaimedInvoices(prisma, config, logger, workerId, claimed, options = {}) {
if (!claimed.length) {
return { claimed: 0, available: 0, duplicate: 0, failed: 0 };
}
@@ -341,11 +367,31 @@ async function runOcrOnce(prisma, config, logger, workerId = `ocr:${process.pid}
} else if (result?.outcome === "failed") {
stats.failed += 1;
}
if (typeof options.onInvoiceProcessed === "function") {
await options.onInvoiceProcessed({
invoiceId: item.id,
status: result?.status ?? item.status ?? null,
outcome: result?.outcome ?? null,
ownerEmployeeId: item.owner_employee_id ?? null,
uploadedByEmployeeId: item.uploaded_by_employee_id ?? null,
});
}
}
return stats;
}
async function runOcrOnce(prisma, config, logger, workerId = `ocr:${process.pid}`, options = {}) {
const claimed = await claimPendingInvoices(prisma, config, workerId);
return processClaimedInvoices(prisma, config, logger, workerId, claimed, options);
}
async function runOcrForInvoiceId(prisma, config, logger, invoiceId, workerId = `ocr:${process.pid}`, options = {}) {
const claimed = await claimPendingInvoiceById(prisma, config, workerId, invoiceId);
return processClaimedInvoices(prisma, config, logger, workerId, claimed, options);
}
function startOcrJob(fastify) {
if (!["tencent", "mock"].includes(fastify.config.ocrProvider)) {
fastify.log.info({ provider: fastify.config.ocrProvider }, "OCR job disabled");
@@ -359,6 +405,12 @@ function startOcrJob(fastify) {
let running = false;
const workerId = `ocr-worker:${process.pid}`;
const onInvoiceProcessed = (event) =>
publishInvoiceUpdated(fastify, {
invoiceId: event.invoiceId,
status: event.status,
audienceEmployeeIds: [event.ownerEmployeeId, event.uploadedByEmployeeId],
});
const task = cron.schedule(fastify.config.ocrPollCron, async () => {
if (running) {
@@ -367,7 +419,9 @@ function startOcrJob(fastify) {
running = true;
try {
const result = await runOcrOnce(fastify.prisma, fastify.config, fastify.log, workerId);
const result = await runOcrOnce(fastify.prisma, fastify.config, fastify.log, workerId, {
onInvoiceProcessed,
});
fastify.metrics?.jobRunsTotal?.inc?.({
job: "ocr.tick",
driver: "cron",
@@ -423,9 +477,11 @@ function startOcrJob(fastify) {
module.exports = {
OCR_DUPLICATE_ERROR,
claimPendingInvoiceById,
claimPendingInvoices,
getEffectiveOcrProvider,
markInvoiceAsQualityFailed,
runOcrForInvoiceId,
runOcrOnce,
startOcrJob,
};
+91 -3
查看文件
@@ -2,16 +2,18 @@ const PgBoss = require("pg-boss");
const { revokeExpiredDelegations } = require("./delegationRevokeJob");
const { archiveUnusedInvoices } = require("./invoiceArchiveJob");
const { runOcrOnce } = require("./ocrJob");
const { runOcrForInvoiceId, runOcrOnce } = require("./ocrJob");
const { runVerificationForRecord } = require("./verificationJob");
const { runIntegrationOutboxOnce } = require("./integrationOutboxJob");
const { runErpMasterDataSyncOnce } = require("./erpMasterDataSyncJob");
const { publishInvoiceUpdated } = require("../services/events");
const {
FEATURE_FLAG_KEYS,
evaluateFeatureFlag,
} = require("../services/featureFlags");
const JOB_OCR_TICK = "ocr.tick";
const JOB_OCR_INVOICE = "ocr.invoice";
const JOB_DELEGATION_REVOKE = "delegation.revoke-expired";
const JOB_INVOICE_ARCHIVE = "invoice.archive-unused";
const JOB_INVOICE_VERIFY = "invoice.verify";
@@ -43,6 +45,7 @@ async function startPgBossJobs(fastify) {
});
await boss.createQueue(JOB_OCR_TICK);
await boss.createQueue(JOB_OCR_INVOICE);
await boss.createQueue(JOB_DELEGATION_REVOKE);
await boss.createQueue(JOB_INVOICE_ARCHIVE);
await boss.createQueue(JOB_INVOICE_VERIFY);
@@ -51,6 +54,13 @@ async function startPgBossJobs(fastify) {
await boss.createQueue(JOB_ERP_MASTER_DATA_SYNC);
}
const onOcrInvoiceProcessed = (event) =>
publishInvoiceUpdated(fastify, {
invoiceId: event.invoiceId,
status: event.status,
audienceEmployeeIds: [event.ownerEmployeeId, event.uploadedByEmployeeId],
});
await boss.work(JOB_DELEGATION_REVOKE, { batchSize: 1 }, async (jobs) => {
const job = jobs?.[0];
try {
@@ -122,7 +132,9 @@ async function startPgBossJobs(fastify) {
const job = jobs?.[0];
const workerId = `pgboss:ocr:${process.pid}`;
try {
const stats = await runOcrOnce(fastify.prisma, fastify.config, fastify.log, workerId);
const stats = await runOcrOnce(fastify.prisma, fastify.config, fastify.log, workerId, {
onInvoiceProcessed: onOcrInvoiceProcessed,
});
fastify.metrics?.jobRunsTotal?.inc?.({
job: JOB_OCR_TICK,
driver: "pgboss",
@@ -164,6 +176,67 @@ async function startPgBossJobs(fastify) {
}
});
await boss.work(JOB_OCR_INVOICE, { batchSize: 1 }, async (jobs) => {
if (!["tencent", "mock"].includes(fastify.config.ocrProvider)) {
return;
}
const job = jobs?.[0];
const invoiceId = job?.data?.invoiceId;
if (!invoiceId) {
return;
}
const workerId = `pgboss:ocr-invoice:${process.pid}`;
try {
const stats = await runOcrForInvoiceId(
fastify.prisma,
fastify.config,
fastify.log,
invoiceId,
workerId,
{ onInvoiceProcessed: onOcrInvoiceProcessed },
);
fastify.metrics?.jobRunsTotal?.inc?.({
job: JOB_OCR_INVOICE,
driver: "pgboss",
outcome: "success",
});
if (stats.available > 0) {
fastify.metrics?.jobProcessedTotal?.inc?.(
{ job: JOB_OCR_INVOICE, driver: "pgboss", kind: "available" },
stats.available,
);
}
if (stats.duplicate > 0) {
fastify.metrics?.jobProcessedTotal?.inc?.(
{ job: JOB_OCR_INVOICE, driver: "pgboss", kind: "duplicate" },
stats.duplicate,
);
}
if (stats.failed > 0) {
fastify.metrics?.jobProcessedTotal?.inc?.(
{ job: JOB_OCR_INVOICE, driver: "pgboss", kind: "failed" },
stats.failed,
);
}
if (stats.claimed > 0) {
fastify.log.info(
{ jobId: job?.id, invoiceId, workerId, ...stats },
"pg-boss OCR invoice job processed",
);
}
} catch (error) {
fastify.metrics?.jobRunsTotal?.inc?.({
job: JOB_OCR_INVOICE,
driver: "pgboss",
outcome: "failure",
});
fastify.log.error({ err: error, jobId: job?.id, invoiceId }, "pg-boss OCR invoice job failed");
throw error;
}
});
await boss.work(JOB_INVOICE_VERIFY, { batchSize: 1 }, async (jobs) => {
const job = jobs?.[0];
const recordId = job?.data?.recordId;
@@ -301,7 +374,7 @@ async function startPgBossJobs(fastify) {
await boss.schedule(JOB_OCR_TICK, fastify.config.ocrPollCron, null, {
tz: fastify.config.appTimezone,
singletonKey: JOB_OCR_TICK,
singletonSeconds: 55,
singletonSeconds: 1,
});
} else {
await boss.unschedule(JOB_OCR_TICK).catch(() => {});
@@ -359,12 +432,27 @@ async function startPgBossJobs(fastify) {
return boss;
}
async function enqueueOcrInvoice(fastify, invoiceId) {
if (
fastify?.config?.jobDriver === "pgboss" &&
fastify.boss &&
["tencent", "mock"].includes(fastify.config.ocrProvider)
) {
await fastify.boss.send(JOB_OCR_INVOICE, { invoiceId });
return { mode: "queued" };
}
return { mode: "fallback" };
}
module.exports = {
JOB_OCR_INVOICE,
JOB_OCR_TICK,
JOB_DELEGATION_REVOKE,
JOB_INVOICE_ARCHIVE,
JOB_INVOICE_VERIFY,
JOB_INTEGRATION_OUTBOX,
JOB_ERP_MASTER_DATA_SYNC,
enqueueOcrInvoice,
startPgBossJobs,
};
+52
查看文件
@@ -0,0 +1,52 @@
const { canReceiveEvent, toClientEvent } = require("../services/events");
const HEARTBEAT_INTERVAL_MS = 25_000;
function writeSseEvent(raw, event) {
const clientEvent = toClientEvent(event);
if (!clientEvent) {
return;
}
raw.write(`id: ${clientEvent.id}\n`);
raw.write(`event: ${clientEvent.type}\n`);
raw.write(`data: ${JSON.stringify(clientEvent)}\n\n`);
}
async function eventsRoutes(fastify) {
fastify.get(
"/events",
{
preHandler: fastify.authenticate,
},
async (request, reply) => {
const canManageInvoice = fastify.hasPermission(request.employee, "invoice.manage");
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",
});
reply.raw.write(": connected\n\n");
const unsubscribe = fastify.appEvents.subscribe((event) => {
if (!canReceiveEvent(request.employee, event, { canManageInvoice })) {
return;
}
writeSseEvent(reply.raw, event);
});
const heartbeat = setInterval(() => {
reply.raw.write(": heartbeat\n\n");
}, HEARTBEAT_INTERVAL_MS);
request.raw.on("close", () => {
clearInterval(heartbeat);
unsubscribe();
});
},
);
}
module.exports = eventsRoutes;
+18
查看文件
@@ -10,6 +10,7 @@ const { sha256Hex } = require("../utils/crypto");
const { toMoneyString } = require("../utils/workflow");
const { findActiveDelegationForDelegate } = require("../utils/delegation");
const { getMissingRequiredOcrFields, isOcrEnabledForType } = require("../services/ocr");
const { enqueueOcrInvoice } = require("../jobs/pgBossJobs");
const { registerInvoiceVerificationRoutes } = require("./invoicePool/verification");
const { registerInvoicePoolReadRoutes } = require("./invoicePool/read");
const { normalizeInvoiceResponse } = require("./invoicePool/shared/mappers");
@@ -17,6 +18,7 @@ const { buildInvoiceDetailSelect } = require("./invoicePool/shared/selects");
const { writeBuffer } = require("../services/storage");
const {
mapInvoiceUnavailableReason,
resolveInvoiceAvailability,
resolveInvoiceDedupBlocker,
resolveUploadDedupDecision,
} = require("../services/invoiceGovernance");
@@ -509,6 +511,8 @@ async function invoicePoolRoutes(fastify) {
uploadAction: uploadDedupDecision.action,
unavailableReason: uploadDedupDecision.reason,
unavailableMessage: message,
availabilityReason: uploadDedupDecision.reason,
availabilityMessage: message,
});
}
@@ -570,10 +574,20 @@ async function invoicePoolRoutes(fastify) {
return created;
});
if (invoice.status === "pending_ocr") {
enqueueOcrInvoice(fastify, invoice.id).catch((error) => {
request.log.warn({ err: error, invoiceId: invoice.id }, "Failed to enqueue OCR invoice job");
});
}
const availability = resolveInvoiceAvailability(invoice);
return sendData(reply, request, {
invoicePoolId: invoice.id,
status: invoice.status,
fileSha256: invoice.file_sha256,
uploadAction: "create",
availabilityReason: availability.reason,
availabilityMessage: availability.message,
});
},
);
@@ -1231,6 +1245,10 @@ async function invoicePoolRoutes(fastify) {
return updated;
});
enqueueOcrInvoice(fastify, invoice.id).catch((error) => {
request.log.warn({ err: error, invoiceId: invoice.id }, "Failed to enqueue retry OCR invoice job");
});
return sendData(reply, request, normalizeInvoiceResponse(invoice));
},
);
+4 -43
查看文件
@@ -6,7 +6,7 @@ const { buildAuditLogData } = require("../../utils/audit");
const { signPayload, verifyToken } = require("../../utils/signedUrl");
const { openReadStream, StorageNotFoundError } = require("../../services/storage");
const { toMoneyString } = require("../../utils/workflow");
const { mapInvoiceUnavailableReason } = require("../../services/invoiceGovernance");
const { resolveInvoiceAvailability } = require("../../services/invoiceGovernance");
const {
mapInvoiceSummary,
normalizeInvoiceResponse,
@@ -106,48 +106,7 @@ function parseInvoiceStatusQuery(query) {
}
function resolveInvoiceStatusReason(invoice) {
if (invoice.status === "pending_ocr") {
return {
reason: "pending_ocr",
message: "发票正在识别,请稍后刷新",
};
}
if (invoice.status === "ocr_failed") {
const rawError = String(invoice.ocr_last_error ?? "");
if (rawError === "duplicate_invoice_code_number" || rawError.includes("duplicate_invoice")) {
return {
reason: "duplicate_invoice_code_number",
message: mapInvoiceUnavailableReason("used_by_company") ?? "发票重复,不能添加",
};
}
return {
reason: "ocr_failed",
message: rawError || "发票识别失败,请重新识别或手动处理",
};
}
if (invoice.status === "bound") {
return {
reason: "already_added",
message: "发票已添加到报销单",
};
}
if (invoice.status === "archived") {
return {
reason: invoice.archived_reason === "used_complete" ? "used_by_company" : "archived",
message:
invoice.archived_reason === "used_complete"
? "该发票已用于报销,不能重复添加"
: "发票已归档,不可添加",
};
}
return {
reason: "available",
message: "发票可添加",
};
return resolveInvoiceAvailability(invoice);
}
function ensureInvoiceReadableByUser(request, invoice) {
@@ -254,6 +213,8 @@ function registerInvoicePoolReadRoutes(fastify) {
status: invoice.status,
reason: statusReason.reason,
message: statusReason.message,
availabilityReason: statusReason.reason,
availabilityMessage: statusReason.message,
invoiceCode: invoice.invoice_code ?? null,
invoiceNumber: invoice.invoice_number ?? null,
invoiceDate: invoice.invoice_date?.toISOString?.()?.slice(0, 10) ?? null,
@@ -1,4 +1,5 @@
const { toMoneyString } = require("../../../utils/workflow");
const { resolveInvoiceAvailability } = require("../../../services/invoiceGovernance");
function mapVerificationRecord(item) {
if (!item) {
@@ -26,10 +27,13 @@ function mapInvoiceSummary(item) {
const latestVerification = Array.isArray(item.verification_records)
? item.verification_records[0] ?? null
: null;
const availability = resolveInvoiceAvailability(item);
return {
id: item.id,
status: item.status,
availabilityReason: availability.reason,
availabilityMessage: availability.message,
ocrType: item.ocr_type,
ocrProvider: item.ocr_provider,
ocrAttempts: item.ocr_attempts,
@@ -71,10 +75,13 @@ function normalizeInvoiceResponse(invoice) {
const latestVerification = Array.isArray(invoice.verification_records)
? invoice.verification_records[0] ?? null
: null;
const availability = resolveInvoiceAvailability(invoice);
return {
id: invoice.id,
status: invoice.status,
availabilityReason: availability.reason,
availabilityMessage: availability.message,
archivedReason: invoice.archived_reason ?? null,
archivedAt: invoice.archived_at?.toISOString?.() ?? null,
ownerEmployeeId: invoice.owner_employee_id,
@@ -28,6 +28,7 @@ function buildInvoiceSummarySelect() {
return {
id: true,
status: true,
archived_reason: true,
ocr_type: true,
ocr_provider: true,
ocr_attempts: true,
+104
查看文件
@@ -0,0 +1,104 @@
const { EventEmitter } = require("node:events");
function uniqueStrings(values) {
return Array.from(
new Set(
(Array.isArray(values) ? values : [])
.map((value) => (typeof value === "string" ? value.trim() : ""))
.filter(Boolean),
),
);
}
function createAppEventBus() {
const emitter = new EventEmitter();
emitter.setMaxListeners(500);
let sequence = 0;
return {
publish(event) {
const payload = {
id: `${Date.now()}-${++sequence}`,
createdAt: new Date().toISOString(),
...event,
audienceEmployeeIds: uniqueStrings(event?.audienceEmployeeIds),
};
emitter.emit("event", payload);
return payload;
},
subscribe(handler) {
emitter.on("event", handler);
return () => emitter.off("event", handler);
},
};
}
function buildInvoiceUpdatedEvent({
invoiceId,
status,
updatedAt = null,
audienceEmployeeIds = [],
}) {
return {
type: "invoice.updated",
invoiceId,
status,
updatedAt: updatedAt ?? new Date().toISOString(),
audienceEmployeeIds: uniqueStrings(audienceEmployeeIds),
};
}
function publishInvoiceUpdated(fastify, event) {
if (!fastify?.appEvents || typeof fastify.appEvents.publish !== "function") {
return null;
}
return fastify.appEvents.publish(buildInvoiceUpdatedEvent(event));
}
function canReceiveEvent(employee, event, options = {}) {
if (!employee) {
return false;
}
if (event?.type === "invoice.updated" && options.canManageInvoice) {
return true;
}
const audience = uniqueStrings(event?.audienceEmployeeIds);
if (!audience.length) {
return false;
}
return audience.includes(employee.id);
}
function toClientEvent(event) {
if (!event) {
return null;
}
if (event.type === "invoice.updated") {
return {
id: event.id,
type: event.type,
invoiceId: event.invoiceId,
status: event.status ?? null,
updatedAt: event.updatedAt ?? event.createdAt ?? null,
createdAt: event.createdAt ?? null,
};
}
return {
id: event.id,
type: event.type,
createdAt: event.createdAt ?? null,
};
}
module.exports = {
buildInvoiceUpdatedEvent,
canReceiveEvent,
createAppEventBus,
publishInvoiceUpdated,
toClientEvent,
};
+95
查看文件
@@ -8,6 +8,14 @@ const INVOICE_UNAVAILABLE_REASONS = {
usedByCompany: "used_by_company",
};
const OCR_FIELD_LABELS = {
invoiceCode: "发票代码",
invoiceNumber: "发票号码",
invoiceDate: "开票日期",
totalAmount: "总金额",
sellerName: "销售方",
};
function buildInvoiceDedupFilters({ fileSha256 = null, invoiceCode = null, invoiceNumber = null }) {
const filters = [];
if (fileSha256) {
@@ -182,13 +190,100 @@ function mapInvoiceUnavailableReason(reason) {
}
}
function parseMissingInvoiceOcrFields(rawError) {
if (typeof rawError !== "string") {
return [];
}
const matched = rawError.match(/^missing_required_fields:(.+)$/);
if (!matched?.[1]) {
return [];
}
return matched[1]
.split(",")
.map((field) => field.trim())
.filter(Boolean);
}
function formatInvoiceOcrFailureMessage(rawError) {
const missingFields = parseMissingInvoiceOcrFields(rawError);
if (missingFields.length) {
const labels = missingFields.map((field) => OCR_FIELD_LABELS[field] ?? field);
return `票据识别缺少关键字段:${labels.join("、")}。`;
}
if (rawError === "mock_ocr_failed") {
return "票据识别失败,请重试或人工补齐关键信息。";
}
if (!rawError) {
return "票据识别失败,请稍后重试。";
}
return "票据识别失败,请重新识别或手动处理。";
}
function resolveInvoiceAvailability(invoice) {
if (!invoice) {
return {
reason: "not_found",
message: "未找到发票",
};
}
if (invoice.status === "pending_ocr") {
return {
reason: "pending_ocr",
message: "发票正在识别,完成后会自动刷新",
};
}
if (invoice.status === "ocr_failed") {
const rawError = String(invoice.ocr_last_error ?? "");
if (rawError === "duplicate_invoice_code_number" || rawError.includes("duplicate_invoice")) {
return {
reason: "duplicate_invoice",
message: "已识别出相同发票,不能重复添加",
};
}
return {
reason: "ocr_failed",
message: formatInvoiceOcrFailureMessage(rawError),
};
}
if (isUsedInvoice(invoice)) {
return {
reason: INVOICE_UNAVAILABLE_REASONS.usedByCompany,
message: mapInvoiceUnavailableReason(INVOICE_UNAVAILABLE_REASONS.usedByCompany),
};
}
if (invoice.status === "archived") {
return {
reason: "archived",
message: "发票已归档,不可添加",
};
}
return {
reason: "available",
message: "发票可添加",
};
}
module.exports = {
INVOICE_UNAVAILABLE_REASONS,
buildInvoiceDedupFilters,
classifyInvoiceConflict,
findInvoiceDedupConflicts,
formatInvoiceOcrFailureMessage,
isUsedInvoice,
mapInvoiceUnavailableReason,
parseMissingInvoiceOcrFields,
resolveInvoiceAvailability,
resolveInvoiceDedupBlocker,
resolveUploadDedupDecision,
};
@@ -1,6 +1,8 @@
const {
INVOICE_UNAVAILABLE_REASONS,
classifyInvoiceConflict,
formatInvoiceOcrFailureMessage,
resolveInvoiceAvailability,
resolveUploadDedupDecision,
} = require("../../src/services/invoiceGovernance");
@@ -17,6 +19,7 @@ function buildInvoice(overrides = {}) {
file_sha256: overrides.file_sha256 ?? "sha256-a",
invoice_code: overrides.invoice_code ?? null,
invoice_number: overrides.invoice_number ?? null,
ocr_last_error: overrides.ocr_last_error ?? null,
};
}
@@ -90,4 +93,34 @@ describe("invoiceGovernance dedup", () => {
expect(decision.action).toBe("blocked");
expect(decision.reason).toBe(INVOICE_UNAVAILABLE_REASONS.duplicateOtherOwner);
});
it("formats invoice availability without leaking raw OCR error codes", () => {
expect(
resolveInvoiceAvailability(
buildInvoice({
status: "ocr_failed",
ocr_last_error: "missing_required_fields:invoiceCode,sellerName",
}),
),
).toEqual({
reason: "ocr_failed",
message: "票据识别缺少关键字段:发票代码、销售方。",
});
expect(
resolveInvoiceAvailability(
buildInvoice({
status: "ocr_failed",
ocr_last_error: "duplicate_invoice_code_number",
}),
),
).toEqual({
reason: "duplicate_invoice",
message: "已识别出相同发票,不能重复添加",
});
expect(formatInvoiceOcrFailureMessage("mock_ocr_failed")).toBe(
"票据识别失败,请重试或人工补齐关键信息。",
);
});
});
+42 -1
查看文件
@@ -2,12 +2,14 @@ const fs = require("node:fs/promises");
const os = require("node:os");
const path = require("node:path");
const { runOcrOnce } = require("../../../src/jobs/ocrJob");
const { runOcrForInvoiceId, runOcrOnce } = require("../../../src/jobs/ocrJob");
function buildClaimedInvoice(overrides = {}) {
return {
id: "11111111-1111-4111-8111-111111111111",
file_sha256: "sha256-1",
owner_employee_id: "employee-owner-1",
uploaded_by_employee_id: "employee-upload-1",
ocr_provider: "mock",
ocr_type: "vat",
storage_path: "20260428/invoice.txt",
@@ -91,6 +93,45 @@ describe("ocrJob quality gate", () => {
);
});
it("processes a directed invoice and emits processed metadata", async () => {
await fs.writeFile(
path.join(tempDir, "20260428/invoice.txt"),
"MOCK_DIRECTED_SUCCESS_R73",
"utf8",
);
const prisma = buildPrisma([buildClaimedInvoice()]);
const logger = {
warn: vi.fn(),
error: vi.fn(),
};
const onInvoiceProcessed = vi.fn();
const result = await runOcrForInvoiceId(
prisma,
{
storageDir: tempDir,
ocrBatchSize: 10,
ocrLockTtlSeconds: 300,
ocrMaxAttempts: 3,
ocrProvider: "mock",
},
logger,
"11111111-1111-4111-8111-111111111111",
"ocr-directed-worker",
{ onInvoiceProcessed },
);
expect(result).toEqual({ claimed: 1, available: 1, duplicate: 0, failed: 0 });
expect(onInvoiceProcessed).toHaveBeenCalledWith({
invoiceId: "11111111-1111-4111-8111-111111111111",
status: "available",
outcome: "available",
ownerEmployeeId: "employee-owner-1",
uploadedByEmployeeId: "employee-upload-1",
});
});
it("marks invoice ocr_failed when required VAT fields are missing", async () => {
await fs.writeFile(
path.join(tempDir, "20260428/invoice.txt"),
+56
查看文件
@@ -0,0 +1,56 @@
const { enqueueOcrInvoice, JOB_OCR_INVOICE } = require("../../../src/jobs/pgBossJobs");
describe("pgBossJobs enqueueOcrInvoice", () => {
it("queues a directed OCR invoice job when pg-boss is available", async () => {
const send = vi.fn(async () => ({ id: "job-1" }));
const result = await enqueueOcrInvoice(
{
config: {
jobDriver: "pgboss",
ocrProvider: "mock",
},
boss: { send },
},
"11111111-1111-4111-8111-111111111111",
);
expect(result).toEqual({ mode: "queued" });
expect(send).toHaveBeenCalledWith(JOB_OCR_INVOICE, {
invoiceId: "11111111-1111-4111-8111-111111111111",
});
});
it("falls back cleanly when pg-boss is unavailable", async () => {
await expect(
enqueueOcrInvoice(
{
config: {
jobDriver: "pgboss",
ocrProvider: "mock",
},
boss: null,
},
"11111111-1111-4111-8111-111111111111",
),
).resolves.toEqual({ mode: "fallback" });
});
it("does not queue OCR when the OCR provider is disabled", async () => {
const send = vi.fn();
const result = await enqueueOcrInvoice(
{
config: {
jobDriver: "pgboss",
ocrProvider: "off",
},
boss: { send },
},
"11111111-1111-4111-8111-111111111111",
);
expect(result).toEqual({ mode: "fallback" });
expect(send).not.toHaveBeenCalled();
});
});
+52
查看文件
@@ -0,0 +1,52 @@
const Fastify = require("fastify");
const eventsRoutes = require("../../../src/routes/events");
const { createAppEventBus } = require("../../../src/services/events");
const { AppError } = require("../../../src/utils/AppError");
async function buildApp() {
const app = Fastify();
app.decorate("appEvents", createAppEventBus());
app.decorate("authenticate", async (request) => {
const auth = request.headers.authorization ?? "";
if (!auth.startsWith("Bearer ")) {
throw new AppError({ code: "AUTH_001", message: "未登录", statusCode: 401 });
}
request.employee = { id: "employee-1", permissionCodes: [] };
});
app.decorate("hasPermission", (employee, permissionCode) =>
(employee?.permissionCodes ?? []).includes(permissionCode),
);
app.setErrorHandler((error, request, reply) => {
if (error instanceof AppError) {
return reply.status(error.statusCode).send({
error: {
code: error.code,
message: error.message,
details: error.details ?? {},
},
});
}
return reply.status(500).send({ error: { code: "SYS_001", message: "服务异常" } });
});
await app.register(eventsRoutes, { prefix: "/api" });
return app;
}
describe("GET /api/events", () => {
it("requires bearer authentication before opening the event stream", async () => {
const app = await buildApp();
try {
const response = await app.inject({
method: "GET",
url: "/api/events",
});
expect(response.statusCode).toBe(401);
expect(response.json().error.code).toBe("AUTH_001");
} finally {
await app.close();
}
});
});
@@ -38,6 +38,8 @@ function buildMultipartRequest({ filename, contentType, content, fields = [] })
async function buildApp({
ocrProvider = "off",
jobDriver = "cron",
bossSend = null,
invoiceConflicts = [],
statusRows = [],
permissionCodes = ["invoice.upload"],
@@ -75,7 +77,11 @@ async function buildApp({
maxUploadSizeBytes: 5 * 1024 * 1024,
invoiceArchiveUnusedDays: 30,
ocrProvider,
jobDriver,
});
if (bossSend) {
app.decorate("boss", { send: bossSend });
}
app.decorate("authenticate", async (request) => {
if (!request.headers.authorization) {
throw new AppError({ code: "AUTH_001", message: "未登录", statusCode: 401 });
@@ -177,8 +183,13 @@ describe("invoicePool upload route", () => {
}
});
it("puts VAT uploads into pending_ocr when OCR provider is enabled", async () => {
const { app, prisma, tempDir } = await buildApp({ ocrProvider: "mock" });
it("puts VAT uploads into pending_ocr and enqueues a directed OCR job when OCR provider is enabled", async () => {
const bossSend = vi.fn(async () => ({ id: "ocr-job-1" }));
const { app, prisma, tempDir } = await buildApp({
ocrProvider: "mock",
jobDriver: "pgboss",
bossSend,
});
try {
const requestPayload = buildMultipartRequest({
@@ -214,6 +225,9 @@ describe("invoicePool upload route", () => {
}),
}),
);
expect(bossSend).toHaveBeenCalledWith("ocr.invoice", {
invoiceId: "11111111-1111-4111-8111-111111111111",
});
} finally {
await app.close();
await fs.rm(tempDir, { recursive: true, force: true });
@@ -133,6 +133,7 @@ async function buildApp(options = {}) {
],
travelRequests: (options.travelRequests ?? [buildTravelRequest()]).map((item) => cloneTravelRequest(item)),
auditLogs: [],
sequenceNumber: 0,
};
function buildReimbursementSelect(record, select) {
@@ -236,6 +237,28 @@ async function buildApp(options = {}) {
return found ? cloneTravelRequest(found) : null;
}),
},
documentSequence: {
findUnique: vi.fn(async () => null),
upsert: vi.fn(async ({ create, update }) => {
stateRef.sequenceNumber = (stateRef.sequenceNumber ?? 0) + 1;
return {
document_type: create.document_type,
prefix: update.prefix ?? create.prefix,
date_format: update.date_format ?? create.date_format,
separator: update.separator ?? create.separator,
number_length: update.number_length ?? create.number_length,
reset_period: update.reset_period ?? create.reset_period,
current_number: stateRef.sequenceNumber,
last_reset_date: null,
updated_at: new Date("2026-04-08T12:00:00.000Z"),
updated_by: update.updated_by ?? create.updated_by ?? null,
};
}),
},
documentCategory: {
findUnique: vi.fn(async () => null),
findFirst: vi.fn(async () => null),
},
reimbursementForm: {
create: vi.fn(async ({ data, select }) => {
const record = {
@@ -317,6 +340,7 @@ async function buildApp(options = {}) {
expenseTypes: state.expenseTypes.map((item) => ({ ...item })),
travelRequests: state.travelRequests.map((item) => cloneTravelRequest(item)),
auditLogs: state.auditLogs.map((item) => ({ ...item })),
sequenceNumber: state.sequenceNumber,
};
const tx = buildPrismaLike(txState);
const result = await callback(tx);
@@ -327,6 +351,7 @@ async function buildApp(options = {}) {
state.expenseTypes = txState.expenseTypes;
state.travelRequests = txState.travelRequests;
state.auditLogs = txState.auditLogs;
state.sequenceNumber = txState.sequenceNumber;
return result;
});
@@ -44,6 +44,7 @@ function buildReimbursementRecord(overrides = {}) {
},
},
],
attachments: [],
invoice_pool_items: [
{
id: "invoice-1",
+67
查看文件
@@ -0,0 +1,67 @@
const {
buildInvoiceUpdatedEvent,
canReceiveEvent,
createAppEventBus,
toClientEvent,
} = require("../../../src/services/events");
describe("app events", () => {
it("publishes minimal invoice updated events", () => {
const event = buildInvoiceUpdatedEvent({
invoiceId: "11111111-1111-4111-8111-111111111111",
status: "available",
audienceEmployeeIds: ["employee-1", "employee-1", ""],
updatedAt: "2026-05-09T10:00:00.000Z",
});
expect(event).toEqual({
type: "invoice.updated",
invoiceId: "11111111-1111-4111-8111-111111111111",
status: "available",
updatedAt: "2026-05-09T10:00:00.000Z",
audienceEmployeeIds: ["employee-1"],
});
expect(toClientEvent({ id: "evt-1", createdAt: "2026-05-09T10:00:01.000Z", ...event })).toEqual({
id: "evt-1",
type: "invoice.updated",
invoiceId: "11111111-1111-4111-8111-111111111111",
status: "available",
updatedAt: "2026-05-09T10:00:00.000Z",
createdAt: "2026-05-09T10:00:01.000Z",
});
});
it("filters invoice events by employee audience unless invoice manager", () => {
const event = buildInvoiceUpdatedEvent({
invoiceId: "11111111-1111-4111-8111-111111111111",
status: "available",
audienceEmployeeIds: ["employee-1"],
});
expect(canReceiveEvent({ id: "employee-1" }, event)).toBe(true);
expect(canReceiveEvent({ id: "employee-2" }, event)).toBe(false);
expect(canReceiveEvent({ id: "employee-2" }, event, { canManageInvoice: true })).toBe(true);
});
it("supports subscribe and unsubscribe on the in-process event bus", () => {
const bus = createAppEventBus();
const handler = vi.fn();
const unsubscribe = bus.subscribe(handler);
bus.publish({ type: "invoice.updated", invoiceId: "invoice-1", audienceEmployeeIds: ["employee-1"] });
unsubscribe();
bus.publish({ type: "invoice.updated", invoiceId: "invoice-2", audienceEmployeeIds: ["employee-1"] });
expect(handler).toHaveBeenCalledTimes(1);
expect(handler.mock.calls[0][0]).toEqual(
expect.objectContaining({
id: expect.any(String),
createdAt: expect.any(String),
type: "invoice.updated",
invoiceId: "invoice-1",
audienceEmployeeIds: ["employee-1"],
}),
);
});
});