chunkr-ai 0.0.49__tar.gz → 0.1.0__tar.gz

{chunkr_ai-0.0.49/src/chunkr_ai.egg-info → chunkr_ai-0.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: chunkr-ai
-Version: 0.0.49
+Version: 0.1.0
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License

{chunkr_ai-0.0.49 → chunkr_ai-0.1.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "chunkr-ai"
-version = "0.0.49"
+version = "0.1.0"
 authors = [{"name" = "Ishaan Kapoor", "email" = "ishaan@lumina.sh"}]
 description = "Python client for Chunkr: open source document intelligence"
 readme = "README.md"

{chunkr_ai-0.0.49 → chunkr_ai-0.1.0}/src/chunkr_ai/api/configuration.py

@@ -3,27 +3,34 @@ from enum import Enum
 from typing import Any, List, Optional, Union
 from pydantic import field_validator, field_serializer
 
-class GenerationStrategy(str, Enum):
-    LLM = "LLM"
-    AUTO = "Auto"
-
 class CroppingStrategy(str, Enum):
     ALL = "All"
     AUTO = "Auto"
 
-class EmbedSource(str, Enum):
-    HTML = "HTML"
+class SegmentFormat(str, Enum):
+    HTML = "Html"
     MARKDOWN = "Markdown"
-    LLM = "LLM"
+
+class EmbedSource(str, Enum):
     CONTENT = "Content"
+    HTML = "HTML"  # Deprecated
+    MARKDOWN = "Markdown"  # Deprecated
+    LLM = "LLM"
+
+class GenerationStrategy(str, Enum):
+    LLM = "LLM"
+    AUTO = "Auto"
 
 class GenerationConfig(BaseModel):
-    html: Optional[GenerationStrategy] = None
+    format: Optional[SegmentFormat] = None
+    strategy: Optional[GenerationStrategy] = None
     llm: Optional[str] = None
-    markdown: Optional[GenerationStrategy] = None
     crop_image: Optional[CroppingStrategy] = None
-    embed_sources: Optional[List[EmbedSource]] = Field(default_factory=lambda: [EmbedSource.MARKDOWN])
+    embed_sources: Optional[List[EmbedSource]] = None
     extended_context: Optional[bool] = None
+    # Deprecated fields for backwards compatibility
+    html: Optional[GenerationStrategy] = None  # Deprecated: Use format=SegmentFormat.HTML and strategy instead
+    markdown: Optional[GenerationStrategy] = None  # Deprecated: Use format=SegmentFormat.MARKDOWN and strategy instead
 
 class SegmentProcessing(BaseModel):
     model_config = ConfigDict(populate_by_name=True, alias_generator=str.title)
@@ -246,7 +253,7 @@ class SegmentType(str, Enum):
 
 class Segment(BaseModel):
     bbox: BoundingBox
-    content: str
+    content: str = ""
     page_height: float
     llm: Optional[str] = None
     html: Optional[str] = None
@@ -258,6 +265,7 @@ class Segment(BaseModel):
     segment_id: str
     segment_type: SegmentType
     confidence: Optional[float]
+    text: str = ""
 
 class Chunk(BaseModel):
     chunk_id: str

{chunkr_ai-0.0.49 → chunkr_ai-0.1.0}/src/chunkr_ai/models.py

@@ -15,6 +15,7 @@ from .api.configuration import (
     OcrStrategy,
     OutputResponse,
     Segment,
+    SegmentFormat,
     SegmentProcessing,
     SegmentType,
     SegmentationStrategy,
@@ -42,6 +43,7 @@ __all__ = [
     "OcrStrategy",
     "OutputResponse",
     "Segment",
+    "SegmentFormat",
     "SegmentProcessing",
     "SegmentType",
     "SegmentationStrategy",

{chunkr_ai-0.0.49 → chunkr_ai-0.1.0/src/chunkr_ai.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: chunkr-ai
-Version: 0.0.49
+Version: 0.1.0
 Summary: Python client for Chunkr: open source document intelligence
 Author-email: Ishaan Kapoor <ishaan@lumina.sh>
 License: MIT License

{chunkr_ai-0.0.49 → chunkr_ai-0.1.0}/tests/test_chunkr.py

@@ -2,7 +2,6 @@ import pytest
 from pathlib import Path
 from PIL import Image
 import asyncio
-from typing import Awaitable
 
 from chunkr_ai import Chunkr
 from chunkr_ai.models import (
@@ -21,6 +20,7 @@ from chunkr_ai.models import (
     Status,
     TaskResponse,
     Tokenizer,
+    SegmentFormat,
 )
 
 @pytest.fixture
@@ -53,9 +53,9 @@ def markdown_embed_config():
     return Configuration(
         segment_processing=SegmentProcessing(
             Page=GenerationConfig(
-                html=GenerationStrategy.LLM,
-                markdown=GenerationStrategy.LLM,
-                embed_sources=[EmbedSource.MARKDOWN]
+                format=SegmentFormat.MARKDOWN,
+                strategy=GenerationStrategy.LLM,
+                embed_sources=[EmbedSource.CONTENT]
             )
         ),
     )
@@ -65,9 +65,9 @@ def html_embed_config():
     return Configuration(
         segment_processing=SegmentProcessing(
             Page=GenerationConfig(
-                html=GenerationStrategy.LLM,
-                markdown=GenerationStrategy.LLM,
-                embed_sources=[EmbedSource.HTML]
+                format=SegmentFormat.HTML,
+                strategy=GenerationStrategy.LLM,
+                embed_sources=[EmbedSource.HTML]  # Keep this for backwards compatibility testing
             )
         ),
     )
@@ -77,10 +77,10 @@ def multiple_embed_config():
     return Configuration(
         segment_processing=SegmentProcessing(
             Page=GenerationConfig(
-                html=GenerationStrategy.LLM,
-                markdown=GenerationStrategy.LLM,
+                format=SegmentFormat.MARKDOWN,
+                strategy=GenerationStrategy.LLM,
                 llm="Generate a summary of this content",
-                embed_sources=[EmbedSource.MARKDOWN, EmbedSource.LLM, EmbedSource.HTML]
+                embed_sources=[EmbedSource.CONTENT, EmbedSource.LLM, EmbedSource.HTML]
             )
         ),
     )
@@ -169,13 +169,15 @@ def model_fallback_config():
 def extended_context_config():
     return Configuration(
         segment_processing=SegmentProcessing(
-            picture=GenerationConfig(
+            Picture=GenerationConfig(
                 extended_context=True,
-                html=GenerationStrategy.LLM,
+                format=SegmentFormat.HTML,
+                strategy=GenerationStrategy.LLM,
             ),
-            table=GenerationConfig(
+            Table=GenerationConfig(
                 extended_context=True,
-                html=GenerationStrategy.LLM,
+                format=SegmentFormat.HTML,
+                strategy=GenerationStrategy.LLM,
             )
         ),
     )
@@ -315,7 +317,8 @@ async def test_update_task_direct(client, sample_path):
         segmentation_strategy=SegmentationStrategy.PAGE,
     )
     task = await client.upload(sample_path, original_config)
-    task = await task.update(new_config)
+    task = await (await task.update(new_config))
+    assert isinstance(task, TaskResponse)
     assert task.status == "Succeeded"
     assert task.output is not None
     assert task.configuration.segmentation_strategy == SegmentationStrategy.PAGE
@@ -470,12 +473,12 @@ async def test_tokenizer_custom_string(client, sample_path, custom_tokenizer_con
     assert response.status == "Succeeded"
     assert response.output is not None
 
-@pytest.mark.asyncio
-async def test_embed_sources_with_different_tokenizer(client, sample_path, xlm_roberta_with_html_content_config):
-    response = await client.upload(sample_path, xlm_roberta_with_html_content_config)
-    assert response.task_id is not None
-    assert response.status == "Succeeded"
-    assert response.output is not None
+# @pytest.mark.asyncio
+# async def test_embed_sources_with_different_tokenizer(client, sample_path, xlm_roberta_with_html_content_config):
+#     response = await client.upload(sample_path, xlm_roberta_with_html_content_config)
+#     assert response.task_id is not None
+#     assert response.status == "Succeeded"
+#     assert response.output is not None
 
 @pytest.mark.asyncio
 async def test_error_handling_continue(client, sample_path):