romfast/roa2web-service-auto
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: multi-Oracle server support with runtime switching

Browse Source 
Complete implementation of multi-server Oracle database support:

Backend:
- Multi-pool Oracle with lazy loading per server
- Email-to-server cache for automatic server discovery
- JWT tokens include server_id claim
- /auth/check-identity and /auth/check-email endpoints
- /auth/my-servers endpoint for listing user's accessible servers
- Server switch with password re-authentication

Frontend:
- New ServerSelector component for header dropdown
- Multi-step login flow (identity → server → password)
- Server switching from header with password modal
- Mobile drawer menu with server selection
- Dark mode support for all new components
- URL bookmark support with ?server= query param

Scripts:
- Unified start.sh replacing start-prod.sh/start-test.sh
- Unified ssh-tunnel.sh with multi-server support
- Updated status.sh for new architecture

Tests:
- E2E tests for multi-server and single-server login flows
- Backend unit tests for all new endpoints
- Oracle multi-pool integration tests

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Claude Agent
2026-01-26 22:39:06 +00:00
parent 5f99ee2fd0
commit b137e80b71
102 changed files with 9398 additions and 2787 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
backend/modules/data_entry/services/sync_service.py
View File

@@ -19,24 +19,30 @@ from backend.modules.data_entry.db.models.nomenclature import SyncedSupplier, Lo
logger = logging.getLogger(__name__)
# Cache for schema lookups (populated dynamically from Oracle)
_schema_cache: dict[int, str] = {}
# Key format: (server_id, company_id) for multi-server support
_schema_cache: dict[tuple, str] = {}
class SyncService:
"""Service for syncing nomenclatures from Oracle."""
@staticmethod
async def get_schema_for_company(company_id: int) -> Optional[str]:
async def get_schema_for_company(company_id: int, server_id: Optional[str] = None) -> Optional[str]:
"""
Get Oracle schema for company ID from V_NOM_FIRME view.
Results are cached in memory for performance.
Args:
company_id: The company ID to look up
server_id: Optional Oracle server ID for multi-server mode
"""
# Check cache first
if company_id in _schema_cache:
return _schema_cache[company_id]
# Check cache first - use (server_id, company_id) as key for multi-server support
cache_key = (server_id, company_id)
if cache_key in _schema_cache:
return _schema_cache[cache_key]
try:
async with oracle_pool.get_connection() as connection:
async with oracle_pool.get_connection(server_id) as connection:
with connection.cursor() as cursor:
cursor.execute("""
SELECT SCHEMA
@@ -47,34 +53,39 @@ class SyncService:
if result:
schema = result[0]
_schema_cache[company_id] = schema
logger.info(f"Resolved schema for company {company_id}: {schema}")
_schema_cache[cache_key] = schema
logger.info(f"Resolved schema for company {company_id} on server {server_id}: {schema}")
return schema
else:
logger.warning(f"No schema found for company {company_id}")
logger.warning(f"No schema found for company {company_id} on server {server_id}")
return None
except Exception as e:
logger.error(f"Error fetching schema for company {company_id}: {e}")
logger.error(f"Error fetching schema for company {company_id} on server {server_id}: {e}")
return None
@staticmethod
async def sync_suppliers(session: AsyncSession, company_id: int) -> Tuple[int, int]:
async def sync_suppliers(session: AsyncSession, company_id: int, server_id: Optional[str] = None) -> Tuple[int, int]:
"""
Sync suppliers (furnizori, id_tip_part=17) from Oracle to SQLite.
Uses CORESP_TIP_PART joined with VNOM_PARTENERI view.
Returns (synced_count, error_count).
Args:
session: SQLAlchemy async session for SQLite
company_id: The company ID to sync suppliers for
server_id: Optional Oracle server ID for multi-server mode
"""
schema = await SyncService.get_schema_for_company(company_id)
schema = await SyncService.get_schema_for_company(company_id, server_id)
if not schema:
logger.warning(f"No schema mapping for company {company_id}")
logger.warning(f"No schema mapping for company {company_id} on server {server_id}")
return 0, 0
synced = 0
errors = 0
try:
async with oracle_pool.get_connection() as connection:
async with oracle_pool.get_connection(server_id) as connection:
with connection.cursor() as cursor:
# Fetch active suppliers from Oracle
# id_tip_part = 17 means "furnizori" (suppliers)
@@ -139,7 +150,7 @@ class SyncService:
return synced, errors
@staticmethod
async def sync_cash_registers(session: AsyncSession, company_id: int) -> Tuple[int, int]:
async def sync_cash_registers(session: AsyncSession, company_id: int, server_id: Optional[str] = None) -> Tuple[int, int]:
"""
Sync cash registers and bank accounts from Oracle to SQLite.
Returns (synced_count, error_count).
@@ -149,10 +160,15 @@ class SyncService:
- id_tip_part = 23: CASA VALUTA
- id_tip_part = 24: BANCA LEI
- id_tip_part = 25: BANCA VALUTA
Args:
session: SQLAlchemy async session for SQLite
company_id: The company ID to sync cash registers for
server_id: Optional Oracle server ID for multi-server mode
"""
schema = await SyncService.get_schema_for_company(company_id)
schema = await SyncService.get_schema_for_company(company_id, server_id)
if not schema:
logger.warning(f"No schema mapping for company {company_id}")
logger.warning(f"No schema mapping for company {company_id} on server {server_id}")
return 0, 0
synced = 0
@@ -164,7 +180,7 @@ class SyncService:
partner_types = [22, 23, 24, 25]
try:
async with oracle_pool.get_connection() as connection:
async with oracle_pool.get_connection(server_id) as connection:
with connection.cursor() as cursor:
# Fetch cash/bank partners from CORESP_TIP_PART
cursor.execute(f"""