migrate url validators to skyvern.utils (#2014)

skyvern/forge/sdk/core/validators.py

@@ -1,61 +0,0 @@
-import ipaddress
-from urllib.parse import urlparse
-
-from fastapi import status
-from pydantic import HttpUrl, ValidationError
-
-from skyvern.config import settings
-from skyvern.exceptions import BlockedHost, InvalidUrl, SkyvernHTTPException
-
-
-def prepend_scheme_and_validate_url(url: str) -> str:
-    if not url:
-        return url
-
-    parsed_url = urlparse(url=url)
-    if parsed_url.scheme and parsed_url.scheme not in ["http", "https"]:
-        raise InvalidUrl(url=url)
-
-    # if url doesn't contain any scheme, we prepend `https` to it by default
-    if not parsed_url.scheme:
-        url = f"https://{url}"
-
-    try:
-        HttpUrl(url)
-    except ValidationError:
-        raise InvalidUrl(url=url)
-
-    return url
-
-
-def is_blocked_host(host: str) -> bool:
-    if host.lower() in (h.lower() for h in settings.ALLOWED_HOSTS):
-        return False
-    try:
-        ip = ipaddress.ip_address(host)
-        # Check if the IP is private, link-local, loopback, or reserved
-        return ip.is_private or ip.is_link_local or ip.is_loopback or ip.is_reserved
-    except ValueError:
-        # If the host is not a valid IP address (e.g., it's a domain name like localhost), handle it here
-        for blocked_host in settings.BLOCKED_HOSTS:
-            if blocked_host == host:
-                return True
-        return False
-    except Exception:
-        return False
-
-
-def validate_url(url: str) -> str | None:
-    try:
-        url = prepend_scheme_and_validate_url(url=url)
-        v = HttpUrl(url=url)
-    except Exception as e:
-        raise SkyvernHTTPException(message=str(e), status_code=status.HTTP_400_BAD_REQUEST)
-
-    if not v.host:
-        return None
-    host = v.host
-    blocked = is_blocked_host(host)
-    if blocked:
-        raise BlockedHost(host=host)
-    return str(v)

skyvern/forge/sdk/schemas/task_v2.py

@@ -4,8 +4,8 @@ from typing import Any
 
 from pydantic import BaseModel, ConfigDict, Field, field_validator
 
-from skyvern.forge.sdk.core.validators import validate_url
 from skyvern.schemas.runs import ProxyLocation
+from skyvern.utils.url_validators import validate_url
 
 DEFAULT_WORKFLOW_TITLE = "New Workflow"
 

skyvern/forge/sdk/schemas/tasks.py

@@ -7,10 +7,10 @@ from typing import Any
 from pydantic import BaseModel, Field, field_validator
 
 from skyvern.exceptions import InvalidTaskStatusTransition, TaskAlreadyCanceled, TaskAlreadyTimeout
-from skyvern.forge.sdk.core.validators import validate_url
 from skyvern.forge.sdk.db.enums import TaskType
 from skyvern.forge.sdk.schemas.files import FileInfo
 from skyvern.schemas.runs import ProxyLocation
+from skyvern.utils.url_validators import validate_url
 
 
 class TaskBase(BaseModel):

skyvern/forge/sdk/workflow/models/block.py

@@ -49,7 +49,6 @@ from skyvern.forge.sdk.api.files import (
 )
 from skyvern.forge.sdk.api.llm.api_handler_factory import LLMAPIHandlerFactory
 from skyvern.forge.sdk.artifact.models import ArtifactType
-from skyvern.forge.sdk.core.validators import prepend_scheme_and_validate_url
 from skyvern.forge.sdk.db.enums import TaskType
 from skyvern.forge.sdk.schemas.files import FileInfo
 from skyvern.forge.sdk.schemas.task_v2 import TaskV2Status
@@ -72,6 +71,7 @@ from skyvern.forge.sdk.workflow.models.parameter import (
     OutputParameter,
     WorkflowParameter,
 )
+from skyvern.utils.url_validators import prepend_scheme_and_validate_url
 from skyvern.webeye.browser_factory import BrowserState
 from skyvern.webeye.utils.page import SkyvernFrame
 

skyvern/forge/sdk/workflow/models/workflow.py

@@ -4,13 +4,13 @@ from typing import Any, List
 
 from pydantic import BaseModel, field_validator
 
-from skyvern.forge.sdk.core.validators import validate_url
 from skyvern.forge.sdk.schemas.files import FileInfo
 from skyvern.forge.sdk.schemas.task_v2 import TaskV2
 from skyvern.forge.sdk.workflow.exceptions import WorkflowDefinitionHasDuplicateBlockLabels
 from skyvern.forge.sdk.workflow.models.block import BlockTypeVar
 from skyvern.forge.sdk.workflow.models.parameter import PARAMETER_TYPE
 from skyvern.schemas.runs import ProxyLocation
+from skyvern.utils.url_validators import validate_url
 
 
 class WorkflowRequestBody(BaseModel):