chunkr-ai 0.0.22__tar.gz → 0.0.23__tar.gz

{chunkr_ai-0.0.22/src/chunkr_ai.egg-info → chunkr_ai-0.0.23}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: chunkr-ai
-Version: 0.0.22
+Version: 0.0.23
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "chunkr-ai"
-version = "0.0.22"
+version = "0.0.23"
 authors = [{"name" = "Ishaan Kapoor", "email" = "ishaan@lumina.sh"}]
 description = "Python client for Chunkr: open source document intelligence"
 readme = "README.md"

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23}/src/chunkr_ai/api/chunkr.py

@@ -34,7 +34,7 @@ class Chunkr(ChunkrBase):
         file: Union[str, Path, BinaryIO, Image.Image],
         config: Configuration = None,
     ) -> TaskResponse:
-        files = prepare_upload_data(file, config)
+        files = await prepare_upload_data(file, config, self._client)
         r = await self._client.post(
             f"{self.url}/api/v1/task", files=files, headers=self._headers()
         )
@@ -44,7 +44,7 @@ class Chunkr(ChunkrBase):
     @anywhere()
     @ensure_client()
     async def update_task(self, task_id: str, config: Configuration) -> TaskResponse:
-        files = prepare_upload_data(None, config)
+        files = await prepare_upload_data(None, config, self._client)
         r = await self._client.patch(
             f"{self.url}/api/v1/task/{task_id}",
             files=files,

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23}/src/chunkr_ai/api/chunkr_base.py

@@ -4,13 +4,10 @@ from .auth import HeadersMixin
 from abc import abstractmethod
 from dotenv import load_dotenv
 import httpx
-import io
-import json
 import os
 from pathlib import Path
 from PIL import Image
-import requests
-from typing import BinaryIO, Tuple, Union
+from typing import BinaryIO, Union
 
 
 class ChunkrBase(HeadersMixin):
@@ -28,130 +25,6 @@ class ChunkrBase(HeadersMixin):
         self.url = self.url.rstrip("/")
         self._client = httpx.AsyncClient()
 
-    def _prepare_file(
-        self, file: Union[str, Path, BinaryIO, Image.Image]
-    ) -> Tuple[str, BinaryIO]:
-        """Convert various file types into a tuple of (filename, file-like object).
-
-        Args:
-            file: Input file, can be:
-                - String or Path to a file
-                - URL string starting with http:// or https://
-                - Base64 string
-                - Opened binary file (mode='rb')
-                - PIL/Pillow Image object
-
-        Returns:
-            Tuple[str, BinaryIO]: (filename, file-like object) ready for upload
-
-        Raises:
-            FileNotFoundError: If the file path doesn't exist
-            TypeError: If the file type is not supported
-            ValueError: If the URL is invalid or unreachable
-            ValueError: If the MIME type is unsupported
-        """
-        # Handle URLs
-        if isinstance(file, str) and (
-            file.startswith("http://") or file.startswith("https://")
-        ):
-            response = requests.get(file)
-            response.raise_for_status()
-            file_obj = io.BytesIO(response.content)
-            filename = Path(file.split("/")[-1]).name or "downloaded_file"
-            return filename, file_obj
-
-        # Handle base64 strings
-        if isinstance(file, str) and "," in file and ";base64," in file:
-            try:
-                # Split header and data
-                header, base64_data = file.split(",", 1)
-                import base64
-
-                file_bytes = base64.b64decode(base64_data)
-                file_obj = io.BytesIO(file_bytes)
-
-                # Try to determine format from header
-                format = "bin"
-                mime_type = header.split(":")[-1].split(";")[0].lower()
-
-                # Map MIME types to file extensions
-                mime_to_ext = {
-                    "application/pdf": "pdf",
-                    "application/vnd.openxmlformats-officedocument.wordprocessingml.document": "docx",
-                    "application/msword": "doc",
-                    "application/vnd.openxmlformats-officedocument.presentationml.presentation": "pptx",
-                    "application/vnd.ms-powerpoint": "ppt",
-                    "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet": "xlsx",
-                    "application/vnd.ms-excel": "xls",
-                    "image/jpeg": "jpg",
-                    "image/png": "png",
-                    "image/jpg": "jpg",
-                }
-
-                if mime_type in mime_to_ext:
-                    format = mime_to_ext[mime_type]
-                else:
-                    raise ValueError(f"Unsupported MIME type: {mime_type}")
-
-                return f"file.{format}", file_obj
-            except Exception as e:
-                raise ValueError(f"Invalid base64 string: {str(e)}")
-
-        # Handle file paths
-        if isinstance(file, (str, Path)):
-            path = Path(file).resolve()
-            if not path.exists():
-                raise FileNotFoundError(f"File not found: {file}")
-            return path.name, open(path, "rb")
-
-        # Handle PIL Images
-        if isinstance(file, Image.Image):
-            img_byte_arr = io.BytesIO()
-            format = file.format or "PNG"
-            file.save(img_byte_arr, format=format)
-            img_byte_arr.seek(0)
-            return f"image.{format.lower()}", img_byte_arr
-
-        # Handle file-like objects
-        if hasattr(file, "read") and hasattr(file, "seek"):
-            # Try to get the filename from the file object if possible
-            name = (
-                getattr(file, "name", "document")
-                if hasattr(file, "name")
-                else "document"
-            )
-            return Path(name).name, file
-
-        raise TypeError(f"Unsupported file type: {type(file)}")
-
-    def _prepare_upload_data(
-        self,
-        file: Union[str, Path, BinaryIO, Image.Image],
-        config: Configuration = None,
-    ) -> Tuple[dict, dict]:
-        """Prepare files and data dictionaries for upload.
-
-        Args:
-            file: The file to upload
-            config: Optional configuration settings
-
-        Returns:
-            Tuple[dict, dict]: (files dict, data dict) ready for upload
-        """
-        filename, file_obj = self._prepare_file(file)
-        files = {"file": (filename, file_obj)}
-        data = {}
-
-        if config:
-            config_dict = config.model_dump(mode="json", exclude_none=True)
-            for key, value in config_dict.items():
-                if isinstance(value, dict):
-                    files[key] = (None, json.dumps(value), "application/json")
-                else:
-                    data[key] = value
-
-        return files, data
-
     @abstractmethod
     def upload(
         self,

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23}/src/chunkr_ai/api/misc.py

@@ -3,16 +3,36 @@ import io
 import json
 from pathlib import Path
 from PIL import Image
-import requests
+import httpx
 from typing import Union, Tuple, BinaryIO, Optional
 
-def prepare_file(file: Union[str, Path, BinaryIO, Image.Image]) -> Tuple[str, BinaryIO]:
-    """Convert various file types into a tuple of (filename, file-like object)."""
+async def prepare_file(file: Union[str, Path, BinaryIO, Image.Image], client: httpx.AsyncClient = None) -> Tuple[str, BinaryIO]:
+    """Convert various file types into a tuple of (filename, file-like object).
+
+        Args:
+            file: Input file, can be:
+                - String or Path to a file
+                - URL string starting with http:// or https://
+                - Base64 string
+                - Opened binary file (mode='rb')
+                - PIL/Pillow Image object
+
+        Returns:
+            Tuple[str, BinaryIO]: (filename, file-like object) ready for upload
+
+        Raises:
+            FileNotFoundError: If the file path doesn't exist
+            TypeError: If the file type is not supported
+            ValueError: If the URL is invalid or unreachable
+            ValueError: If the MIME type is unsupported
+    """  
     # Handle URLs
     if isinstance(file, str) and (
         file.startswith("http://") or file.startswith("https://")
     ):
-        response = requests.get(file)
+        if not client:
+            raise ValueError("Client must be provided to download files from URLs")
+        response = client.get(file)
         response.raise_for_status()
 
         # Try to get filename from Content-Disposition header first
@@ -108,9 +128,10 @@ def prepare_file(file: Union[str, Path, BinaryIO, Image.Image]) -> Tuple[str, Bi
     raise TypeError(f"Unsupported file type: {type(file)}")
 
 
-def prepare_upload_data(
+async def prepare_upload_data(
     file: Optional[Union[str, Path, BinaryIO, Image.Image]] = None,
     config: Optional[Configuration] = None,
+    client: httpx.AsyncClient = None,
 ) -> dict:
     """Prepare files and data dictionaries for upload.
 
@@ -123,7 +144,7 @@ def prepare_upload_data(
     """
     files = {}
     if file:
-        filename, file_obj = prepare_file(file)
+        filename, file_obj = await prepare_file(file, client)
         files = {"file": (filename, file_obj)}
 
     if config:

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23}/src/chunkr_ai/api/task_response.py

@@ -35,10 +35,9 @@ class TaskResponse(BaseModel, Generic[T]):
             return self
         return None
 
+    @require_task()
     async def _poll_request(self) -> dict:
         try:
-            if not self._client._client:
-                raise ValueError("Client not found")
             r = await self._client._client.get(
                 self.task_url, headers=self._client._headers()
             )
@@ -51,7 +50,6 @@ class TaskResponse(BaseModel, Generic[T]):
             raise
 
     @anywhere()
-    @require_task()
     async def poll(self) -> T:
         """Poll the task for completion."""
         while True:
@@ -66,7 +64,7 @@ class TaskResponse(BaseModel, Generic[T]):
     @require_task()
     async def update(self, config: Configuration) -> T:
         """Update the task configuration."""
-        f = prepare_upload_data(None, config)
+        f = await prepare_upload_data(None, config, self._client._client)
         r = await self._client._client.patch(
             self.task_url, files=f, headers=self._client._headers()
         )

{chunkr_ai-0.0.22 → chunkr_ai-0.0.23/src/chunkr_ai.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: chunkr-ai
-Version: 0.0.22
+Version: 0.0.23
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License