hyb-oa/server/scripts/regression-dedup.sh

#!/usr/bin/env bash
set -euo pipefail

BASE_URL="${BASE_URL:-http://127.0.0.1:3000}"
DEVICE_FINGERPRINT="${DEVICE_FINGERPRINT:-regression-dedup}"
EMPLOYEE_IDENTIFIER="${EMPLOYEE_IDENTIFIER:-employee@hyinda.com}"
EMPLOYEE_PASSWORD="${EMPLOYEE_PASSWORD:-EmployeePassword1234}"

if ! command -v curl >/dev/null 2>&1; then
  echo "[ERROR] curl not found" >&2
  exit 1
fi

if ! command -v jq >/dev/null 2>&1; then
  echo "[ERROR] jq not found" >&2
  exit 1
fi

if ! command -v node >/dev/null 2>&1; then
  echo "[ERROR] node not found" >&2
  exit 1
fi

ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
cd "$ROOT_DIR"

if [[ -z "${DATABASE_URL:-}" && -f ".env" ]]; then
  set -a
  # shellcheck disable=SC1091
  source .env
  set +a
fi

if [[ -z "${DATABASE_URL:-}" ]]; then
  echo "[ERROR] DATABASE_URL is required (export it or provide server/.env)" >&2
  exit 1
fi

TMP_DIR="$(mktemp -d)"
TMP_INDEX_CREATED=0
RUN_SUFFIX="$(date +%s)-$RANDOM"
DEDUP_CODE="RCODE-$RUN_SUFFIX"
DEDUP_NUMBER="RNO-$RUN_SUFFIX"

cleanup() {
  if [[ "$TMP_INDEX_CREATED" == "1" ]]; then
    drop_tmp_non_dedup_index || true
  fi
  rm -rf "$TMP_DIR"
}
trap cleanup EXIT

assert_status() {
  local actual="$1"
  local expected="$2"
  local name="$3"
  local body_file="${4:-}"

  if [[ "$actual" != "$expected" ]]; then
    echo "[ERROR] $name expected HTTP $expected but got $actual" >&2
    if [[ -n "$body_file" && -f "$body_file" ]]; then
      cat "$body_file" >&2
    fi
    exit 1
  fi
}

api_json() {
  local method="$1"
  local url="$2"
  local out_file="$3"
  local body="${4:-}"

  if [[ -n "$body" ]]; then
    curl -sS -o "$out_file" -w "%{http_code}" \
      -X "$method" "$url" \
      -H "authorization: Bearer $TOKEN" \
      -H "content-type: application/json" \
      -d "$body"
  else
    curl -sS -o "$out_file" -w "%{http_code}" \
      -X "$method" "$url" \
      -H "authorization: Bearer $TOKEN"
  fi
}

upload_file() {
  local file_path="$1"
  local out_file="$2"

  curl -sS -o "$out_file" -w "%{http_code}" \
    -X POST "$BASE_URL/api/invoice-pool/upload" \
    -H "authorization: Bearer $TOKEN" \
    -F "file=@$file_path"
}

create_reimbursement() {
  local out_file="$1"
  local status
  status="$(api_json POST "$BASE_URL/api/reimbursements" "$out_file" '{}')"
  assert_status "$status" "200" "create reimbursement" "$out_file"
  jq -er '.data.id' "$out_file"
}

set_invoice_code_number() {
  local invoice_id="$1"
  local invoice_code="$2"
  local invoice_number="$3"

  DATABASE_URL="$DATABASE_URL" node - "$invoice_id" "$invoice_code" "$invoice_number" <<'NODE'
const invoiceId = process.argv[2];
const invoiceCode = process.argv[3];
const invoiceNumber = process.argv[4];
const { createPrismaClient } = require('./src/prismaClient');

(async () => {
  const { prisma, pool } = createPrismaClient();
  await prisma.$connect();
  try {
    await prisma.invoicePool.update({
      where: { id: invoiceId },
      data: {
        invoice_code: invoiceCode,
        invoice_number: invoiceNumber,
      },
    });
  } finally {
    await prisma.$disconnect();
    await pool.end();
  }
})().catch((error) => {
  console.error(error);
  process.exit(1);
});
NODE
}

create_tmp_non_dedup_index() {
  local reimbursement_id="$1"
  DATABASE_URL="$DATABASE_URL" REIMBURSEMENT_ID="$reimbursement_id" node <<'NODE'
const { createPrismaClient } = require('./src/prismaClient');

(async () => {
  const reimbursementId = process.env.REIMBURSEMENT_ID;
  if (!reimbursementId) {
    throw new Error('REIMBURSEMENT_ID is required');
  }

  const { prisma, pool } = createPrismaClient();
  await prisma.$connect();
  try {
    await prisma.$executeRawUnsafe('DROP INDEX IF EXISTS "tmp_unique_linked_reim"');
    await prisma.$executeRawUnsafe(
      `CREATE UNIQUE INDEX "tmp_unique_linked_reim" ON "invoice_pool" ("linked_reimbursement_id") WHERE "linked_reimbursement_id" = '${reimbursementId}'`,
    );
  } finally {
    await prisma.$disconnect();
    await pool.end();
  }
})().catch((error) => {
  console.error(error);
  process.exit(1);
});
NODE
}

drop_tmp_non_dedup_index() {
  DATABASE_URL="$DATABASE_URL" node <<'NODE'
const { createPrismaClient } = require('./src/prismaClient');

(async () => {
  const { prisma, pool } = createPrismaClient();
  await prisma.$connect();
  try {
    await prisma.$executeRawUnsafe('DROP INDEX IF EXISTS "tmp_unique_linked_reim"');
  } finally {
    await prisma.$disconnect();
    await pool.end();
  }
})().catch((error) => {
  console.error(error);
  process.exit(1);
});
NODE
}

assert_error_code() {
  local file_path="$1"
  local expected="$2"
  local actual
  actual="$(jq -r '.error.code // empty' "$file_path")"
  if [[ "$actual" != "$expected" ]]; then
    echo "[ERROR] expected error code $expected but got $actual" >&2
    cat "$file_path" >&2
    exit 1
  fi
}

echo "[1/7] Login employee"
login_code="$(curl -sS -o "$TMP_DIR/login.json" -w "%{http_code}" \
  -X POST "$BASE_URL/api/auth/login" \
  -H "content-type: application/json" \
  -H "x-device-fingerprint: $DEVICE_FINGERPRINT" \
  -d "{\"identifier\":\"$EMPLOYEE_IDENTIFIER\",\"password\":\"$EMPLOYEE_PASSWORD\"}")"
assert_status "$login_code" "200" "login" "$TMP_DIR/login.json"
TOKEN="$(jq -er '.data.accessToken' "$TMP_DIR/login.json")"

echo "[2/7] Dedup by file_sha256 => VALID_003"
cat > "$TMP_DIR/same-file.pdf" <<PDF
%PDF-1.4
same-content-$RUN_SUFFIX
PDF

RID_A="$(create_reimbursement "$TMP_DIR/rA.json")"
status="$(upload_file "$TMP_DIR/same-file.pdf" "$TMP_DIR/uA.json")"
assert_status "$status" "200" "upload first same-file" "$TMP_DIR/uA.json"
IID_A="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uA.json")"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_A/bind" "$TMP_DIR/bA.json" "{\"reimbursementId\":\"$RID_A\"}")"
assert_status "$status" "200" "bind first same-file" "$TMP_DIR/bA.json"

RID_B="$(create_reimbursement "$TMP_DIR/rB.json")"
status="$(upload_file "$TMP_DIR/same-file.pdf" "$TMP_DIR/uB.json")"
assert_status "$status" "200" "upload second same-file" "$TMP_DIR/uB.json"
IID_B="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uB.json")"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_B/bind" "$TMP_DIR/bB.json" "{\"reimbursementId\":\"$RID_B\"}")"
assert_status "$status" "409" "file_sha256 dedup bind" "$TMP_DIR/bB.json"
assert_error_code "$TMP_DIR/bB.json" "VALID_003"

echo "[3/7] Dedup by invoice_code+invoice_number => VALID_003"
cat > "$TMP_DIR/code-1.pdf" <<PDF
%PDF-1.4
code-one-$RUN_SUFFIX
PDF
cat > "$TMP_DIR/code-2.pdf" <<PDF
%PDF-1.4
code-two-$RUN_SUFFIX
PDF

RID_C="$(create_reimbursement "$TMP_DIR/rC.json")"
status="$(upload_file "$TMP_DIR/code-1.pdf" "$TMP_DIR/uC.json")"
assert_status "$status" "200" "upload code-1" "$TMP_DIR/uC.json"
IID_C="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uC.json")"
set_invoice_code_number "$IID_C" "$DEDUP_CODE" "$DEDUP_NUMBER"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_C/bind" "$TMP_DIR/bC.json" "{\"reimbursementId\":\"$RID_C\"}")"
assert_status "$status" "200" "bind code-1" "$TMP_DIR/bC.json"

RID_D="$(create_reimbursement "$TMP_DIR/rD.json")"
status="$(upload_file "$TMP_DIR/code-2.pdf" "$TMP_DIR/uD.json")"
assert_status "$status" "200" "upload code-2" "$TMP_DIR/uD.json"
IID_D="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uD.json")"
set_invoice_code_number "$IID_D" "$DEDUP_CODE" "$DEDUP_NUMBER"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_D/bind" "$TMP_DIR/bD.json" "{\"reimbursementId\":\"$RID_D\"}")"
assert_status "$status" "409" "code+number dedup bind" "$TMP_DIR/bD.json"
assert_error_code "$TMP_DIR/bD.json" "VALID_003"

echo "[4/7] Non-dedup P2002 should not map to VALID_003"
cat > "$TMP_DIR/non-dedup-1.pdf" <<PDF
%PDF-1.4
non-dedup-1-$RUN_SUFFIX
PDF
cat > "$TMP_DIR/non-dedup-2.pdf" <<PDF
%PDF-1.4
non-dedup-2-$RUN_SUFFIX
PDF

RID_E="$(create_reimbursement "$TMP_DIR/rE.json")"
create_tmp_non_dedup_index "$RID_E"
TMP_INDEX_CREATED=1
status="$(upload_file "$TMP_DIR/non-dedup-1.pdf" "$TMP_DIR/uE1.json")"
assert_status "$status" "200" "upload non-dedup-1" "$TMP_DIR/uE1.json"
IID_E1="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uE1.json")"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_E1/bind" "$TMP_DIR/bE1.json" "{\"reimbursementId\":\"$RID_E\"}")"
assert_status "$status" "200" "bind non-dedup-1" "$TMP_DIR/bE1.json"

status="$(upload_file "$TMP_DIR/non-dedup-2.pdf" "$TMP_DIR/uE2.json")"
assert_status "$status" "200" "upload non-dedup-2" "$TMP_DIR/uE2.json"
IID_E2="$(jq -er '.data.invoicePoolId' "$TMP_DIR/uE2.json")"
status="$(api_json POST "$BASE_URL/api/invoice-pool/$IID_E2/bind" "$TMP_DIR/bE2.json" "{\"reimbursementId\":\"$RID_E\"}")"
if [[ "$status" == "409" ]] && [[ "$(jq -r '.error.code // empty' "$TMP_DIR/bE2.json")" == "VALID_003" ]]; then
  echo "[ERROR] non-dedup P2002 was incorrectly mapped to VALID_003" >&2
  cat "$TMP_DIR/bE2.json" >&2
  exit 1
fi

echo "[5/7] Validate non-dedup result is server-side failure path"
if [[ "$status" != "500" ]]; then
  echo "[ERROR] expected HTTP 500 for temporary non-dedup unique index violation, got $status" >&2
  cat "$TMP_DIR/bE2.json" >&2
  exit 1
fi

echo "[6/7] Ensure non-dedup response is not VALID_003"
unexpected_code="$(jq -r '.error.code // empty' "$TMP_DIR/bE2.json")"
if [[ "$unexpected_code" == "VALID_003" ]]; then
  echo "[ERROR] unexpected VALID_003 for non-dedup conflict" >&2
  cat "$TMP_DIR/bE2.json" >&2
  exit 1
fi

drop_tmp_non_dedup_index
TMP_INDEX_CREATED=0

echo "[7/7] Done"
echo "[OK] Dedup regression passed"