chunkr-ai 0.3.1__tar.gz → 0.3.3__tar.gz

{chunkr_ai-0.3.1/src/chunkr_ai.egg-info → chunkr_ai-0.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: chunkr-ai
-Version: 0.3.1
+Version: 0.3.3
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/pyproject.toml

@@ -4,12 +4,12 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "chunkr-ai"
-version = "0.3.1"
-authors = [{"name" = "Ishaan Kapoor", "email" = "ishaan@lumina.sh"}]
+version = "0.3.3"
+authors = [{ "name" = "Ishaan Kapoor", "email" = "ishaan@lumina.sh" }]
 description = "Python client for Chunkr: open source document intelligence"
 readme = "README.md"
-license = {"file" = "LICENSE"}
-urls = {Homepage = "https://chunkr.ai"} 
+license = { "file" = "LICENSE" }
+urls = { Homepage = "https://chunkr.ai" }
 dependencies = [
     "httpx>=0.25.0",
     "matplotlib>=3.10.3",
@@ -27,3 +27,5 @@ test = [
     "ruff>=0.9.3",
 ]
 
+[dependency-groups]
+dev = ["mypy>=1.17.1"]

chunkr_ai-0.3.3/src/chunkr_ai/__init__.py

@@ -0,0 +1,14 @@
+from .api.chunkr import Chunkr
+import tomllib
+from pathlib import Path
+
+# Read version from pyproject.toml
+try:
+    pyproject_path = Path(__file__).parent.parent.parent / "pyproject.toml"
+    with open(pyproject_path, "rb") as f:
+        pyproject_data = tomllib.load(f)
+    __version__ = pyproject_data["project"]["version"]
+except Exception:
+    __version__ = "unknown"
+
+__all__ = ["Chunkr", "__version__"]

chunkr_ai-0.3.3/src/chunkr_ai/api/auth.py

@@ -0,0 +1,42 @@
+import platform
+import sys
+import tomllib
+from pathlib import Path
+
+def _find_pyproject_toml(start_path: Path) -> Path | None:
+    """Search for pyproject.toml in current and parent directories."""
+    for parent in [start_path, *start_path.parents]:
+        candidate = parent / "pyproject.toml"
+        if candidate.is_file():
+            return candidate
+    return None
+
+# Read version from pyproject.toml
+try:
+    pyproject_path = _find_pyproject_toml(Path(__file__).resolve().parent)
+    if pyproject_path is not None:
+        with open(pyproject_path, "rb") as f:
+            pyproject_data = tomllib.load(f)
+        __version__ = pyproject_data["project"]["version"]
+    else:
+        __version__ = "unknown"
+except Exception:
+    __version__ = "unknown"
+
+class HeadersMixin:
+    """Mixin class for handling authorization headers"""
+    _api_key: str = ""
+
+    def get_api_key(self) -> str:
+        """Get the API key"""
+        if not hasattr(self, "_api_key") or not self._api_key:
+            raise ValueError("API key not set")
+        return self._api_key
+
+    def _headers(self) -> dict:
+        """Generate authorization headers and version information"""
+        user_agent = f"chunkr-ai/{__version__} (Python/{sys.version.split()[0]}; {platform.system()}/{platform.release()})"
+        return {
+            "Authorization": self.get_api_key(),
+            "User-Agent": user_agent
+        }

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/src/chunkr_ai/api/chunkr.py

@@ -42,7 +42,7 @@ class Chunkr(ChunkrBase):
         data = await prepare_upload_data(file, filename, config)
         assert self._client is not None
         r = await self._client.post(
-            f"{self.url}/api/v1/task/parse", json=data, headers=self._headers()
+            f"{self.url}/task/parse", json=data, headers=self._headers()
         )
         r.raise_for_status()
         return TaskResponse(**r.json()).with_client(cast(ChunkrClientProtocol, self), True, False)
@@ -55,7 +55,7 @@ class Chunkr(ChunkrBase):
         data = await prepare_upload_data(None, None, config)
         assert self._client is not None
         r = await self._client.patch(
-            f"{self.url}/api/v1/task/{task_id}/parse",
+            f"{self.url}/task/{task_id}/parse",
             json=data,
             headers=self._headers(),
         )
@@ -71,7 +71,7 @@ class Chunkr(ChunkrBase):
         }
         assert self._client is not None
         r = await self._client.get(
-            f"{self.url}/api/v1/task/{task_id}",
+            f"{self.url}/task/{task_id}",
             params=params,
             headers=self._headers()
         )
@@ -83,7 +83,7 @@ class Chunkr(ChunkrBase):
     async def delete_task(self, task_id: str) -> None:
         assert self._client is not None
         r = await self._client.delete(
-            f"{self.url}/api/v1/task/{task_id}", headers=self._headers()
+            f"{self.url}/task/{task_id}", headers=self._headers()
         )
         r.raise_for_status()
 
@@ -92,7 +92,7 @@ class Chunkr(ChunkrBase):
     async def cancel_task(self, task_id: str) -> None:
         assert self._client is not None
         r = await self._client.get(
-            f"{self.url}/api/v1/task/{task_id}/cancel", headers=self._headers()
+            f"{self.url}/task/{task_id}/cancel", headers=self._headers()
         )
         r.raise_for_status()
 

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/src/chunkr_ai/api/configuration.py

@@ -1,6 +1,6 @@
 from pydantic import BaseModel, Field, ConfigDict
 from enum import Enum
-from typing import Any, List, Optional, Union
+from typing import List, Optional, Union
 from pydantic import field_validator, field_serializer
 
 class CroppingStrategy(str, Enum):
@@ -20,15 +20,17 @@ class EmbedSource(str, Enum):
 class GenerationStrategy(str, Enum):
     LLM = "LLM"
     AUTO = "Auto"
+    IGNORE = "Ignore"
 
 class GenerationConfig(BaseModel):
     format: Optional[SegmentFormat] = None
     strategy: Optional[GenerationStrategy] = None
-    llm: Optional[str] = None
     crop_image: Optional[CroppingStrategy] = None
-    embed_sources: Optional[List[EmbedSource]] = None
     extended_context: Optional[bool] = None
+    description: Optional[bool] = None
     # Deprecated fields for backwards compatibility
+    llm: Optional[str] = None # Deprecated
+    embed_sources: Optional[List[EmbedSource]] = None # Deprecated
     html: Optional[GenerationStrategy] = None  # Deprecated: Use format=SegmentFormat.HTML and strategy instead
     markdown: Optional[GenerationStrategy] = None  # Deprecated: Use format=SegmentFormat.MARKDOWN and strategy instead
 
@@ -83,7 +85,7 @@ class TokenizerType(BaseModel):
         return {}
 
 class ChunkProcessing(BaseModel):
-    ignore_headers_and_footers: Optional[bool] = True
+    ignore_headers_and_footers: Optional[bool] = None # Deprecated
     target_length: Optional[int] = None
     tokenizer: Optional[Union[TokenizerType, Tokenizer, str]] = None
 
@@ -286,6 +288,7 @@ class Page(BaseModel):
     page_height: float
     page_width: float
     ss_sheet_name: Optional[str] = None
+    dpi: Optional[float] = None
 
 class Segment(BaseModel):
     bbox: BoundingBox
@@ -303,6 +306,8 @@ class Segment(BaseModel):
     confidence: Optional[float]
     text: str = ""
     segment_length: Optional[int] = None
+    embed: Optional[str] = None
+    description: Optional[str] = None
     # Spreadsheet-specific fields
     ss_cells: Optional[List[Cell]] = None
     ss_header_bbox: Optional[BoundingBox] = None
@@ -317,6 +322,7 @@ class Chunk(BaseModel):
     chunk_length: int
     segments: List[Segment]
     embed: Optional[str] = None
+    content: Optional[str] = None
 
 class OutputResponse(BaseModel):
     chunks: List[Chunk]
@@ -347,10 +353,6 @@ class Configuration(BaseModel):
     
 class OutputConfiguration(Configuration):
     input_file_url: Optional[str] = None
-    # Deprecated
-    json_schema: Optional[Any] = None
-    model: Optional[Model] = None
-    target_chunk_length: Optional[int] = None
     
 class Status(str, Enum):
     STARTING = "Starting"

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/src/chunkr_ai/api/decorators.py

@@ -2,10 +2,12 @@ import asyncio
 import functools
 import httpx
 import nest_asyncio
-from typing import Callable, Any, TypeVar, Awaitable, Union, overload
-try:
+from typing import Callable, Any, TypeVar, Awaitable, Union
+import sys
+
+if sys.version_info >= (3, 10):
     from typing import ParamSpec
-except ImportError:
+else:
     from typing_extensions import ParamSpec
 
 T = TypeVar('T')

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/src/chunkr_ai/api/misc.py

@@ -3,7 +3,7 @@ import base64
 import io
 from pathlib import Path
 from PIL import Image
-from typing import Union, Tuple, BinaryIO, Optional, Any
+from typing import Union, Tuple, BinaryIO, Optional
 
 async def prepare_file(file: Union[str, Path, BinaryIO, Image.Image, bytes, bytearray, memoryview]) -> Tuple[Optional[str], str]:
     """Convert various file types into a tuple of (filename, file content).
@@ -39,7 +39,7 @@ async def prepare_file(file: Union[str, Path, BinaryIO, Image.Image, bytes, byte
             base64.b64decode(potential_base64)
             # If we get here, it was a valid base64 string in bytes form
             return None, potential_base64
-        except:
+        except Exception:
             # Not a base64 string in bytes form, encode it as base64
             base64_str = base64.b64encode(file_bytes).decode()
             return None, base64_str
@@ -66,14 +66,14 @@ async def prepare_file(file: Union[str, Path, BinaryIO, Image.Image, bytes, byte
                     # Just test if it's valid base64, don't store the result
                     base64.b64decode(file)
                     return None, file
-                except:
+                except Exception:
                     raise ValueError(f"File not found: {file} and it's not a valid base64 string")
         except Exception as e:
             # If string can't be converted to Path or decoded as base64, it might still be a base64 string
             try:
-                base64.b64decode(file)
-                return None, file
-            except:
+                base64.b64decode(str(file))
+                return None, str(file)
+            except Exception:
                 raise ValueError(f"Unable to process file: {e}")
 
     # Handle file paths - convert to base64

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/src/chunkr_ai/api/task_response.py

@@ -1,5 +1,5 @@
 from datetime import datetime
-from typing import Optional, cast, Awaitable, Union
+from typing import Optional, cast, Union
 from pydantic import BaseModel, PrivateAttr
 import asyncio
 import json

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3/src/chunkr_ai.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: chunkr-ai
-Version: 0.3.1
+Version: 0.3.3
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/tests/test_chunkr.py

@@ -680,6 +680,74 @@ class TestNewFields:
         for segment in segments_with_length:
             assert segment.segment_length > 0
 
+    @pytest.mark.asyncio
+    async def test_segment_embed_field(self, client, sample_path):
+        """Test that Segment objects include the embed field"""
+        response = await client.upload(sample_path)
+        assert response.task_id is not None
+        assert response.status == "Succeeded"
+        assert response.output is not None
+        assert len(response.output.chunks) > 0
+        
+        segment = response.output.chunks[0].segments[0]
+        
+        # embed field should be accessible
+        assert hasattr(segment, 'embed')
+        # embed can be None or a string
+        assert segment.embed is None or isinstance(segment.embed, str)
+
+    @pytest.mark.asyncio
+    async def test_segment_description_field(self, client, sample_path):
+        """Test that Segment objects include the description field"""
+        response = await client.upload(sample_path)
+        assert response.task_id is not None
+        assert response.status == "Succeeded"
+        assert response.output is not None
+        assert len(response.output.chunks) > 0
+        
+        segment = response.output.chunks[0].segments[0]
+        
+        # description field should be accessible
+        assert hasattr(segment, 'description')
+        # description can be None or a string
+        assert segment.description is None or isinstance(segment.description, str)
+
+    @pytest.mark.asyncio
+    async def test_chunk_content_field(self, client, sample_path):
+        """Test that Chunk objects include the content field"""
+        response = await client.upload(sample_path)
+        assert response.task_id is not None
+        assert response.status == "Succeeded"
+        assert response.output is not None
+        assert len(response.output.chunks) > 0
+        
+        chunk = response.output.chunks[0]
+        
+        # content field should be accessible
+        assert hasattr(chunk, 'content')
+        # content can be None or a string
+        assert chunk.content is None or isinstance(chunk.content, str)
+
+    @pytest.mark.asyncio
+    async def test_page_dpi_field_type(self, client, sample_path):
+        """Test that Page objects have dpi field as float type"""
+        response = await client.upload(sample_path)
+        assert response.task_id is not None
+        assert response.status == "Succeeded"
+        assert response.output is not None
+        
+        # pages should be accessible (might be None for some configurations)
+        assert hasattr(response.output, 'pages')
+        
+        # If pages exist, validate dpi type
+        if response.output.pages:
+            assert len(response.output.pages) > 0
+            page = response.output.pages[0]
+            assert hasattr(page, 'dpi')
+            # dpi can be None or a float/int (numbers)
+            if page.dpi is not None:
+                assert isinstance(page.dpi, (int, float))
+
     @pytest.mark.asyncio
     async def test_backwards_compatibility_preserved(self, client, sample_path):
         """Test that all existing fields still work after adding new ones"""
@@ -716,4 +784,10 @@ class TestNewFields:
         assert segment.segment_id is not None
         assert segment.segment_type is not None
         assert segment.confidence is not None or segment.confidence is None
-        assert segment.text is not None or segment.text == ""
+        assert segment.text is not None or segment.text == ""
+        # New fields added for server compatibility
+        assert hasattr(segment, 'embed')
+        assert hasattr(segment, 'description')
+        
+        # Chunk should include new content field
+        assert hasattr(chunk, 'content')

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/tests/test_excel.py

@@ -14,12 +14,6 @@ from chunkr_ai.models import (
     Tokenizer,
     OcrStrategy,
     SegmentationStrategy,
-    Cell,
-    CellStyle,
-    Alignment,
-    VerticalAlignment,
-    Page,
-    Segment,
     SegmentType,
 )
 
@@ -316,6 +310,17 @@ class TestExcelEmbedding:
         for chunk in chunks_with_embed:
             assert len(chunk.embed.strip()) > 0, "Empty embed content found"
 
+    @pytest.mark.asyncio
+    async def test_chunks_have_content_field(self, client, excel_sample_path, excel_config):
+        """Test that chunks include the new content field for Excel data"""
+        response = await client.upload(excel_sample_path, excel_config)
+        
+        # All chunks should have the content field accessible
+        for chunk in response.output.chunks:
+            assert hasattr(chunk, 'content'), "Chunk missing content field"
+            # content can be None or a string
+            assert chunk.content is None or isinstance(chunk.content, str)
+
     @pytest.mark.asyncio
     async def test_segment_length_calculation(self, client, excel_sample_path, excel_config):
         """Test that segments have length calculations"""

{chunkr_ai-0.3.1 → chunkr_ai-0.3.3}/tests/test_pages.py

@@ -4,16 +4,8 @@ from pathlib import Path
 from chunkr_ai import Chunkr
 from chunkr_ai.models import (
     Configuration,
-    ChunkProcessing,
-    SegmentProcessing,
-    GenerationConfig,
-    SegmentFormat,
-    GenerationStrategy,
-    EmbedSource,
-    Tokenizer,
     OcrStrategy,
     SegmentationStrategy,
-    Page,
 )
 
 

chunkr_ai-0.3.1/src/chunkr_ai/__init__.py

@@ -1,3 +0,0 @@
-from .api.chunkr import Chunkr
-
-__all__ = ["Chunkr"]

chunkr_ai-0.3.1/src/chunkr_ai/api/auth.py

@@ -1,13 +0,0 @@
-class HeadersMixin:
-    """Mixin class for handling authorization headers"""
-    _api_key: str = ""
-
-    def get_api_key(self) -> str:
-        """Get the API key"""
-        if not hasattr(self, "_api_key") or not self._api_key:
-            raise ValueError("API key not set")
-        return self._api_key
-
-    def _headers(self) -> dict:
-        """Generate authorization headers"""
-        return {"Authorization": self.get_api_key()}