datalab-python-sdk 0.2.0__tar.gz → 0.2.2__tar.gz

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datalab-python-sdk
-Version: 0.2.0
+Version: 0.2.2
 Summary: SDK for the Datalab document intelligence API
 Author-email: Datalab Team <hi@datalab.to>
 License-Expression: MIT
@@ -12,6 +12,7 @@ Requires-Dist: click>=8.2.1
 Requires-Dist: pydantic-settings>=2.10.1
 Requires-Dist: pydantic>=2.11.7
 Requires-Dist: tenacity>=8.2.3
+Requires-Dist: tqdm>=4.66.0
 Description-Content-Type: text/markdown
 
 # Datalab SDK
@@ -54,7 +55,7 @@ Workflows allow you to chain multiple document processing steps together. Each w
 
 **Note:** All workflow operations require authentication. Make sure you have set your `DATALAB_API_KEY` environment variable or pass `api_key` when creating the client (see [Authentication](#authentication) section above).
 
-For more Workflow tips, see our [examples here](./examples/README.md) and [documentation](https://documentation.datalab.to/docs/recipes/workflows/workflow-concepts).
+For more Workflow tips, see our [documentation](https://documentation.datalab.to/docs/recipes/workflows/workflow-concepts).
 
 ## CLI Usage
 
@@ -64,9 +65,6 @@ The SDK includes a command-line interface:
 # Convert document to markdown
 datalab convert document.pdf
 
-# OCR with JSON output
-datalab ocr document.pdf --output-format json
-
 # Workflow commands
 datalab create-workflow --help
 datalab execute-workflow --help

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/README.md

@@ -38,7 +38,7 @@ Workflows allow you to chain multiple document processing steps together. Each w
 
 **Note:** All workflow operations require authentication. Make sure you have set your `DATALAB_API_KEY` environment variable or pass `api_key` when creating the client (see [Authentication](#authentication) section above).
 
-For more Workflow tips, see our [examples here](./examples/README.md) and [documentation](https://documentation.datalab.to/docs/recipes/workflows/workflow-concepts).
+For more Workflow tips, see our [documentation](https://documentation.datalab.to/docs/recipes/workflows/workflow-concepts).
 
 ## CLI Usage
 
@@ -48,9 +48,6 @@ The SDK includes a command-line interface:
 # Convert document to markdown
 datalab convert document.pdf
 
-# OCR with JSON output
-datalab ocr document.pdf --output-format json
-
 # Workflow commands
 datalab create-workflow --help
 datalab execute-workflow --help

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/datalab_sdk/cli.py

@@ -9,6 +9,7 @@ import asyncio
 from pathlib import Path
 from typing import Optional, List
 import click
+from tqdm import tqdm
 
 from datalab_sdk.client import AsyncDatalabClient, DatalabClient
 from datalab_sdk.mimetypes import SUPPORTED_EXTENSIONS
@@ -122,6 +123,25 @@ async def process_files_async(
     """Process files asynchronously"""
     semaphore = asyncio.Semaphore(max_concurrent)
 
+    async def call_api(client, file_path, output_path):
+        """Make API call - client handles retries for rate limits"""
+        if method == "convert":
+            return await client.convert(
+                file_path,
+                options=options,
+                save_output=output_path,
+                max_polls=max_polls,
+                poll_interval=poll_interval,
+            )
+        else:  # method == 'ocr'
+            return await client.ocr(
+                file_path,
+                options=options,
+                save_output=output_path,
+                max_polls=max_polls,
+                poll_interval=poll_interval,
+            )
+
     async def process_single_file(file_path: Path) -> dict:
         async with semaphore:
             try:
@@ -134,22 +154,7 @@ async def process_files_async(
                 async with AsyncDatalabClient(
                     api_key=api_key, base_url=base_url
                 ) as client:
-                    if method == "convert":
-                        result = await client.convert(
-                            file_path,
-                            options=options,
-                            save_output=output_path,
-                            max_polls=max_polls,
-                            poll_interval=poll_interval,
-                        )
-                    else:  # method == 'ocr'
-                        result = await client.ocr(
-                            file_path,
-                            options=options,
-                            save_output=output_path,
-                            max_polls=max_polls,
-                            poll_interval=poll_interval,
-                        )
+                    result = await call_api(client, file_path, output_path)
 
                 return {
                     "file_path": str(file_path),
@@ -167,9 +172,19 @@ async def process_files_async(
                     "page_count": None,
                 }
 
-    # Process all files concurrently
-    tasks = [process_single_file(file_path) for file_path in files]
-    results = await asyncio.gather(*tasks)
+    # Process all files concurrently with progress bar
+    tasks = [asyncio.create_task(process_single_file(file_path)) for file_path in files]
+    results = []
+
+    with tqdm(total=len(tasks), desc="Processing", unit="file") as pbar:
+        for coro in asyncio.as_completed(tasks):
+            result = await coro
+            results.append(result)
+            # Update progress bar description with current file
+            filename = Path(result["file_path"]).name
+            status = "✓" if result["success"] else "✗"
+            pbar.set_postfix_str(f"{status} {filename[:30]}")
+            pbar.update(1)
 
     return results
 

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/datalab_sdk/client.py

@@ -112,7 +112,9 @@ class AsyncDatalabClient:
             try:
                 error_data = await response.json()
                 # FastAPI returns errors in "detail" field, but some APIs use "error"
-                error_message = error_data.get("detail") or error_data.get("error") or str(e)
+                error_message = (
+                    error_data.get("detail") or error_data.get("error") or str(e)
+                )
             except Exception:
                 error_message = str(e)
             raise DatalabAPIError(
@@ -123,6 +125,19 @@ class AsyncDatalabClient:
         except aiohttp.ClientError as e:
             raise DatalabAPIError(f"Request failed: {str(e)}")
 
+    @retry(
+        retry=retry_if_exception(
+            lambda e: isinstance(e, DatalabAPIError)
+            and getattr(e, "status_code", None) == 429
+        ),
+        stop=stop_after_attempt(10),
+        wait=wait_exponential_jitter(initial=5, max=120),
+        reraise=True,
+    )
+    async def _submit_with_retry(self, endpoint: str, data) -> Dict[str, Any]:
+        """POST submission with retry for rate limits (429)"""
+        return await self._make_request("POST", endpoint, data=data)
+
     async def _poll_result(
         self, check_url: str, max_polls: int = 300, poll_interval: int = 1
     ) -> Dict[str, Any]:
@@ -168,8 +183,8 @@ class AsyncDatalabClient:
                 )
             )
         ),
-        stop=stop_after_attempt(2),
-        wait=wait_exponential_jitter(max=0.5),
+        stop=stop_after_attempt(10),
+        wait=wait_exponential_jitter(initial=5, max=120),
         reraise=True,
     )
     async def _poll_get_with_retry(self, url: str) -> Dict[str, Any]:
@@ -185,7 +200,7 @@ class AsyncDatalabClient:
 
         # Read file content
         file_data = file_path.read_bytes()
-        
+
         # Check if file is empty
         if not file_data:
             raise DatalabFileError(
@@ -252,8 +267,7 @@ class AsyncDatalabClient:
         if options is None:
             options = ConvertOptions()
 
-        initial_data = await self._make_request(
-            "POST",
+        initial_data = await self._submit_with_retry(
             "/api/v1/marker",
             data=self.get_form_params(
                 file_path=file_path, file_url=file_url, options=options
@@ -283,6 +297,7 @@ class AsyncDatalabClient:
             images=result_data.get("images"),
             metadata=result_data.get("metadata"),
             error=result_data.get("error"),
+            error_in=result_data.get("error_in"),
             page_count=result_data.get("page_count"),
             status=result_data.get("status", "complete"),
             checkpoint_id=result_data.get("checkpoint_id"),
@@ -312,8 +327,7 @@ class AsyncDatalabClient:
         if options is None:
             options = OCROptions()
 
-        initial_data = await self._make_request(
-            "POST",
+        initial_data = await self._submit_with_retry(
             "/api/v1/ocr",
             data=self.get_form_params(file_path=file_path, options=options),
         )
@@ -370,8 +384,7 @@ class AsyncDatalabClient:
         if options is None:
             raise ValueError("options must be provided with field_data")
 
-        initial_data = await self._make_request(
-            "POST",
+        initial_data = await self._submit_with_retry(
             "/api/v1/fill",
             data=self.get_form_params(
                 file_path=file_path, file_url=file_url, options=options
@@ -393,6 +406,7 @@ class AsyncDatalabClient:
             status=result_data.get("status", "complete"),
             success=result_data.get("success"),
             error=result_data.get("error"),
+            error_in=result_data.get("error_in"),
             output_format=result_data.get("output_format"),
             output_base64=result_data.get("output_base64"),
             fields_filled=result_data.get("fields_filled"),
@@ -570,7 +584,9 @@ class AsyncDatalabClient:
 
         return {
             "success": response.get("success", True),
-            "message": response.get("message", f"Workflow {workflow_id} deleted successfully"),
+            "message": response.get(
+                "message", f"Workflow {workflow_id} deleted successfully"
+            ),
         }
 
     async def execute_workflow(
@@ -1236,9 +1252,7 @@ class DatalabClient:
         Returns:
             UploadedFileMetadata object with file information
         """
-        return self._run_async(
-            self._async_client.get_file_metadata(file_id=file_id)
-        )
+        return self._run_async(self._async_client.get_file_metadata(file_id=file_id))
 
     def get_file_download_url(
         self,
@@ -1260,7 +1274,9 @@ class DatalabClient:
                 - original_filename: Original filename
         """
         return self._run_async(
-            self._async_client.get_file_download_url(file_id=file_id, expires_in=expires_in)
+            self._async_client.get_file_download_url(
+                file_id=file_id, expires_in=expires_in
+            )
         )
 
     def delete_file(
@@ -1280,9 +1296,7 @@ class DatalabClient:
                 - success: Whether the deletion was successful
                 - message: Confirmation message
         """
-        return self._run_async(
-            self._async_client.delete_file(file_id=file_id)
-        )
+        return self._run_async(self._async_client.delete_file(file_id=file_id))
 
     def delete_workflow(self, workflow_id: int) -> Dict[str, Any]:
         """

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/datalab_sdk/models.py

@@ -3,7 +3,7 @@ Datalab SDK data models
 """
 
 from dataclasses import dataclass, field
-from typing import Dict, List, Optional, Any, Union
+from typing import Dict, List, Optional, Any, Union, Literal
 from pathlib import Path
 import json
 import base64
@@ -41,19 +41,20 @@ class ConvertOptions(ProcessingOptions):
     paginate: bool = False
     disable_image_extraction: bool = False
     disable_image_captions: bool = False
+    fence_synthetic_captions: bool = False
     additional_config: Optional[Dict[str, Any]] = None
     page_schema: Optional[Dict[str, Any]] = None
     segmentation_schema: Optional[str] = None  # JSON string for document segmentation
     save_checkpoint: bool = False
-    extras: Optional[str] = (
-        None  # Comma-separated list: 'track_changes', 'chart_understanding'
-    )
     output_format: str = "markdown"  # markdown, json, html, chunks
     mode: str = "balanced"  # fast, balanced, accurate
     keep_spreadsheet_formatting: bool = False
     webhook_url: Optional[str] = None
-    extras: Optional[str] = None  # comma-separated extras
+    # Comma-separated list of extra features: 'track_changes', 'chart_understanding',
+    # 'table_row_bboxes', 'extract_links', 'infographic', 'new_block_types'
+    extras: Optional[str] = None
     add_block_ids: bool = False  # add block IDs to HTML output
+    include_markdown_in_chunks: bool = False  # include markdown field in chunks/JSON output
 
     def to_form_data(self) -> Dict[str, Any]:
         """Convert to form data format for API requests"""
@@ -121,6 +122,7 @@ class ConversionResult:
     images: Optional[Dict[str, str]] = None
     metadata: Optional[Dict[str, Any]] = None
     error: Optional[str] = None
+    error_in: Optional[Literal["VALIDATION", "INFERENCE", "OTHER"]] = None
     page_count: Optional[int] = None
     status: str = "complete"
     checkpoint_id: Optional[str] = None
@@ -375,6 +377,7 @@ class FormFillingResult:
     status: str
     success: Optional[bool] = None
     error: Optional[str] = None
+    error_in: Optional[Literal["VALIDATION", "INFERENCE", "OTHER"]] = None
     output_format: Optional[str] = None  # "pdf" or "png"
     output_base64: Optional[str] = None  # Base64-encoded filled form
     fields_filled: Optional[List[str]] = (

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/pyproject.toml

@@ -7,7 +7,7 @@ readme = "README.md"
 license = "MIT"
 repository = "https://github.com/datalab-to/sdk"
 keywords = ["datalab", "sdk", "document-intelligence", "api"]
-version = "0.2.0"
+version = "0.2.2"
 description = "SDK for the Datalab document intelligence API"
 requires-python = ">=3.10"
 dependencies = [
@@ -16,6 +16,7 @@ dependencies = [
     "pydantic>=2.11.7",
     "pydantic-settings>=2.10.1",
     "tenacity>=8.2.3",
+    "tqdm>=4.66.0",
 ]
 
 [project.scripts]

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/tests/test_client_methods.py

@@ -504,13 +504,13 @@ class TestClientErrorHandling:
             with patch.object(
                 client, "_make_request", new_callable=AsyncMock
             ) as mock_request:
-                # Setup mock to raise API error
+                # Setup mock to raise API error (use 400 since 429 is retried)
                 mock_request.side_effect = DatalabAPIError(
-                    "API rate limit exceeded", status_code=429
+                    "Bad request", status_code=400
                 )
 
                 # Test that error is propagated
-                with pytest.raises(DatalabAPIError, match="API rate limit exceeded"):
+                with pytest.raises(DatalabAPIError, match="Bad request"):
                     await client.ocr(pdf_file)
 
     def test_convert_unsuccessful_response(self, temp_dir):

{datalab_python_sdk-0.2.0 → datalab_python_sdk-0.2.2}/uv.lock

@@ -212,7 +212,7 @@ wheels = [
 
 [[package]]
 name = "datalab-python-sdk"
-version = "0.2.0"
+version = "0.2.2"
 source = { editable = "." }
 dependencies = [
     { name = "aiohttp" },
@@ -220,6 +220,7 @@ dependencies = [
     { name = "pydantic" },
     { name = "pydantic-settings" },
     { name = "tenacity" },
+    { name = "tqdm" },
 ]
 
 [package.dev-dependencies]
@@ -240,6 +241,7 @@ requires-dist = [
     { name = "pydantic", specifier = ">=2.11.7" },
     { name = "pydantic-settings", specifier = ">=2.10.1" },
     { name = "tenacity", specifier = ">=8.2.3" },
+    { name = "tqdm", specifier = ">=4.66.0" },
 ]
 
 [package.metadata.requires-dev]
@@ -1102,6 +1104,18 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/77/b8/0135fadc89e73be292b473cb820b4f5a08197779206b33191e801feeae40/tomli-2.3.0-py3-none-any.whl", hash = "sha256:e95b1af3c5b07d9e643909b5abbec77cd9f1217e6d0bca72b0234736b9fb1f1b", size = 14408, upload-time = "2025-10-08T22:01:46.04Z" },
 ]
 
+[[package]]
+name = "tqdm"
+version = "4.67.3"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "colorama", marker = "sys_platform == 'win32'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/09/a9/6ba95a270c6f1fbcd8dac228323f2777d886cb206987444e4bce66338dd4/tqdm-4.67.3.tar.gz", hash = "sha256:7d825f03f89244ef73f1d4ce193cb1774a8179fd96f31d7e1dcde62092b960bb", size = 169598, upload-time = "2026-02-03T17:35:53.048Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/16/e1/3079a9ff9b8e11b846c6ac5c8b5bfb7ff225eee721825310c91b3b50304f/tqdm-4.67.3-py3-none-any.whl", hash = "sha256:ee1e4c0e59148062281c49d80b25b67771a127c85fc9676d3be5f243206826bf", size = 78374, upload-time = "2026-02-03T17:35:50.982Z" },
+]
+
 [[package]]
 name = "typing-extensions"
 version = "4.15.0"