retab 0.0.68__py3-none-any.whl → 0.0.70__py3-none-any.whl

retab/types/extractions/types.py

@@ -0,0 +1,3 @@
+from typing import Literal
+
+type HumanReviewStatus = Literal["success", "review_required", "reviewed"]

retab/types/inference_settings.py

@@ -1,17 +1,19 @@
 from openai.types.chat.chat_completion_reasoning_effort import ChatCompletionReasoningEffort
 from pydantic import BaseModel, Field, ConfigDict
 from .modality import Modality
-from .browser_canvas import BrowserCanvas
+from typing import Any
 
 class InferenceSettings(BaseModel):
     model: str = "gpt-5-mini"
     temperature: float = 0.0
     reasoning_effort: ChatCompletionReasoningEffort = "minimal"
     image_resolution_dpi: int = Field(default=192, description="Resolution of the image sent to the LLM", ge=96, le=300)
-    browser_canvas: BrowserCanvas = "A4"
-    n_consensus: int = Field(default=1, description="Number of consensus rounds to perform")
+    n_consensus: int = Field(default=1, ge=1, le=8, description="Number of consensus rounds to perform")
     modality: Modality = "native"
     parallel_ocr_keys: dict[str, str] | None = Field(default=None, description="If set, keys to be used for the extraction of long lists of data using Parallel OCR", examples=[{"properties": "ID", "products": "identity.id"}])
-
     model_config = ConfigDict(extra="ignore")
 
+
+class ExtractionSettings(InferenceSettings):
+    json_schema: dict[str, Any] = Field(..., description="JSON schema format used to validate the output data.")
+    

retab/types/mime.py

@@ -171,44 +171,10 @@ class MIMEData(BaseModel):
         return self.__str__()
 
 
-
-class BaseMIMEData(MIMEData):
-    @classmethod
-    def model_validate(
-        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
-            obj = obj.model_dump()
-        if isinstance(obj, dict) and "url" in obj:
-            # Truncate URL to 1000 chars or less, ensuring it's a valid base64 string
-            if len(obj["url"]) > 1000:
-                # Find the position of the base64 data
-                if "," in obj["url"]:
-                    prefix, base64_data = obj["url"].split(",", 1)
-                    # Calculate how many characters we can keep (must be a multiple of 4)
-                    max_base64_len = 1000 - len(prefix) - 1  # -1 for the comma
-                    # Ensure the length is a multiple of 4
-                    max_base64_len = max_base64_len - (max_base64_len % 4)
-                    # Truncate and reassemble
-                    obj["url"] = prefix + "," + base64_data[:max_base64_len]
-                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, extra=extra, from_attributes=from_attributes, context=context, by_alias=by_alias, by_name=by_name)
-
-    @property
-    def id(self) -> str:
-        raise NotImplementedError("id is not implemented for BaseMIMEData - id is the hash of the content, so it's not possible to generate it from the base class")
-
-    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"BaseMIMEData(filename='{self.filename}', url='{truncated_url}', content='{truncated_content}', mime_type='{self.mime_type}', extension='{self.extension}')"
-
-    def __repr__(self) -> str:
-        return self.__str__()
-
+class BaseMIMEData(BaseModel):
+    id: str = Field(..., description="ID of the file")
+    filename: str = Field(..., description="Filename of the file")
+    mime_type: str = Field(..., description="MIME type of the file")
 
 # **** MIME DATACLASSES ****
 class AttachmentMetadata(BaseModel):

retab/types/pagination.py

@@ -1,6 +1,14 @@
+from typing import Any, List, Literal
 from pydantic import BaseModel
 
 
 class ListMetadata(BaseModel):
     before: str | None
     after: str | None
+
+
+class PaginatedList(BaseModel):
+    data: List[Any]
+    list_metadata: ListMetadata
+
+type PaginationOrder = Literal["asc", "desc"]

retab/types/projects/model.py

@@ -4,7 +4,9 @@ from typing import Any, Optional
 import nanoid  # type: ignore
 from pydantic import BaseModel, Field, ConfigDict
 
+from ..mime import BaseMIMEData, MIMEData
 from ..inference_settings import InferenceSettings
+from .predictions import PredictionData
 
 default_inference_settings = InferenceSettings(
     model="auto-small",
@@ -12,45 +14,29 @@ default_inference_settings = InferenceSettings(
     reasoning_effort="minimal",
     modality="native",
     image_resolution_dpi=192,
-    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):
+
+class Computation(BaseModel):
+    expression: str = Field(description="The expression to use for the computation")
+
+class ComputationSpec(BaseModel):
+    computations: dict[str, Computation] = Field(default_factory=dict, description="The computations to use for the project")
+
+class HilCriterion(BaseModel):
     path: str
     agentic_fix: bool = Field(default=False, description="Whether to use agentic fix for the criterion")
 
-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 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
+    inference_settings: InferenceSettings = Field(default=default_inference_settings, description="The inference settings of the project")
     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)
+    human_in_the_loop_criteria: list[HilCriterion] = Field(default_factory=list)
+    computation_spec: ComputationSpec = Field(default_factory=ComputationSpec, description="The computation spec of the project")
+
+class PublishedConfig(DraftConfig):
+    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 Project(BaseModel):
     model_config = ConfigDict(extra="ignore")
     id: str = Field(default_factory=lambda: "project_" + nanoid.generate())
@@ -59,8 +45,7 @@ class Project(BaseModel):
     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")
+    is_schema_generated: bool = Field(default=True, description="Whether the schema has been generated for the project")
 
 class StoredProject(Project):
     """Project model with organization_id for database storage"""
@@ -78,5 +63,33 @@ class PatchProjectRequest(BaseModel):
     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")
+    computation_spec: Optional[ComputationSpec] = Field(default=None, description="The computation spec of the project")
+# ----------------------------
+# ----------------------------
+# ----------------------------
+
+class BuilderDocument(BaseModel):
+    model_config = ConfigDict(extra="ignore")
+    id: str = Field(default_factory=lambda: "builder_doc_" + nanoid.generate())
+    updated_at: datetime.datetime = Field(default_factory=lambda: datetime.datetime.now(tz=datetime.timezone.utc))
+    project_id: str
+    mime_data: BaseMIMEData = Field(description="The mime data of the document")
+    
+    prediction_data: PredictionData = Field(default=PredictionData(), description="The prediction data of the document")
+    extraction_id: str | None = Field(default=None, description="The extraction id of the document")
+
+class StoredBuilderDocument(BuilderDocument):
+    """Builder document model with organization_id and project_id for database storage"""
+    organization_id: str
+
+class PatchBuilderDocumentRequest(BaseModel):
+    model_config = ConfigDict(extra="ignore")
+    extraction_id: Optional[str] = Field(default=None, description="The extraction id of the builder document")
+    prediction_data: Optional[PredictionData] = Field(default=None, description="The prediction data of the document")
+
+
+class AddBuilderDocumentRequest(BaseModel):
+    model_config = ConfigDict(extra="ignore")
+    mime_data: MIMEData
+    prediction_data: PredictionData = Field(default=PredictionData(), description="The prediction data of the document")
+    project_id: str

retab/types/schemas/generate.py

@@ -2,7 +2,6 @@ from openai.types.chat.chat_completion_reasoning_effort import ChatCompletionRea
 from pydantic import BaseModel, ConfigDict, Field
 
 from ..mime import MIMEData
-from ..browser_canvas import BrowserCanvas
 
 
 class GenerateSchemaRequest(BaseModel):
@@ -14,9 +13,6 @@ class GenerateSchemaRequest(BaseModel):
     instructions: str | None = None
     """The modality of the document to load."""
     image_resolution_dpi: int = Field(default=192, description="Resolution of the image sent to the LLM", ge=96, le=300)
-    browser_canvas: BrowserCanvas = Field(
-        default="A4", description="Sets the size of the browser canvas for rendering documents in browser-based processing. Choose a size that matches the document type."
-    )
 
     """The image operations to apply to the document."""
     stream: bool = False

{retab-0.0.68.dist-info → retab-0.0.70.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: retab
-Version: 0.0.68
+Version: 0.0.70
 Summary: Retab official python library
 Home-page: https://github.com/retab-dev/retab
 Author: Retab

{retab-0.0.68.dist-info → retab-0.0.70.dist-info}/RECORD

@@ -1,36 +1,38 @@
 retab/__init__.py,sha256=s4GawWTRBYz4VY-CyAV5-ZdFtdw8V5oopGIYm9GgdSo,188
 retab/_resource.py,sha256=JfAU4UTa05ugWfbrpO7fsVr_pFewht99NkoIfK6kBQM,577
-retab/client copy.py,sha256=RqKYBvLB8Bp6QpS3DodOXffWQFdr36bobQRSU09vES0,27883
-retab/client.py,sha256=wOiTAzGw6Hi7FVbrs-TwSuPR6tmgOLMYjY1eBgrgBBk,29929
+retab/client.py,sha256=fP9dI9YLFRNCqhODB85Gzp_sxhBJ5KF9R3U0qogP0dc,30075
 retab/generate_types.py,sha256=cUu1IX65uU__MHivmEb_PZtzAi8DYsvppZvcY30hj90,8425
 retab/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 retab/resources/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 retab/resources/models.py,sha256=4WidFBnTGZEA65DSn2pLP2SRnCVXkMTw7o_m8xVCFC4,2469
-retab/resources/schemas.py,sha256=qWkOAZrimtbTHzhDof3z4PG72F7MIPrb3A495f6mrHE,5020
+retab/resources/schemas.py,sha256=rZ6OzfmoYv-mGaRVzvXjO09dD-KxP74mZhOO8sMgcDQ,4632
 retab/resources/documents/__init__.py,sha256=OjXmngFN0RKqO4SI-mJBNzr6Ex6rMxfq0DxaqzP0RQs,89
-retab/resources/documents/client.py,sha256=7-BaveXUlwZqtMTvM3dNiltZS83jJVXDRuyfFMg_eGw,37446
+retab/resources/documents/client.py,sha256=H-ylTrFcaC36iVj5XSWbwJB489n3GEWwbqsPgQQ50G4,31980
+retab/resources/extractions/__init__.py,sha256=2H1ezUG8hI5SmTRy6NFzXdYLOdGFFsFrI60uzkitV20,97
+retab/resources/extractions/client.py,sha256=sEoNjOgX91FTOgoJUV-I1A9A9xl1ciCdPlhYwjhEjbA,11035
 retab/resources/projects/__init__.py,sha256=tPR3_3tr7bsoYd618qmGjnYN2R23PmF5oCFd7Z5_HGY,85
-retab/resources/projects/client.py,sha256=-T5iSRp-jODNVg-cgQiQ3S6D53cwCTwNQsI0Rr9vqsU,14654
+retab/resources/projects/client.py,sha256=nvqsDiVyeRrXsoYddwyXNmpZxaBLYWAGO2e0n6qkCzY,14988
 retab/types/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-retab/types/browser_canvas.py,sha256=LdW1SnaWb4WdFgFBCo8p9k78K2fmzV5YtloQtIR7hk0,82
 retab/types/chat.py,sha256=x9VbtPMa4w6Gc0HrFC3ILl6cCnfEn5ytDnwJtZmlcys,1436
-retab/types/inference_settings.py,sha256=co0C5ynGm7z7Bvi7gP3CFwhkRJUd-bMNBVbavFu7Iuo,939
-retab/types/mime.py,sha256=1k3REXEVJ4dlFErs27Xx05sIWHWr3KjLzuKSDXEMNQQ,11973
+retab/types/inference_settings.py,sha256=e4NDZHFdtrqi30Za1z0JhRU5PBO94yKxUGoK7S2kA3M,1053
+retab/types/mime.py,sha256=3Zk7vIbV8o4uJQiclVH-ncKKhs_ZeVi-UQV68TTu7s0,10039
 retab/types/modality.py,sha256=4B8LctdUBZVgIjtS2FjrJpljn2Eyse0XE1bpFsGb9O4,131
-retab/types/pagination.py,sha256=-XrKILKX_5isTHTfShLiK3Kwp21Y6Wqy0Jci8lIFQig,109
+retab/types/pagination.py,sha256=A0Fw06baPTfEaYwo3kvNs4vaupzlqylBc6tQH-2DFuY,279
 retab/types/standards.py,sha256=7aGtuvzzkKidvqY8JB2Cjfn43V80FeKwrTtp162kjKc,1477
 retab/types/documents/__init__.py,sha256=RaD6PnvRJw7QEVTh_PYNX6gckpLcxUJH7FKaopRKJzY,114
 retab/types/documents/correct_orientation.py,sha256=e-ivsslI6L6Gl0YkcslXw_DH620xMGEYVp4tdeviXeM,261
-retab/types/documents/create_messages.py,sha256=fibThBlJg1iKQfVkvIVL1QqZxrraIl9qeO0lJHLo_nQ,7509
-retab/types/documents/extract.py,sha256=KeEDnbvXTtgamrGvpBINsg5MJAue-Xf_lNi1a71M7oE,20067
-retab/types/documents/parse.py,sha256=ekXnjnFEx7adZ8dJKivo5QRtYnKBEUFzJiIM9WY-LPY,1454
+retab/types/documents/create_messages.py,sha256=Cox0QgIyLhTXIvw1Nzd2BCnB9-5KAYgw_gads5eTaDw,7272
+retab/types/documents/extract.py,sha256=Z-Qsj5-lmWb9dkgyTTt3Z7Lpd4_7SZdRGErtV_aw8oE,16492
+retab/types/documents/parse.py,sha256=Jd6i-1UXhAtgntRBZItEHGHeevyLdLmbTQa1-HNrico,1305
+retab/types/extractions/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+retab/types/extractions/types.py,sha256=mnCYSfJoEKsXN2eG-PrahnnQyR6RDjP5VO9sHC1Opmg,102
 retab/types/projects/__init__.py,sha256=I7P_dems5_LOLgYQ-4Bzt9B6P6jRlQwP-D_9GxRDhVk,155
 retab/types/projects/metrics.py,sha256=J8aZdVbqlszfxosAZyTB7l6lp9WgdL5QgLMlLrckN7k,1946
-retab/types/projects/model.py,sha256=0He-lMiT4jX1SwukNowq41LFPeozWDnTVsbjlmCNWJw,4374
+retab/types/projects/model.py,sha256=f5NSIvwQQXhax4gHu57CE0pUkU_5S8t3aSfzAAqFFlI,4615
 retab/types/projects/predictions.py,sha256=GeQX-nCmbzKL50aaSFlSyn0sgnbzKserPK6DSQT8ACk,1114
 retab/types/schemas/__init__.py,sha256=9ODWiC_4pUVKxoIKglYZjvRjRyd1ZCVxG8GBdQgHNbU,57
 retab/types/schemas/chat.py,sha256=ppTidxsNslTKE5aBva04i9IxeARMqYpXYLjtR7V6pBc,21219
-retab/types/schemas/generate.py,sha256=DN4pMJihoADyAZgA2RW7BddFzhbNOoaEnfEjjjUyMKM,1007
+retab/types/schemas/generate.py,sha256=kB69Sc8cSw4oqaw-Rfzc-vz3ModfKuX86KOqp47y2-8,744
 retab/types/schemas/layout.py,sha256=JLPwQGIWfPBoe1Y5r-MhiNDJigzZ-yKZnVGgox0uqMk,1487
 retab/types/schemas/model.py,sha256=MwiSrck_5NCDvxjDSqgiDtmenGowXnfoOZ7KTINKW34,69588
 retab/types/schemas/templates.py,sha256=XihWTHi6t_6QjxN07n_1dee5KdhHiuoHAYfmKwI7gQg,1708
@@ -42,7 +44,7 @@ retab/utils/mime.py,sha256=mTP_lqSPttOP5DYJxopiWaeFXrUCPjhwd7y53nCVGO4,6189
 retab/utils/stream_context_managers.py,sha256=gI1gVQSj3nWz6Mvjz7Ix5AiY0g6vSL-c2tPfuP04izo,2314
 retab/utils/usage/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 retab/utils/usage/json_schema.py,sha256=kRbL5E5OStlhlNlBXlxHNwaVHKd6MVhyqIb0y4aj8JA,84322
-retab-0.0.68.dist-info/METADATA,sha256=ZqKChL45ZWbRxN1B7QMPk-VWdz0qKezYOQOgPRhDFc0,4532
-retab-0.0.68.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-retab-0.0.68.dist-info/top_level.txt,sha256=waQR0EGdhLIQtztoE3AXg7ik5ONQ9q_bsKVpyFuJdq0,6
-retab-0.0.68.dist-info/RECORD,,
+retab-0.0.70.dist-info/METADATA,sha256=fmovHmnRSdICMGaxGo7aOrYiOXbqyMmXBxUiG7gtdEQ,4532
+retab-0.0.70.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+retab-0.0.70.dist-info/top_level.txt,sha256=waQR0EGdhLIQtztoE3AXg7ik5ONQ9q_bsKVpyFuJdq0,6
+retab-0.0.70.dist-info/RECORD,,