retab 0.0.66__py3-none-any.whl → 0.0.68__py3-none-any.whl

retab/__init__.py

@@ -1,4 +1,4 @@
-from .client import AsyncRetab, Retab
+from .client import AsyncRetab, Retab, SignatureVerificationError
 from . import utils
 from . import types
-__all__ = ["Retab", "AsyncRetab", "utils", "types"]
+__all__ = ["Retab", "AsyncRetab", "SignatureVerificationError", "utils", "types"]

retab/client.py

@@ -1,3 +1,5 @@
+import hashlib
+import hmac
 import json
 import os
 from types import TracebackType
@@ -12,6 +14,11 @@ from .resources import documents, models, schemas, projects
 from .types.standards import PreparedRequest, FieldUnset
 
 
+class SignatureVerificationError(Exception):
+    """Raised when webhook signature verification fails."""
+    pass
+
+
 class MaxRetriesExceeded(Exception):
     pass
 
@@ -34,7 +41,7 @@ class BaseRetab:
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
         base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
-        timeout (float): Request timeout in seconds. Defaults to 240.0
+        timeout (float): Request timeout in seconds. Defaults to 1800.0 (30 minutes)
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
 
@@ -46,7 +53,7 @@ class BaseRetab:
         self,
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
-        timeout: float = 800.0,
+        timeout: float = 1800.0,
         max_retries: int = 3,
         openai_api_key: Optional[str] = FieldUnset,
         gemini_api_key: Optional[str] = FieldUnset,
@@ -140,7 +147,7 @@ class Retab(BaseRetab):
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
         base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
-        timeout (float): Request timeout in seconds. Defaults to 240.0
+        timeout (float): Request timeout in seconds. Defaults to 1800.0 (30 minutes)
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
         gemini_api_key (str, optional): Gemini API key. Will look for GEMINI_API_KEY env variable if not provided
@@ -161,7 +168,7 @@ class Retab(BaseRetab):
         self,
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
-        timeout: float = 240.0,
+        timeout: float = 1800.0,
         max_retries: int = 3,
         openai_api_key: Optional[str] = FieldUnset,
         gemini_api_key: Optional[str] = FieldUnset,
@@ -385,6 +392,44 @@ class Retab(BaseRetab):
         """
         self.close()
 
+    @staticmethod
+    def verify_event(event_body: bytes, event_signature: str, secret: str) -> Any:
+        """Verify the signature of a webhook event.
+
+        Args:
+            event_body: The raw request body as bytes
+            event_signature: The signature from the request header (x-retab-signature)
+            secret: The webhook secret key used for signing
+
+        Returns:
+            Any: The parsed event payload (JSON)
+
+        Raises:
+            SignatureVerificationError: If the signature verification fails
+
+        Example:
+            ```python
+            from retab import Retab
+
+            # In your webhook handler
+            secret = "your_webhook_secret"
+            body = request.body  # Raw bytes
+            signature = request.headers.get("x-retab-signature")
+
+            try:
+                event = Retab.verify_event(body, signature, secret)
+                print(f"Verified event: {event}")
+            except SignatureVerificationError:
+                print("Invalid signature!")
+            ```
+        """
+        expected_signature = hmac.new(secret.encode(), event_body, hashlib.sha256).hexdigest()
+
+        if not hmac.compare_digest(event_signature, expected_signature):
+            raise SignatureVerificationError("Invalid signature")
+
+        return json.loads(event_body.decode("utf-8"))
+
 
 class AsyncRetab(BaseRetab):
     """Asynchronous client for interacting with the Retab API.
@@ -395,7 +440,7 @@ class AsyncRetab(BaseRetab):
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
         base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
-        timeout (float): Request timeout in seconds. Defaults to 240.0
+        timeout (float): Request timeout in seconds. Defaults to 1800.0 (30 minutes)
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
         claude_api_key (str, optional): Claude API key. Will look for CLAUDE_API_KEY env variable if not provided
@@ -418,7 +463,7 @@ class AsyncRetab(BaseRetab):
         self,
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
-        timeout: float = 240.0,
+        timeout: float = 1800.0,
         max_retries: int = 3,
         openai_api_key: Optional[str] = FieldUnset,
         gemini_api_key: Optional[str] = FieldUnset,
@@ -661,3 +706,41 @@ class AsyncRetab(BaseRetab):
             traceback: The traceback of the exception that was raised, if any
         """
         await self.close()
+    @staticmethod
+    def verify_event(event_body: bytes, event_signature: str, secret: str) -> Any:
+        """Verify the signature of a webhook event.
+
+        Args:
+            event_body: The raw request body as bytes
+            event_signature: The signature from the request header (x-retab-signature)
+            secret: The webhook secret key used for signing
+
+        Returns:
+            Any: The parsed event payload (JSON)
+
+        Raises:
+            SignatureVerificationError: If the signature verification fails
+
+        Example:
+            ```python
+            from retab import AsyncRetab
+
+            # In your async webhook handler
+            secret = "your_webhook_secret"
+            body = await request.body()  # Raw bytes
+            signature = request.headers.get("x-retab-signature")
+
+            try:
+                event = AsyncRetab.verify_event(body, signature, secret)
+                print(f"Verified event: {event}")
+            except SignatureVerificationError:
+                print("Invalid signature!")
+            ```
+        """
+        expected_signature = hmac.new(secret.encode(), event_body, hashlib.sha256).hexdigest()
+
+        if not hmac.compare_digest(event_signature, expected_signature):
+            raise SignatureVerificationError("Invalid signature")
+
+        return json.loads(event_body.decode("utf-8"))
+

retab/generate_types.py

@@ -59,7 +59,7 @@ def type_to_zod(field_type: Any, put_names: bool = True, ts: bool = False) -> st
             optional = True
         typename = make_union([type_to_zod(x) for x in args])
         ts_typename = make_ts_union([type_to_zod(x, ts=True) for x in args])
-    elif issubclass(origin, BaseModel) or is_typeddict(origin) or is_typeddict_ext(origin):
+    elif isinstance(origin, type) and (issubclass(origin, BaseModel) or is_typeddict(origin) or is_typeddict_ext(origin)):
         if put_names:
             name = get_class_name(origin)
             typename = "Z" + name
@@ -77,7 +77,7 @@ def type_to_zod(field_type: Any, put_names: bool = True, ts: bool = False) -> st
 
             typename += "z.object({\n"
             ts_typename += "{\n"
-            props = [(n, f.annotation, f.default) for n, f in origin.model_fields.items() if not f.exclude] if issubclass(origin, BaseModel) else \
+            props = [(n, f.annotation, f.default) for n, f in origin.model_fields.items() if not f.exclude] if isinstance(origin, type) and issubclass(origin, BaseModel) else \
                     [(n, f, PydanticUndefined) for n, f in origin.__annotations__.items()]
 
             for field_name, field, default in props:

retab/resources/projects/client.py

@@ -8,11 +8,8 @@ from pydantic import HttpUrl
 from ..._resource import AsyncAPIResource, SyncAPIResource
 from ...utils.mime import MIMEData, prepare_mime_document
 from ...types.documents.extract import RetabParsedChatCompletion
-from ...types.projects import Project, PatchProjectRequest, BaseProject
+from ...types.projects import Project, PatchProjectRequest, CreateProjectRequest
 from ...types.standards import PreparedRequest, DeleteResponse, FieldUnset
-from .documents import Documents, AsyncDocuments
-from .iterations import Iterations, AsyncIterations
-
 
 class ProjectsMixin:
     def prepare_create(
@@ -29,7 +26,7 @@ class ProjectsMixin:
         if extra_body:
             eval_dict.update(extra_body)
 
-        eval_data = BaseProject(**eval_dict)
+        eval_data = CreateProjectRequest(**eval_dict)
         return PreparedRequest(method="POST", url="/v1/projects", data=eval_data.model_dump(exclude_unset=True, mode="json"))
 
     def prepare_get(self, project_id: str) -> PreparedRequest:
@@ -76,6 +73,10 @@ class ProjectsMixin:
     def prepare_delete(self, id: str) -> PreparedRequest:
         return PreparedRequest(method="DELETE", url=f"/v1/projects/{id}")
 
+    def prepare_publish(self, project_id: str, **extra_body: Any) -> PreparedRequest:
+        data = extra_body or None
+        return PreparedRequest(method="POST", url=f"/v1/projects/{project_id}/publish", data=data)
+
     def prepare_extract(
         self,
         project_id: str,
@@ -160,8 +161,6 @@ class Projects(SyncAPIResource, ProjectsMixin):
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
-        self.documents = Documents(self._client)
-        self.iterations = Iterations(self._client)
 
     def create(
         self,
@@ -174,9 +173,7 @@ class Projects(SyncAPIResource, ProjectsMixin):
 
         Args:
             name: The name of the project
-            json_schema: The JSON schema for the project
-            documents: The documents to associate with the project
-
+            json_schema: The json schema of the project
         Returns:
             Project: The created project
         Raises:
@@ -202,37 +199,6 @@ class Projects(SyncAPIResource, ProjectsMixin):
         response = self._client._prepared_request(request)
         return Project(**response)
 
-    def update(
-        self,
-        project_id: str,
-        name: str = FieldUnset,
-        json_schema: dict[str, Any] = FieldUnset,
-        **extra_body: Any,
-    ) -> Project:
-        """
-        Update an project with partial updates.
-
-        Args:
-            project_id: The ID of the project to update
-            name: Optional new name for the project
-            json_schema: Optional new JSON schema
-            documents: Optional list of documents to update
-            iterations: Optional list of iterations to update
-
-        Returns:
-            Project: The updated project
-        Raises:
-            HTTPException if the request fails
-        """
-        request = self.prepare_update(
-            project_id=project_id,
-            name=name,
-            json_schema=json_schema,
-            **extra_body,
-        )
-        response = self._client._prepared_request(request)
-        return Project(**response)
-
     def list(self, **extra_params: Any) -> List[Project]:
         """
         List projects for a project.
@@ -262,6 +228,12 @@ class Projects(SyncAPIResource, ProjectsMixin):
         request = self.prepare_delete(project_id)
         return self._client._prepared_request(request)
 
+    def publish(self, project_id: str, **extra_body: Any) -> Project:
+        """Publish a project's draft configuration."""
+        request = self.prepare_publish(project_id, **extra_body)
+        response = self._client._prepared_request(request)
+        return Project(**response)
+
     def extract(
         self,
         project_id: str,
@@ -314,23 +286,20 @@ class AsyncProjects(AsyncAPIResource, ProjectsMixin):
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
-        self.documents = AsyncDocuments(self._client)
-        self.iterations = AsyncIterations(self._client)
 
-    async def create(self, name: str, json_schema: Dict[str, Any]) -> Project:
+    async def create(self, name: str, json_schema: dict[str, Any], **extra_body: Any) -> Project:
         """
         Create a new project.
 
         Args:
             name: The name of the project
-            json_schema: The JSON schema for the project
-
+            json_schema: The json schema of the project
         Returns:
             Project: The created project
         Raises:
             HTTPException if the request fails
         """
-        request = self.prepare_create(name, json_schema)
+        request = self.prepare_create(name, json_schema, **extra_body)
         response = await self._client._prepared_request(request)
         return Project(**response)
 
@@ -350,36 +319,7 @@ class AsyncProjects(AsyncAPIResource, ProjectsMixin):
         response = await self._client._prepared_request(request)
         return Project(**response)
 
-    async def update(
-        self,
-        project_id: str,
-        name: str = FieldUnset,
-        json_schema: dict[str, Any] = FieldUnset,
-    ) -> Project:
-        """
-        Update an project with partial updates.
-
-        Args:
-            id: The ID of the project to update
-            name: Optional new name for the project
-            json_schema: Optional new JSON schema
-            documents: Optional list of documents to update
-            iterations: Optional list of iterations to update
-
-        Returns:
-            Project: The updated project
-        Raises:
-            HTTPException if the request fails
-        """
-        request = self.prepare_update(
-            project_id=project_id,
-            name=name,
-            json_schema=json_schema,
-        )
-        response = await self._client._prepared_request(request)
-        return Project(**response)
-
-    async def list(self) -> List[Project]:
+    async def list(self, **extra_params: Any) -> List[Project]:
         """
         List projects for a project.
 
@@ -388,7 +328,7 @@ class AsyncProjects(AsyncAPIResource, ProjectsMixin):
         Raises:
             HTTPException if the request fails
         """
-        request = self.prepare_list()
+        request = self.prepare_list(**extra_params)
         response = await self._client._prepared_request(request)
         return [Project(**item) for item in response.get("data", [])]
 
@@ -407,6 +347,12 @@ class AsyncProjects(AsyncAPIResource, ProjectsMixin):
         request = self.prepare_delete(project_id)
         return await self._client._prepared_request(request)
 
+    async def publish(self, project_id: str, **extra_body: Any) -> Project:
+        """Publish a project's draft configuration."""
+        request = self.prepare_publish(project_id, **extra_body)
+        response = await self._client._prepared_request(request)
+        return Project(**response)
+
     async def extract(
         self,
         project_id: str,
@@ -419,6 +365,7 @@ class AsyncProjects(AsyncAPIResource, ProjectsMixin):
         n_consensus: int | None = None,
         seed: int | None = None,
         store: bool = True,
+        **extra_form: Any,
     ) -> RetabParsedChatCompletion:
         """Extract documents from a project.
 

retab/types/mime.py

@@ -8,6 +8,8 @@ from typing import Any, Optional, Self, Sequence
 from pydantic import BaseModel, Field, field_validator
 from ..utils.hashing import generate_blake2b_hash_from_base64
 
+import io
+
 # Add webp and heic to the list of supported mime types
 mimetypes.add_type("image/webp", ".webp")
 mimetypes.add_type("image/heic", ".heic")
@@ -85,8 +87,17 @@ class OCR(BaseModel):
 
 
 class MIMEData(BaseModel):
-    filename: str = Field(description="The filename of the file", examples=["file.pdf", "image.png", "data.txt"])
-    url: str = Field(description="The URL of the file in base64 format", examples=["data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAADIA..."])
+    filename: str = Field(
+        description="The filename of the file",
+        examples=["file.pdf", "image.png", "data.txt"]
+    )
+    url: str = Field(
+        description="The URL of the file in base64 format",
+        examples=["data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAADIA..."]
+    )
+
+    # Internal resource
+    _buffer: Optional[io.BytesIO] = None
 
     @property
     def id(self) -> str:
@@ -99,18 +110,14 @@ class MIMEData(BaseModel):
     @property
     def content(self) -> str:
         if self.url.startswith("data:"):
-            # Extract base64 content from data URL
-            base64_content = self.url.split(",")[1]
-            return base64_content
-        else:
-            raise ValueError("Content is not available for this file")
+            return self.url.split(",")[1]
+        raise ValueError("Content is not available for this file")
 
     @property
     def mime_type(self) -> str:
         if self.url.startswith("data:"):
             return self.url.split(";")[0].split(":")[1]
-        else:
-            return mimetypes.guess_type(self.filename)[0] or "application/octet-stream"
+        return mimetypes.guess_type(self.filename)[0] or "application/octet-stream"
 
     @property
     def unique_filename(self) -> str:
@@ -118,22 +125,57 @@ class MIMEData(BaseModel):
 
     @property
     def size(self) -> int:
-        # size in bytes
         return len(base64.b64decode(self.content))
 
+    # def to_bytesio(self) -> io.BytesIO:
+    #     """Decode base64 and return a BytesIO (without leaking references)."""
+    #     buf = io.BytesIO(base64.b64decode(self.content))
+    #     buf.seek(0)
+    #     return buf
+
+    # # -------- Context manager interface --------
+
+    # def __enter__(self) -> io.BytesIO:
+    #     """Opens the internal buffer so you can use it like a file."""
+    #     if self._buffer is None:
+    #         self._buffer = self.to_bytesio()
+    #     return self._buffer
+
+    # def __exit__(self, exc_type, exc_val, exc_tb):
+    #     """Close and cleanup the buffer."""
+    #     if self._buffer is not None:
+    #         self._buffer.close()
+    #         self._buffer = None
+
+    # # -------- Optional convenience methods --------
+
+    # def open(self) -> io.BytesIO:
+    #     """Manual open without `with`."""
+    #     return self.__enter__()
+
+    # def close(self):
+    #     """Manual close."""
+    #     self.__exit__(None, None, None)
+
     def __str__(self) -> str:
         truncated_url = self.url[:50] + "..." if len(self.url) > 50 else self.url
-        # truncated_content = self.content[:50] + '...' if len(self.content) > 50 else self.content
-        return f"MIMEData(filename='{self.filename}', url='{truncated_url}', mime_type='{self.mime_type}', size='{self.size}', extension='{self.extension}')"
+        return (
+            f"MIMEData(filename='{self.filename}', "
+            f"url='{truncated_url}', "
+            f"mime_type='{self.mime_type}', "
+            f"size='{self.size}', "
+            f"extension='{self.extension}')"
+        )
 
     def __repr__(self) -> str:
         return self.__str__()
 
 
+
 class BaseMIMEData(MIMEData):
     @classmethod
     def model_validate(
-        cls, obj: Any, *, strict: bool | None = None, from_attributes: bool | None = None, context: Any | None = None, by_alias: bool | None = None, by_name: bool | None = None
+        cls, obj: Any, *, strict: bool | None = None, extra: Any | None = None, from_attributes: bool | None = None, context: Any | None = None, by_alias: bool | None = None, by_name: bool | None = None
     ) -> Self:
         if isinstance(obj, MIMEData):
             # Convert MIMEData instance to dict
@@ -153,7 +195,7 @@ class BaseMIMEData(MIMEData):
                 else:
                     # If there's no comma (unexpected format), truncate to 996 chars (multiple of 4)
                     obj["url"] = obj["url"][:996]
-        return super().model_validate(obj, strict=strict, from_attributes=from_attributes, context=context, by_alias=by_alias, by_name=by_name)
+        return super().model_validate(obj, strict=strict, extra=extra, from_attributes=from_attributes, context=context, by_alias=by_alias, by_name=by_name)
 
     @property
     def id(self) -> str:

retab/types/projects/__init__.py

@@ -1,33 +1,8 @@
-from .model import Project, BaseProject, CreateProjectRequest, PatchProjectRequest
-from .documents import AnnotatedDocument, DocumentItem, ProjectDocument, CreateProjectDocumentRequest, PatchProjectDocumentRequest
-from .iterations import (
-    BaseIteration,
-    Iteration,
-    CreateIterationRequest,
-    PatchIterationRequest,
-    ProcessIterationRequest,
-    DocumentStatus,
-    IterationDocumentStatusResponse,
-    AddIterationFromJsonlRequest,
-)
+from .model import Project, CreateProjectRequest, PatchProjectRequest
 
 
 __all__ = [
     "Project",
-    "BaseProject",
     "CreateProjectRequest",
-    "PatchProjectRequest",
-    "AnnotatedDocument",
-    "DocumentItem",
-    "ProjectDocument",
-    "CreateProjectDocumentRequest",
-    "PatchProjectDocumentRequest",
-    "BaseIteration",
-    "Iteration",
-    "CreateIterationRequest",
-    "PatchIterationRequest",
-    "ProcessIterationRequest",
-    "DocumentStatus",
-    "IterationDocumentStatusResponse",
-    "AddIterationFromJsonlRequest",
+    "PatchProjectRequest"
 ]

retab/types/projects/model.py

@@ -4,8 +4,6 @@ from typing import Any, Optional
 import nanoid  # type: ignore
 from pydantic import BaseModel, Field, ConfigDict
 
-from .documents import ProjectDocument
-from .iterations import Iteration
 from ..inference_settings import InferenceSettings
 
 default_inference_settings = InferenceSettings(
@@ -17,42 +15,68 @@ default_inference_settings = InferenceSettings(
     browser_canvas="A4",
     n_consensus=1,
 )
+class Function(BaseModel):
+    model_config = ConfigDict(extra="ignore")
+    id: str = Field(default_factory=lambda: "function_" + nanoid.generate())
+    path: str
+    code: Optional[str] = Field(default=None, description="The code of the function")
+    function_registry_id: Optional[str] = Field(default=None, description="The function registry id of the function")
+
+    # @model_validator(mode="before")
+    # @classmethod
+    # def validate_function(cls, data: Any):
+    #     if isinstance(data, dict):
+    #         code = data.get("code")
+    #         function_registry_id = data.get("function_registry_id")
+    #         if code is None and function_registry_id is None:
+    #             raise ValueError("Either code or function_registry_id must be provided")
+    #     return data
+
+class FunctionHilCriterion(BaseModel):
+    path: str
+    agentic_fix: bool = Field(default=False, description="Whether to use agentic fix for the criterion")
 
-class SheetsIntegration(BaseModel):
-    sheet_id: str
-    spreadsheet_id: str
+class HumanInTheLoopParams(BaseModel):
+    enabled: bool = Field(default=False)
+    url: str = Field(default="", description="The URL of the human in the loop endpoint")
+    headers: dict[str, str] = Field(default_factory=dict, description="The headers to send to the human in the loop endpoint")
+    criteria: list[FunctionHilCriterion] = Field(default_factory=list, description="The criteria to use for the human in the loop")
 
-class BaseProject(BaseModel):
+class PublishedConfig(BaseModel):
+    inference_settings: InferenceSettings = default_inference_settings
+    json_schema: dict[str, Any] = Field(default_factory=dict, description="The json schema of the project")
+    human_in_the_loop_params: HumanInTheLoopParams = Field(default_factory=HumanInTheLoopParams)
+    origin: str = Field(default="manual", description="The origin of the published config. Either 'Manual' or the iteration id that was used to generate the config")
+class DraftConfig(BaseModel):
+    inference_settings: InferenceSettings = default_inference_settings
+    json_schema: dict[str, Any] = Field(default_factory=dict, description="The json schema of the builder config")
+    human_in_the_loop_criteria: list[FunctionHilCriterion] = Field(default_factory=list)
+class Project(BaseModel):
     model_config = ConfigDict(extra="ignore")
-    id: str = Field(default_factory=lambda: "proj_" + nanoid.generate())
+    id: str = Field(default_factory=lambda: "project_" + nanoid.generate())
     name: str = Field(default="", description="The name of the project")
-    json_schema: dict[str, Any] = Field(default_factory=dict, description="The json schema of the project")
     updated_at: datetime.datetime = Field(default_factory=lambda: datetime.datetime.now(tz=datetime.timezone.utc))
-    sheets_integration: SheetsIntegration | None = None
-    validation_flags: dict[str, Any] | None = None
-    inference_settings: InferenceSettings = default_inference_settings
+    published_config: PublishedConfig
+    draft_config: DraftConfig
+    is_published: bool = False
+    #computation_spec: ComputationSpec = Field(default_factory=ComputationSpec, description="The computation spec of the project")
+    functions: list[Function] = Field(default_factory=list, description="The functions of the project")
 
-# Actual Object stored in DB
-class Project(BaseProject):
-    documents: list[ProjectDocument] = Field(default_factory=list)
-    iterations: list[Iteration] = Field(default_factory=list)
+class StoredProject(Project):
+    """Project model with organization_id for database storage"""
+    organization_id: str
 
 class CreateProjectRequest(BaseModel):
     model_config = ConfigDict(extra="ignore")
     name: str
-    json_schema: dict[str, Any]
-
+    json_schema: dict[str, Any] = Field(default_factory=dict, description="The json schema of the project")
 
-# This is basically the same as BaseProject, but everything is optional.
-# Could be achieved by convert_basemodel_to_partial_basemodel(BaseProject) but we prefer explicitness
+# This is basically the same as Project, but everything is optional.
 class PatchProjectRequest(BaseModel):
     model_config = ConfigDict(extra="ignore")
     name: Optional[str] = Field(default=None, description="The name of the document")
-    json_schema: Optional[dict[str, Any]] = Field(default=None, description="The json schema of the project")
-    sheets_integration: SheetsIntegration | None = None
-    validation_flags: Optional[dict[str, Any]] = Field(default=None, description="The validation flags of the project")
-    inference_settings: Optional[InferenceSettings] = Field(default=None, description="The inference settings of the project")
-
-class AddIterationFromJsonlRequest(BaseModel):
-    model_config = ConfigDict(extra="ignore")
-    jsonl_gcs_path: str
+    published_config: Optional[PublishedConfig] = Field(default=None, description="The published config of the project")
+    draft_config: Optional[DraftConfig] = Field(default=None, description="The draft config of the project")
+    is_published: Optional[bool] = Field(default=None, description="The published status of the project")
+    #computation_spec: Optional[ComputationSpec] = Field(default=None, description="The computation spec of the project")
+    functions: Optional[list[Function]] = Field(default=None, description="The functions of the project")

retab/types/schemas/chat.py

@@ -25,8 +25,8 @@ from openai.types.responses.response_input_message_content_list_param import Res
 from openai.types.responses.response_input_param import ResponseInputItemParam
 from openai.types.responses.response_input_text_param import ResponseInputTextParam
 
-from retab.types.chat import ChatCompletionRetabMessage
-from retab.types.documents.extract import RetabParsedChatCompletion, RetabParsedChoice
+from ...types.chat import ChatCompletionRetabMessage
+from ...types.documents.extract import RetabParsedChatCompletion, RetabParsedChoice
 
 
 MediaType = Literal["image/jpeg", "image/png", "image/gif", "image/webp"]

retab/types/schemas/model.py

@@ -1,7 +1,7 @@
 import copy
 import json
 from typing import Any, Literal, Optional, Self, Union, Type, MutableMapping, Tuple, MutableSequence
-from retab.utils.hashing import generate_blake2b_hash_from_string
+from ...utils.hashing import generate_blake2b_hash_from_string
 
 import datetime
 from pathlib import Path
@@ -16,7 +16,7 @@ from pydantic import BaseModel, Field, PrivateAttr, computed_field, model_valida
 from .chat import convert_to_anthropic_format, convert_to_google_genai_format
 from .chat import convert_to_openai_completions_api_format
 
-from retab.utils.json_schema import convert_json_schema_to_basemodel, expand_refs, load_json_schema
+from ...utils.json_schema import convert_json_schema_to_basemodel, expand_refs, load_json_schema
 from .chat import convert_to_openai_responses_api_format
 from ..standards import StreamingBaseModel
 from ..chat import ChatCompletionRetabMessage