Files
rengine/shared/models/scan_schedule.py
T
Yogesh Ojha 1945242c1b feat: implement scan scheduling functionality
- Added new types and interfaces for scan scheduling in the frontend.
- Created a new Svelte page for managing scheduled scans, including listing, creating, editing, and deleting schedules.
- Implemented backend models and services for scan schedules, including handling different schedule types (one-off, interval, daily, cron).
- Introduced a Celery task to manage the execution of scheduled scans based on their defined timing.
- Enhanced notification system to include alerts for newly discovered subdomains during scans.
- Added utility functions for handling time zones and scheduling logic.
2026-06-24 08:10:21 +05:30

108 lines
3.5 KiB
Python

import uuid
from datetime import datetime
from pydantic import BaseModel
from sqlalchemy import Column
from sqlalchemy.types import JSON
from sqlmodel import Field, SQLModel
from shared.definitions.schedule_constants import MAX_SCHEDULE_TARGETS
from shared.enums.scan_schedule import ScheduleStatus
from shared.utils.datetime import utc_now
class ScanSchedule(SQLModel, table=True):
__tablename__ = "scan_schedules"
id: uuid.UUID = Field(default_factory=uuid.uuid4, primary_key=True, index=True)
project_id: uuid.UUID = Field(foreign_key="projects.id", index=True)
name: str = Field(max_length=200)
target_ids: list = Field(
default_factory=list, sa_column=Column(JSON, nullable=False)
)
engine_id: uuid.UUID = Field(index=True)
engine_name: str = Field(max_length=200)
context_id: uuid.UUID | None = Field(default=None)
context_name: str | None = Field(default=None, max_length=200)
schedule_type: str = Field(index=True)
run_at: datetime | None = Field(default=None)
interval_every: int | None = Field(default=None)
interval_unit: str | None = Field(default=None, max_length=10)
daily_at_time: str | None = Field(default=None, max_length=5)
cron_expression: str | None = Field(default=None, max_length=120)
timezone: str = Field(default="UTC", max_length=64)
status: str = Field(default=ScheduleStatus.ACTIVE.value, index=True)
next_run_at: datetime | None = Field(default=None, index=True)
last_run_at: datetime | None = Field(default=None)
last_error: str | None = Field(default=None, max_length=2000)
total_run_count: int = Field(default=0)
created_by: uuid.UUID = Field(foreign_key="users.id")
created_at: datetime = Field(default_factory=utc_now)
updated_at: datetime = Field(default_factory=utc_now)
class ScheduleTargetRef(BaseModel):
id: uuid.UUID
target_value: str
target_type: str
class ScanScheduleCreate(BaseModel):
name: str
target_ids: list[uuid.UUID] = Field(min_length=1, max_length=MAX_SCHEDULE_TARGETS)
engine_id: uuid.UUID
context_id: uuid.UUID | None = None
schedule_type: str
run_at: datetime | None = None
interval_every: int | None = None
interval_unit: str | None = None
daily_at_time: str | None = None
cron_expression: str | None = None
class ScanScheduleUpdate(BaseModel):
name: str | None = None
target_ids: list[uuid.UUID] | None = Field(
default=None, min_length=1, max_length=MAX_SCHEDULE_TARGETS
)
engine_id: uuid.UUID | None = None
context_id: uuid.UUID | None = None
schedule_type: str | None = None
run_at: datetime | None = None
interval_every: int | None = None
interval_unit: str | None = None
daily_at_time: str | None = None
cron_expression: str | None = None
class ScanScheduleRead(BaseModel):
id: uuid.UUID
project_id: uuid.UUID
name: str
target_ids: list[uuid.UUID]
targets: list[ScheduleTargetRef]
engine_id: uuid.UUID
engine_name: str
context_id: uuid.UUID | None
context_name: str | None
schedule_type: str
run_at: datetime | None
interval_every: int | None
interval_unit: str | None
daily_at_time: str | None
cron_expression: str | None
timezone: str
status: str
next_run_at: datetime | None
last_run_at: datetime | None
last_error: str | None
total_run_count: int
cadence: str
timezone_stale: bool = False
created_by: uuid.UUID
created_at: datetime
updated_at: datetime