mistralai 1.10.1__py3-none-any.whl → 1.11.1__py3-none-any.whl

mistralai_azure/models/ocrpageobject.py

@@ -3,10 +3,17 @@
 from __future__ import annotations
 from .ocrimageobject import OCRImageObject, OCRImageObjectTypedDict
 from .ocrpagedimensions import OCRPageDimensions, OCRPageDimensionsTypedDict
-from mistralai_azure.types import BaseModel, Nullable, UNSET_SENTINEL
+from .ocrtableobject import OCRTableObject, OCRTableObjectTypedDict
+from mistralai_azure.types import (
+    BaseModel,
+    Nullable,
+    OptionalNullable,
+    UNSET,
+    UNSET_SENTINEL,
+)
 from pydantic import model_serializer
-from typing import List
-from typing_extensions import TypedDict
+from typing import List, Optional
+from typing_extensions import NotRequired, TypedDict
 
 
 class OCRPageObjectTypedDict(TypedDict):
@@ -18,6 +25,14 @@ class OCRPageObjectTypedDict(TypedDict):
     r"""List of all extracted images in the page"""
     dimensions: Nullable[OCRPageDimensionsTypedDict]
     r"""The dimensions of the PDF Page's screenshot image"""
+    tables: NotRequired[List[OCRTableObjectTypedDict]]
+    r"""List of all extracted tables in the page"""
+    hyperlinks: NotRequired[List[str]]
+    r"""List of all hyperlinks in the page"""
+    header: NotRequired[Nullable[str]]
+    r"""Header of the page"""
+    footer: NotRequired[Nullable[str]]
+    r"""Footer of the page"""
 
 
 class OCRPageObject(BaseModel):
@@ -33,10 +48,22 @@ class OCRPageObject(BaseModel):
     dimensions: Nullable[OCRPageDimensions]
     r"""The dimensions of the PDF Page's screenshot image"""
 
+    tables: Optional[List[OCRTableObject]] = None
+    r"""List of all extracted tables in the page"""
+
+    hyperlinks: Optional[List[str]] = None
+    r"""List of all hyperlinks in the page"""
+
+    header: OptionalNullable[str] = UNSET
+    r"""Header of the page"""
+
+    footer: OptionalNullable[str] = UNSET
+    r"""Footer of the page"""
+
     @model_serializer(mode="wrap")
     def serialize_model(self, handler):
-        optional_fields = []
-        nullable_fields = ["dimensions"]
+        optional_fields = ["tables", "hyperlinks", "header", "footer"]
+        nullable_fields = ["header", "footer", "dimensions"]
         null_default_fields = []
 
         serialized = handler(self)

mistralai_azure/models/ocrrequest.py

@@ -13,7 +13,7 @@ from mistralai_azure.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import List, Optional, Union
+from typing import List, Literal, Optional, Union
 from typing_extensions import NotRequired, TypeAliasType, TypedDict
 
 
@@ -28,6 +28,12 @@ Document = TypeAliasType("Document", Union[FileChunk, ImageURLChunk, DocumentURL
 r"""Document to run OCR on"""
 
 
+TableFormat = Literal[
+    "markdown",
+    "html",
+]
+
+
 class OCRRequestTypedDict(TypedDict):
     model: Nullable[str]
     document: DocumentTypedDict
@@ -45,6 +51,9 @@ class OCRRequestTypedDict(TypedDict):
     r"""Structured output class for extracting useful information from each extracted bounding box / image from document. Only json_schema is valid for this field"""
     document_annotation_format: NotRequired[Nullable[ResponseFormatTypedDict]]
     r"""Structured output class for extracting useful information from the entire document. Only json_schema is valid for this field"""
+    table_format: NotRequired[Nullable[TableFormat]]
+    extract_header: NotRequired[bool]
+    extract_footer: NotRequired[bool]
 
 
 class OCRRequest(BaseModel):
@@ -73,6 +82,12 @@ class OCRRequest(BaseModel):
     document_annotation_format: OptionalNullable[ResponseFormat] = UNSET
     r"""Structured output class for extracting useful information from the entire document. Only json_schema is valid for this field"""
 
+    table_format: OptionalNullable[TableFormat] = UNSET
+
+    extract_header: Optional[bool] = None
+
+    extract_footer: Optional[bool] = None
+
     @model_serializer(mode="wrap")
     def serialize_model(self, handler):
         optional_fields = [
@@ -83,6 +98,9 @@ class OCRRequest(BaseModel):
             "image_min_size",
             "bbox_annotation_format",
             "document_annotation_format",
+            "table_format",
+            "extract_header",
+            "extract_footer",
         ]
         nullable_fields = [
             "model",
@@ -92,6 +110,7 @@ class OCRRequest(BaseModel):
             "image_min_size",
             "bbox_annotation_format",
             "document_annotation_format",
+            "table_format",
         ]
         null_default_fields = []
 

mistralai_azure/models/ocrtableobject.py

@@ -0,0 +1,34 @@
+"""Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
+
+from __future__ import annotations
+from mistralai_azure.types import BaseModel
+import pydantic
+from typing import Literal
+from typing_extensions import Annotated, TypedDict
+
+
+Format = Literal[
+    "markdown",
+    "html",
+]
+r"""Format of the table"""
+
+
+class OCRTableObjectTypedDict(TypedDict):
+    id: str
+    r"""Table ID for extracted table in a page"""
+    content: str
+    r"""Content of the table in the given format"""
+    format_: Format
+    r"""Format of the table"""
+
+
+class OCRTableObject(BaseModel):
+    id: str
+    r"""Table ID for extracted table in a page"""
+
+    content: str
+    r"""Content of the table in the given format"""
+
+    format_: Annotated[Format, pydantic.Field(alias="format")]
+    r"""Format of the table"""

mistralai_azure/models/referencechunk.py

@@ -6,7 +6,7 @@ from typing import List, Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-ReferenceChunkType = Literal["reference"]
+ReferenceChunkType = Literal["reference",]
 
 
 class ReferenceChunkTypedDict(TypedDict):

mistralai_azure/models/responseformats.py

@@ -4,4 +4,8 @@ from __future__ import annotations
 from typing import Literal
 
 
-ResponseFormats = Literal["text", "json_object", "json_schema"]
+ResponseFormats = Literal[
+    "text",
+    "json_object",
+    "json_schema",
+]

mistralai_azure/models/responsevalidationerror.py

@@ -2,10 +2,12 @@
 
 import httpx
 from typing import Optional
+from dataclasses import dataclass
 
 from mistralai_azure.models import MistralAzureError
 
 
+@dataclass(unsafe_hash=True)
 class ResponseValidationError(MistralAzureError):
     """Error raised when there is a type mismatch between the response data and the expected Pydantic model."""
 

mistralai_azure/models/sdkerror.py

@@ -2,12 +2,14 @@
 
 import httpx
 from typing import Optional
+from dataclasses import dataclass
 
 from mistralai_azure.models import MistralAzureError
 
 MAX_MESSAGE_LEN = 10_000
 
 
+@dataclass(unsafe_hash=True)
 class SDKError(MistralAzureError):
     """The fallback error class if no more specific error class is matched."""
 

mistralai_azure/models/systemmessage.py

@@ -21,7 +21,7 @@ SystemMessageContent = TypeAliasType(
 )
 
 
-Role = Literal["system"]
+Role = Literal["system",]
 
 
 class SystemMessageTypedDict(TypedDict):

mistralai_azure/models/textchunk.py

@@ -6,7 +6,7 @@ from typing import Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-Type = Literal["text"]
+Type = Literal["text",]
 
 
 class TextChunkTypedDict(TypedDict):

mistralai_azure/models/thinkchunk.py

@@ -16,7 +16,7 @@ ThinkingTypedDict = TypeAliasType(
 Thinking = TypeAliasType("Thinking", Union[ReferenceChunk, TextChunk])
 
 
-ThinkChunkType = Literal["thinking"]
+ThinkChunkType = Literal["thinking",]
 
 
 class ThinkChunkTypedDict(TypedDict):

mistralai_azure/models/tool.py

@@ -4,10 +4,8 @@ from __future__ import annotations
 from .function import Function, FunctionTypedDict
 from .tooltypes import ToolTypes
 from mistralai_azure.types import BaseModel
-from mistralai_azure.utils import validate_open_enum
-from pydantic.functional_validators import PlainValidator
 from typing import Optional
-from typing_extensions import Annotated, NotRequired, TypedDict
+from typing_extensions import NotRequired, TypedDict
 
 
 class ToolTypedDict(TypedDict):
@@ -18,6 +16,4 @@ class ToolTypedDict(TypedDict):
 class Tool(BaseModel):
     function: Function
 
-    type: Annotated[Optional[ToolTypes], PlainValidator(validate_open_enum(False))] = (
-        None
-    )
+    type: Optional[ToolTypes] = None

mistralai_azure/models/toolcall.py

@@ -4,10 +4,8 @@ from __future__ import annotations
 from .functioncall import FunctionCall, FunctionCallTypedDict
 from .tooltypes import ToolTypes
 from mistralai_azure.types import BaseModel
-from mistralai_azure.utils import validate_open_enum
-from pydantic.functional_validators import PlainValidator
 from typing import Optional
-from typing_extensions import Annotated, NotRequired, TypedDict
+from typing_extensions import NotRequired, TypedDict
 
 
 class ToolCallTypedDict(TypedDict):
@@ -22,8 +20,6 @@ class ToolCall(BaseModel):
 
     id: Optional[str] = "null"
 
-    type: Annotated[Optional[ToolTypes], PlainValidator(validate_open_enum(False))] = (
-        None
-    )
+    type: Optional[ToolTypes] = None
 
     index: Optional[int] = 0

mistralai_azure/models/toolchoice.py

@@ -4,10 +4,8 @@ from __future__ import annotations
 from .functionname import FunctionName, FunctionNameTypedDict
 from .tooltypes import ToolTypes
 from mistralai_azure.types import BaseModel
-from mistralai_azure.utils import validate_open_enum
-from pydantic.functional_validators import PlainValidator
 from typing import Optional
-from typing_extensions import Annotated, NotRequired, TypedDict
+from typing_extensions import NotRequired, TypedDict
 
 
 class ToolChoiceTypedDict(TypedDict):
@@ -24,6 +22,4 @@ class ToolChoice(BaseModel):
     function: FunctionName
     r"""this restriction of `Function` is used to select a specific function to call"""
 
-    type: Annotated[Optional[ToolTypes], PlainValidator(validate_open_enum(False))] = (
-        None
-    )
+    type: Optional[ToolTypes] = None

mistralai_azure/models/toolchoiceenum.py

@@ -4,4 +4,9 @@ from __future__ import annotations
 from typing import Literal
 
 
-ToolChoiceEnum = Literal["auto", "none", "any", "required"]
+ToolChoiceEnum = Literal[
+    "auto",
+    "none",
+    "any",
+    "required",
+]

mistralai_azure/models/toolmessage.py

@@ -22,7 +22,7 @@ ToolMessageContentTypedDict = TypeAliasType(
 ToolMessageContent = TypeAliasType("ToolMessageContent", Union[str, List[ContentChunk]])
 
 
-ToolMessageRole = Literal["tool"]
+ToolMessageRole = Literal["tool",]
 
 
 class ToolMessageTypedDict(TypedDict):

mistralai_azure/models/tooltypes.py

@@ -5,4 +5,4 @@ from mistralai_azure.types import UnrecognizedStr
 from typing import Literal, Union
 
 
-ToolTypes = Union[Literal["function"], UnrecognizedStr]
+ToolTypes = Union[Literal["function",], UnrecognizedStr]

mistralai_azure/models/usermessage.py

@@ -16,7 +16,7 @@ UserMessageContentTypedDict = TypeAliasType(
 UserMessageContent = TypeAliasType("UserMessageContent", Union[str, List[ContentChunk]])
 
 
-UserMessageRole = Literal["user"]
+UserMessageRole = Literal["user",]
 
 
 class UserMessageTypedDict(TypedDict):

mistralai_azure/ocr.py

@@ -25,11 +25,14 @@ class Ocr(BaseSDK):
         document_annotation_format: OptionalNullable[
             Union[models.ResponseFormat, models.ResponseFormatTypedDict]
         ] = UNSET,
+        table_format: OptionalNullable[models.TableFormat] = UNSET,
+        extract_header: Optional[bool] = None,
+        extract_footer: Optional[bool] = None,
         retries: OptionalNullable[utils.RetryConfig] = UNSET,
         server_url: Optional[str] = None,
         timeout_ms: Optional[int] = None,
         http_headers: Optional[Mapping[str, str]] = None,
-    ) -> Optional[models.OCRResponse]:
+    ) -> models.OCRResponse:
         r"""OCR
 
         :param model:
@@ -41,6 +44,9 @@ class Ocr(BaseSDK):
         :param image_min_size: Minimum height and width of image to extract
         :param bbox_annotation_format: Structured output class for extracting useful information from each extracted bounding box / image from document. Only json_schema is valid for this field
         :param document_annotation_format: Structured output class for extracting useful information from the entire document. Only json_schema is valid for this field
+        :param table_format:
+        :param extract_header:
+        :param extract_footer:
         :param retries: Override the default retry configuration for this method
         :param server_url: Override the default server URL for this method
         :param timeout_ms: Override the default request timeout configuration for this method in milliseconds
@@ -70,6 +76,9 @@ class Ocr(BaseSDK):
             document_annotation_format=utils.get_pydantic_model(
                 document_annotation_format, OptionalNullable[models.ResponseFormat]
             ),
+            table_format=table_format,
+            extract_header=extract_header,
+            extract_footer=extract_footer,
         )
 
         req = self._build_request(
@@ -88,6 +97,7 @@ class Ocr(BaseSDK):
             get_serialized_body=lambda: utils.serialize_request_body(
                 request, False, False, "json", models.OCRRequest
             ),
+            allow_empty_value=None,
             timeout_ms=timeout_ms,
         )
 
@@ -104,7 +114,7 @@ class Ocr(BaseSDK):
                 config=self.sdk_configuration,
                 base_url=base_url or "",
                 operation_id="ocr_v1_ocr_post",
-                oauth2_scopes=[],
+                oauth2_scopes=None,
                 security_source=self.sdk_configuration.security,
             ),
             request=req,
@@ -114,7 +124,7 @@ class Ocr(BaseSDK):
 
         response_data: Any = None
         if utils.match_response(http_res, "200", "application/json"):
-            return unmarshal_json_response(Optional[models.OCRResponse], http_res)
+            return unmarshal_json_response(models.OCRResponse, http_res)
         if utils.match_response(http_res, "422", "application/json"):
             response_data = unmarshal_json_response(
                 models.HTTPValidationErrorData, http_res
@@ -145,11 +155,14 @@ class Ocr(BaseSDK):
         document_annotation_format: OptionalNullable[
             Union[models.ResponseFormat, models.ResponseFormatTypedDict]
         ] = UNSET,
+        table_format: OptionalNullable[models.TableFormat] = UNSET,
+        extract_header: Optional[bool] = None,
+        extract_footer: Optional[bool] = None,
         retries: OptionalNullable[utils.RetryConfig] = UNSET,
         server_url: Optional[str] = None,
         timeout_ms: Optional[int] = None,
         http_headers: Optional[Mapping[str, str]] = None,
-    ) -> Optional[models.OCRResponse]:
+    ) -> models.OCRResponse:
         r"""OCR
 
         :param model:
@@ -161,6 +174,9 @@ class Ocr(BaseSDK):
         :param image_min_size: Minimum height and width of image to extract
         :param bbox_annotation_format: Structured output class for extracting useful information from each extracted bounding box / image from document. Only json_schema is valid for this field
         :param document_annotation_format: Structured output class for extracting useful information from the entire document. Only json_schema is valid for this field
+        :param table_format:
+        :param extract_header:
+        :param extract_footer:
         :param retries: Override the default retry configuration for this method
         :param server_url: Override the default server URL for this method
         :param timeout_ms: Override the default request timeout configuration for this method in milliseconds
@@ -190,6 +206,9 @@ class Ocr(BaseSDK):
             document_annotation_format=utils.get_pydantic_model(
                 document_annotation_format, OptionalNullable[models.ResponseFormat]
             ),
+            table_format=table_format,
+            extract_header=extract_header,
+            extract_footer=extract_footer,
         )
 
         req = self._build_request_async(
@@ -208,6 +227,7 @@ class Ocr(BaseSDK):
             get_serialized_body=lambda: utils.serialize_request_body(
                 request, False, False, "json", models.OCRRequest
             ),
+            allow_empty_value=None,
             timeout_ms=timeout_ms,
         )
 
@@ -224,7 +244,7 @@ class Ocr(BaseSDK):
                 config=self.sdk_configuration,
                 base_url=base_url or "",
                 operation_id="ocr_v1_ocr_post",
-                oauth2_scopes=[],
+                oauth2_scopes=None,
                 security_source=self.sdk_configuration.security,
             ),
             request=req,
@@ -234,7 +254,7 @@ class Ocr(BaseSDK):
 
         response_data: Any = None
         if utils.match_response(http_res, "200", "application/json"):
-            return unmarshal_json_response(Optional[models.OCRResponse], http_res)
+            return unmarshal_json_response(models.OCRResponse, http_res)
         if utils.match_response(http_res, "422", "application/json"):
             response_data = unmarshal_json_response(
                 models.HTTPValidationErrorData, http_res

mistralai_azure/types/basemodel.py

@@ -2,7 +2,8 @@
 
 from pydantic import ConfigDict, model_serializer
 from pydantic import BaseModel as PydanticBaseModel
-from typing import TYPE_CHECKING, Literal, Optional, TypeVar, Union
+from pydantic_core import core_schema
+from typing import TYPE_CHECKING, Any, Literal, Optional, TypeVar, Union
 from typing_extensions import TypeAliasType, TypeAlias
 
 
@@ -35,5 +36,42 @@ else:
         "OptionalNullable", Union[Optional[Nullable[T]], Unset], type_params=(T,)
     )
 
-UnrecognizedInt: TypeAlias = int
-UnrecognizedStr: TypeAlias = str
+
+class UnrecognizedStr(str):
+    @classmethod
+    def __get_pydantic_core_schema__(cls, _source_type: Any, _handler: Any) -> core_schema.CoreSchema:
+        # Make UnrecognizedStr only work in lax mode, not strict mode
+        # This makes it a "fallback" option when more specific types (like Literals) don't match
+        def validate_lax(v: Any) -> 'UnrecognizedStr':
+            if isinstance(v, cls):
+                return v
+            return cls(str(v))
+
+        # Use lax_or_strict_schema where strict always fails
+        # This forces Pydantic to prefer other union members in strict mode
+        # and only fall back to UnrecognizedStr in lax mode
+        return core_schema.lax_or_strict_schema(
+            lax_schema=core_schema.chain_schema([
+                core_schema.str_schema(),
+                core_schema.no_info_plain_validator_function(validate_lax)
+            ]),
+            strict_schema=core_schema.none_schema(),  # Always fails in strict mode
+        )
+
+
+class UnrecognizedInt(int):
+    @classmethod
+    def __get_pydantic_core_schema__(cls, _source_type: Any, _handler: Any) -> core_schema.CoreSchema:
+        # Make UnrecognizedInt only work in lax mode, not strict mode
+        # This makes it a "fallback" option when more specific types (like Literals) don't match
+        def validate_lax(v: Any) -> 'UnrecognizedInt':
+            if isinstance(v, cls):
+                return v
+            return cls(int(v))
+        return core_schema.lax_or_strict_schema(
+            lax_schema=core_schema.chain_schema([
+                core_schema.int_schema(),
+                core_schema.no_info_plain_validator_function(validate_lax)
+            ]),
+            strict_schema=core_schema.none_schema(),  # Always fails in strict mode
+        )

mistralai_azure/utils/__init__.py

@@ -41,7 +41,6 @@ if TYPE_CHECKING:
         validate_decimal,
         validate_float,
         validate_int,
-        validate_open_enum,
     )
     from .url import generate_url, template_url, remove_suffix
     from .values import (
@@ -102,7 +101,6 @@ __all__ = [
     "validate_const",
     "validate_float",
     "validate_int",
-    "validate_open_enum",
     "cast_partial",
 ]
 
@@ -155,7 +153,6 @@ _dynamic_imports: dict[str, str] = {
     "validate_const": ".serializers",
     "validate_float": ".serializers",
     "validate_int": ".serializers",
-    "validate_open_enum": ".serializers",
     "cast_partial": ".values",
 }
 

mistralai_azure/utils/annotations.py

@@ -3,6 +3,7 @@
 from enum import Enum
 from typing import Any, Optional
 
+
 def get_discriminator(model: Any, fieldname: str, key: str) -> str:
     """
     Recursively search for the discriminator attribute in a model.
@@ -25,31 +26,54 @@ def get_discriminator(model: Any, fieldname: str, key: str) -> str:
 
         if isinstance(field, dict):
             if key in field:
-                return f'{field[key]}'
+                return f"{field[key]}"
 
         if hasattr(field, fieldname):
             attr = getattr(field, fieldname)
             if isinstance(attr, Enum):
-                return f'{attr.value}'
-            return f'{attr}'
+                return f"{attr.value}"
+            return f"{attr}"
 
         if hasattr(field, upper_fieldname):
             attr = getattr(field, upper_fieldname)
             if isinstance(attr, Enum):
-                return f'{attr.value}'
-            return f'{attr}'
+                return f"{attr.value}"
+            return f"{attr}"
 
         return None
 
+    def search_nested_discriminator(obj: Any) -> Optional[str]:
+        """Recursively search for discriminator in nested structures."""
+        # First try direct field lookup
+        discriminator = get_field_discriminator(obj)
+        if discriminator is not None:
+            return discriminator
+
+        # If it's a dict, search in nested values
+        if isinstance(obj, dict):
+            for value in obj.values():
+                if isinstance(value, list):
+                    # Search in list items
+                    for item in value:
+                        nested_discriminator = search_nested_discriminator(item)
+                        if nested_discriminator is not None:
+                            return nested_discriminator
+                elif isinstance(value, dict):
+                    # Search in nested dict
+                    nested_discriminator = search_nested_discriminator(value)
+                    if nested_discriminator is not None:
+                        return nested_discriminator
+
+        return None
 
     if isinstance(model, list):
         for field in model:
-            discriminator = get_field_discriminator(field)
+            discriminator = search_nested_discriminator(field)
             if discriminator is not None:
                 return discriminator
 
-    discriminator = get_field_discriminator(model)
+    discriminator = search_nested_discriminator(model)
     if discriminator is not None:
         return discriminator
 
-    raise ValueError(f'Could not find discriminator field {fieldname} in {model}')
+    raise ValueError(f"Could not find discriminator field {fieldname} in {model}")