Watch
1
0
Fork
You've already forked arc
0
forked from hesabix/arc
arc/hesabixAPI/app/services/ai/ai_function_extensions_phase5.py

357 lines
14 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""
ثبت functionهای فاز ۵ AI — writeهای تکمیلی، CRM، workflow، export.
"""
from __future__ import annotations
import base64
from datetime import date, datetime
from typing import Any, Dict, List, TYPE_CHECKING
from app.services.ai.function_registry import AIRole, AIFunction
if TYPE_CHECKING:
from app.services.ai.function_registry import AIFunctionRegistry
_EXPORT_TYPES = [
"persons",
"invoices",
"products",
"expense_income",
"documents",
"checks",
]
_MAX_EXPORT_ROWS = 10_000
def register_phase5_business_functions(registry: "AIFunctionRegistry") -> None:
from app.services.ai.ai_tool_payloads import (
CREATE_EXPENSE_INCOME_DESCRIPTION,
CREATE_EXPENSE_INCOME_PARAMETERS_SCHEMA,
build_create_expense_income_payload,
)
# --- expense / income ---
def create_expense_income_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from app.services.expense_income_service import create_expense_income
db = context["db"]
business_id = int(args.get("business_id") or context.get("business_id"))
user_id = context["user_context"].get_user_id()
body = build_create_expense_income_payload(
args, db=db, business_id=business_id
)
return create_expense_income(db, business_id, user_id, body)
registry.register(
AIFunction(
name="create_expense_income",
description=CREATE_EXPENSE_INCOME_DESCRIPTION,
parameters_schema=CREATE_EXPENSE_INCOME_PARAMETERS_SCHEMA,
handler=create_expense_income_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["expenses_income.write"],
category="financial",
requires_approval=True,
risk_level="high",
is_readonly=False,
)
)
# --- invoice update / delete ---
def update_invoice_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from app.services.ai.ai_tool_payloads import build_update_invoice_payload
from app.services.invoice_service import update_invoice
db = context["db"]
user_id = context["user_context"].get_user_id()
invoice_id = int(args["invoice_id"])
data = build_update_invoice_payload(args, db=db, invoice_id=invoice_id)
return update_invoice(db, invoice_id, user_id, data)
from app.services.ai.ai_tool_payloads import (
UPDATE_INVOICE_DESCRIPTION,
UPDATE_INVOICE_PARAMETERS_SCHEMA,
)
registry.register(
AIFunction(
name="update_invoice",
description=UPDATE_INVOICE_DESCRIPTION,
parameters_schema=UPDATE_INVOICE_PARAMETERS_SCHEMA,
handler=update_invoice_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["invoices.write"],
category="invoices",
requires_approval=True,
risk_level="high",
is_readonly=False,
)
)
def delete_invoice_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from app.services.invoice_service import delete_invoice
db = context["db"]
invoice_id = int(args["invoice_id"])
ok = delete_invoice(db, invoice_id)
return {"deleted": ok, "invoice_id": invoice_id}
registry.register(
AIFunction(
name="delete_invoice",
description="حذف فاکتور (سند). عملیات برگشت‌ناپذیر — نیاز به تأیید.",
parameters_schema={
"type": "object",
"properties": {"invoice_id": {"type": "integer"}},
"required": ["invoice_id"],
},
handler=delete_invoice_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["invoices.write"],
category="invoices",
requires_approval=True,
risk_level="high",
is_readonly=False,
)
)
# --- CRM lead ---
def create_lead_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from datetime import date as date_cls
from adapters.db.models.crm import CrmProcessDefinition, CrmProcessStage, Lead
from app.services.document_numbering_service import generate_document_code
db = context["db"]
business_id = int(args.get("business_id") or context.get("business_id"))
user_id = context["user_context"].get_user_id()
process_id = int(args["process_definition_id"])
stage_id = int(args["stage_id"])
proc = (
db.query(CrmProcessDefinition)
.filter(
CrmProcessDefinition.id == process_id,
CrmProcessDefinition.business_id == business_id,
)
.first()
)
if not proc:
raise ValueError("فرایند CRM یافت نشد")
stage = (
db.query(CrmProcessStage)
.filter(
CrmProcessStage.id == stage_id,
CrmProcessStage.process_definition_id == process_id,
)
.first()
)
if not stage:
raise ValueError("مرحله CRM یافت نشد")
code_val = args.get("code")
if code_val and str(code_val).strip():
code_val = str(code_val).strip()
else:
code_val = generate_document_code(db, business_id, "crm_lead", date_cls.today())
lead = Lead(
business_id=business_id,
process_definition_id=process_id,
stage_id=stage_id,
code=code_val,
source_code=args.get("source_code"),
name=args["name"],
company_name=args.get("company_name"),
mobile=args.get("mobile"),
email=args.get("email"),
description=args.get("description"),
assigned_to_user_id=args.get("assigned_to_user_id"),
created_by_user_id=user_id,
)
db.add(lead)
db.commit()
db.refresh(lead)
try:
from app.services.workflow.workflow_trigger_service import trigger_lead_created
trigger_lead_created(
db,
business_id,
lead_id=lead.id,
process_definition_id=lead.process_definition_id,
stage_id=lead.stage_id,
name=lead.name,
user_id=user_id,
)
except Exception:
pass
return {
"id": lead.id,
"code": lead.code,
"name": lead.name,
"stage_id": lead.stage_id,
"process_definition_id": lead.process_definition_id,
}
registry.register(
AIFunction(
name="create_lead",
description=(
"ایجاد سرنخ CRM. process_definition_id و stage_id را از قیف/pipeline بگیر. "
"name الزامی است. نیاز به تأیید."
),
parameters_schema={
"type": "object",
"properties": {
"process_definition_id": {
"type": "integer",
"description": "شناسه تعریف قیف/فرآیند CRM",
},
"stage_id": {"type": "integer", "description": "شناسه مرحله قیف"},
"name": {"type": "string", "description": "نام سرنخ"},
"code": {"type": "string", "description": "کد سرنخ (اختیاری)"},
"source_code": {"type": "string", "description": "کد منبع جذب (اختیاری)"},
"company_name": {"type": "string", "description": "نام شرکت (اختیاری)"},
"mobile": {"type": "string", "description": "موبایل (اختیاری)"},
"email": {"type": "string", "format": "email", "description": "ایمیل (اختیاری)"},
"description": {"type": "string", "description": "توضیحات (اختیاری)"},
"assigned_to_user_id": {
"type": "integer",
"description": "شناسه کاربر مسئول (اختیاری)",
},
},
"required": ["process_definition_id", "stage_id", "name"],
},
handler=create_lead_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["crm.write"],
category="crm",
requires_approval=True,
risk_level="medium",
is_readonly=False,
)
)
# --- workflow execute ---
def execute_workflow_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from adapters.db.models.workflow import Workflow, WorkflowStatus
from app.services.workflow.workflow_engine import WorkflowEngine
db = context["db"]
business_id = int(args.get("business_id") or context.get("business_id"))
user_id = context["user_context"].get_user_id()
workflow_id = int(args["workflow_id"])
trigger_data = args.get("trigger_data") or {}
if not isinstance(trigger_data, dict):
trigger_data = {}
dry_run = bool(args.get("dry_run", False))
if dry_run:
from app.services.workflow.dry_run import DRY_RUN_TRIGGER_KEY
trigger_data = {**trigger_data, DRY_RUN_TRIGGER_KEY: True}
workflow = db.get(Workflow, workflow_id)
if not workflow or workflow.business_id != business_id:
raise ValueError("Workflow یافت نشد")
if workflow.status != WorkflowStatus.ACTIVE:
raise ValueError("Workflow فعال نیست")
engine = WorkflowEngine(db, business_id, user_id)
execution = engine.execute_workflow(workflow, trigger_data)
return {
"execution_id": execution.id,
"workflow_id": workflow_id,
"status": getattr(execution.status, "value", str(execution.status)),
"dry_run": dry_run,
}
registry.register(
AIFunction(
name="execute_workflow",
description="اجرای دستی یک workflow فعال. trigger_data اختیاری. dry_run برای آزمایش.",
parameters_schema={
"type": "object",
"properties": {
"workflow_id": {"type": "integer"},
"trigger_data": {"type": "object"},
"dry_run": {"type": "boolean"},
},
"required": ["workflow_id"],
},
handler=execute_workflow_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["workflows.write"],
category="workflow",
requires_approval=True,
risk_level="high",
is_readonly=False,
)
)
# --- export ---
def export_business_data_handler(args: Dict[str, Any], context: Dict[str, Any]) -> Any:
from app.services.ai.ai_export_service import export_business_file
db = context["db"]
business_id = int(args.get("business_id") or context.get("business_id"))
export_type = str(args.get("export_type", "")).strip().lower()
fmt = str(args.get("format", "excel")).strip().lower()
if export_type not in _EXPORT_TYPES:
raise ValueError(f"export_type نامعتبر. مجاز: {', '.join(_EXPORT_TYPES)}")
if fmt not in ("excel", "xlsx", "pdf"):
raise ValueError("format باید excel یا pdf باشد")
if export_type == "checks":
raise ValueError("export checks از AI هنوز پشتیبانی نمی‌شود؛ از search_checks استفاده کنید")
filters = dict(args.get("filters") or {})
file_bytes, filename, mime = export_business_file(
db, business_id, export_type, fmt, filters
)
if len(file_bytes) > 8 * 1024 * 1024:
return {
"error": "EXPORT_TOO_LARGE",
"message": "حجم فایل بیش از ۸ مگابایت است. فیلترها را محدود کنید.",
"size_bytes": len(file_bytes),
}
return {
"filename": filename,
"mime_type": mime,
"size_bytes": len(file_bytes),
"content_base64": base64.b64encode(file_bytes).decode("ascii"),
"export_type": export_type,
"format": fmt,
}
registry.register(
AIFunction(
name="export_business_data",
description=(
"خروجی Excel/PDF داده‌های کسب‌وکار. "
f"export_type: {', '.join(_EXPORT_TYPES)}. "
"format: excel یا pdf. نتیجه شامل content_base64 برای دانلود."
),
parameters_schema={
"type": "object",
"properties": {
"export_type": {"type": "string", "enum": _EXPORT_TYPES},
"format": {"type": "string", "enum": ["excel", "pdf"]},
"filters": {
"type": "object",
"description": "from_date, to_date, search, document_type, ...",
},
},
"required": ["export_type"],
},
handler=export_business_data_handler,
allowed_roles={AIRole.USER, AIRole.BUSINESS_OWNER, AIRole.OPERATOR, AIRole.ADMIN},
required_permissions=["reports.read"],
category="export",
requires_approval=True,
risk_level="medium",
is_readonly=False,
)
)