retab 0.0.41__tar.gz → 0.0.43__tar.gz

retab-0.0.43/PKG-INFO

@@ -0,0 +1,117 @@
+Metadata-Version: 2.1
+Name: retab
+Version: 0.0.43
+Summary: Retab official python library
+Home-page: https://github.com/Retab-dev/retab
+Author: Retab
+Author-email: contact@retab.com
+Project-URL: Team website, https://retab.com
+Classifier: Programming Language :: Python :: 3
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Operating System :: MacOS
+Classifier: Intended Audience :: Science/Research
+Requires-Python: >=3.6
+Description-Content-Type: text/markdown
+Requires-Dist: Pillow
+Requires-Dist: httpx
+Requires-Dist: pydantic
+Requires-Dist: pydantic_core
+Requires-Dist: requests
+Requires-Dist: tqdm
+Requires-Dist: types-tqdm
+Requires-Dist: backoff
+Requires-Dist: termplotlib
+Requires-Dist: Levenshtein
+Requires-Dist: pandas
+Requires-Dist: numpy
+Requires-Dist: motor
+Requires-Dist: rich
+Requires-Dist: puremagic
+Requires-Dist: pycountry
+Requires-Dist: phonenumbers
+Requires-Dist: email_validator
+Requires-Dist: python-stdnum
+Requires-Dist: nanoid
+Requires-Dist: openai
+Requires-Dist: google-genai
+Requires-Dist: google-generativeai
+Requires-Dist: anthropic
+Requires-Dist: tiktoken
+Requires-Dist: truststore
+Requires-Dist: ruff
+
+# Retab
+
+<div align="center" style="margin-bottom: 1em;">
+
+<img src="https://raw.githubusercontent.com/Retab-dev/retab/refs/heads/main/assets/retab-logo.png" alt="Retab Logo" width="150">
+
+
+  *The AI Automation Platform*
+
+Made with love by the team at [Retab](https://retab.com) 🤍.
+
+[Our Website](https://retab.com) | [Documentation](https://docs.retab.com/get-started/introduction) | [Discord](https://discord.com/invite/vc5tWRPqag) | [Twitter](https://x.com/retabdev)
+
+
+</div>
+
+---
+
+### What is Retab?
+
+Retab solves all the major challenges in document processing with Large Language Models:
+
+1. **Universal Document Preprocessing**: Convert any file type (PDFs, Excel, emails, etc.) into LLM-ready format without writing custom parsers
+2. **Structured, Schema-driven Extraction**: Get consistent, reliable outputs using schema-based prompt engineering
+3. **Processors**: Publish a live, stable, shareable document processor.
+4. **Automations**: Create document processing workflows that can be triggered by events (mailbox, upload link, endpoint, outlook plugin).
+5. **Evaluations**: Evaluate the performance of models against annotated datasets
+6. **Optimizations**: Identify the most used processors and help you finetune models to reduce costs and improve performance
+
+We are offering you all the software-defined primitives to build your own document processing solutions. We see it as **Stripe** for document processing.
+
+Our goal is to make the process of analyzing documents and unstructured data as **easy** and **transparent** as possible.
+
+**A new, lighter paradigm**
+Large Language Models collapse entire layers of legacy OCR pipelines into a single, elegant abstraction. When a model can read, reason, and structure text natively, we no longer need brittle heuristics, handcrafted parsers, or heavyweight ETL jobs. Instead, we can expose a small, principled API: "give me the document, tell me the schema, and get back structured truth." Complexity evaporates, reliability rises, speed follows, and costs fall—because every component you remove is one that can no longer break. LLM‑first design lets us focus less on plumbing and more on the questions we actually want answered.
+
+Many people haven't yet realized how powerful LLMs have become at document processing tasks - we're here to help **unlock these capabilities**.
+
+---
+
+## Go further
+
+* [Quickstart](/get-started/quickstart)
+* [API Reference](/api-reference/introduction)
+
+---
+
+## Code examples
+
+You can check our Github repository to see code examples: [python examples](https://github.com/Retab-dev/retab/tree/main/examples) and [jupyter notebooks](https://github.com/Retab-dev/retab-nodejs/tree/main/notebooks).
+
+## Community
+
+Let's create the future of document processing together!
+
+Join our [discord community](https://discord.com/invite/vc5tWRPqag) to share tips, discuss best practices, and showcase what you build. Or just [tweet](https://x.com/retabdev) at us.
+
+We can't wait to see how you'll use Retab.
+
+* [Discord](https://discord.com/invite/vc5tWRPqag)
+* [Twitter](https://x.com/retabdev)
+
+---
+
+## Roadmap
+
+We share our roadmap publicly on [Github](https://github.com/Retab-dev/retab)
+
+Among the features we're working on:
+
+* [ ] Node.js SDK
+* [ ] Schema optimization autopilot
+* [ ] Sources API
+

retab-0.0.43/README.md

@@ -0,0 +1,74 @@
+# Retab
+
+<div align="center" style="margin-bottom: 1em;">
+
+<img src="https://raw.githubusercontent.com/Retab-dev/retab/refs/heads/main/assets/retab-logo.png" alt="Retab Logo" width="150">
+
+
+  *The AI Automation Platform*
+
+Made with love by the team at [Retab](https://retab.com) 🤍.
+
+[Our Website](https://retab.com) | [Documentation](https://docs.retab.com/get-started/introduction) | [Discord](https://discord.com/invite/vc5tWRPqag) | [Twitter](https://x.com/retabdev)
+
+
+</div>
+
+---
+
+### What is Retab?
+
+Retab solves all the major challenges in document processing with Large Language Models:
+
+1. **Universal Document Preprocessing**: Convert any file type (PDFs, Excel, emails, etc.) into LLM-ready format without writing custom parsers
+2. **Structured, Schema-driven Extraction**: Get consistent, reliable outputs using schema-based prompt engineering
+3. **Processors**: Publish a live, stable, shareable document processor.
+4. **Automations**: Create document processing workflows that can be triggered by events (mailbox, upload link, endpoint, outlook plugin).
+5. **Evaluations**: Evaluate the performance of models against annotated datasets
+6. **Optimizations**: Identify the most used processors and help you finetune models to reduce costs and improve performance
+
+We are offering you all the software-defined primitives to build your own document processing solutions. We see it as **Stripe** for document processing.
+
+Our goal is to make the process of analyzing documents and unstructured data as **easy** and **transparent** as possible.
+
+**A new, lighter paradigm**
+Large Language Models collapse entire layers of legacy OCR pipelines into a single, elegant abstraction. When a model can read, reason, and structure text natively, we no longer need brittle heuristics, handcrafted parsers, or heavyweight ETL jobs. Instead, we can expose a small, principled API: "give me the document, tell me the schema, and get back structured truth." Complexity evaporates, reliability rises, speed follows, and costs fall—because every component you remove is one that can no longer break. LLM‑first design lets us focus less on plumbing and more on the questions we actually want answered.
+
+Many people haven't yet realized how powerful LLMs have become at document processing tasks - we're here to help **unlock these capabilities**.
+
+---
+
+## Go further
+
+* [Quickstart](/get-started/quickstart)
+* [API Reference](/api-reference/introduction)
+
+---
+
+## Code examples
+
+You can check our Github repository to see code examples: [python examples](https://github.com/Retab-dev/retab/tree/main/examples) and [jupyter notebooks](https://github.com/Retab-dev/retab-nodejs/tree/main/notebooks).
+
+## Community
+
+Let's create the future of document processing together!
+
+Join our [discord community](https://discord.com/invite/vc5tWRPqag) to share tips, discuss best practices, and showcase what you build. Or just [tweet](https://x.com/retabdev) at us.
+
+We can't wait to see how you'll use Retab.
+
+* [Discord](https://discord.com/invite/vc5tWRPqag)
+* [Twitter](https://x.com/retabdev)
+
+---
+
+## Roadmap
+
+We share our roadmap publicly on [Github](https://github.com/Retab-dev/retab)
+
+Among the features we're working on:
+
+* [ ] Node.js SDK
+* [ ] Schema optimization autopilot
+* [ ] Sources API
+

{retab-0.0.41 → retab-0.0.43}/retab/__init__.py

@@ -1,4 +1,5 @@
 from .client import AsyncRetab, Retab
 from .types.schemas.object import Schema
+from . import utils
 
-__all__ = ["Retab", "AsyncRetab", "Schema"]
+__all__ = ["Retab", "AsyncRetab", "Schema", "utils"]

{retab-0.0.41 → retab-0.0.43}/retab/client.py

@@ -7,10 +7,9 @@ import backoff
 import backoff.types
 import httpx
 import truststore
-from pydantic_core import PydanticUndefined
 
-from .resources import consensus, documents, evals, files, finetuning, models, processors, schemas, secrets, usage, evaluations
-from .types.standards import PreparedRequest
+from .resources import consensus, documents, files, finetuning, models, processors, schemas, secrets, usage, evaluations
+from .types.standards import PreparedRequest, FieldUnset
 
 
 class MaxRetriesExceeded(Exception):
@@ -34,7 +33,7 @@ class BaseRetab:
 
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
-        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.dev
+        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
         timeout (float): Request timeout in seconds. Defaults to 240.0
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
@@ -43,32 +42,27 @@ class BaseRetab:
         ValueError: If no API key is provided through arguments or environment variables
     """
 
-    #   claude_api_key (str, optional): Claude API key. Will look for CLAUDE_API_KEY env variable if not provided
-    #   xai_api_key (str, optional): XAI API key. Will look for XAI_API_KEY env variable if not provided
-    #   gemini_api_key (str, optional): Gemini API key. Will look for GEMINI_API_KEY env variable if not provided
-
     def __init__(
         self,
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
         timeout: float = 240.0,
         max_retries: int = 3,
-        openai_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        gemini_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        # claude_api_key: Optional[str] = PydanticUndefined,   # type: ignore[assignment]
-        xai_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
+        openai_api_key: Optional[str] = FieldUnset,
+        gemini_api_key: Optional[str] = FieldUnset,
+        xai_api_key: Optional[str] = FieldUnset,
     ) -> None:
         if api_key is None:
             api_key = os.environ.get("RETAB_API_KEY")
 
         if api_key is None:
             raise ValueError(
-                "No API key provided. You can create an API key at https://retab.dev\n"
+                "No API key provided. You can create an API key at https://retab.com\n"
                 "Then either pass it to the client (api_key='your-key') or set the RETAB_API_KEY environment variable"
             )
 
         if base_url is None:
-            base_url = os.environ.get("RETAB_API_BASE_URL", "https://api.retab.dev")
+            base_url = os.environ.get("RETAB_API_BASE_URL", "https://api.retab.com")
 
         truststore.inject_into_ssl()
         self.api_key = api_key
@@ -80,30 +74,21 @@ class BaseRetab:
             "Content-Type": "application/json",
         }
 
-        # Only check environment variables if the value is PydanticUndefined
-        if openai_api_key is PydanticUndefined:
+        # Only check environment variables if the value is FieldUnset
+        if openai_api_key is FieldUnset:
             openai_api_key = os.environ.get("OPENAI_API_KEY")
 
-        # if claude_api_key is PydanticUndefined:
-        #    claude_api_key = os.environ.get("CLAUDE_API_KEY")
-
-        # if xai_api_key is PydanticUndefined:
-        #    xai_api_key = os.environ.get("XAI_API_KEY")
-
-        if gemini_api_key is PydanticUndefined:
+        if gemini_api_key is FieldUnset:
             gemini_api_key = os.environ.get("GEMINI_API_KEY")
 
-        # Only add headers if the values are actual strings (not None or PydanticUndefined)
-        if openai_api_key and openai_api_key is not PydanticUndefined:
+        # Only add headers if the values are actual strings (not None or FieldUnset)
+        if openai_api_key and openai_api_key is not FieldUnset:
             self.headers["OpenAI-Api-Key"] = openai_api_key
 
-        # if claude_api_key and claude_api_key is not PydanticUndefined:
-        #    self.headers["Anthropic-Api-Key"] = claude_api_key
-
-        if xai_api_key and xai_api_key is not PydanticUndefined:
+        if xai_api_key and xai_api_key is not FieldUnset:
             self.headers["XAI-Api-Key"] = xai_api_key
 
-        if gemini_api_key and gemini_api_key is not PydanticUndefined:
+        if gemini_api_key and gemini_api_key is not FieldUnset:
             self.headers["Gemini-Api-Key"] = gemini_api_key
 
     def _prepare_url(self, endpoint: str) -> str:
@@ -154,12 +139,10 @@ class Retab(BaseRetab):
 
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
-        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.dev
+        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
         timeout (float): Request timeout in seconds. Defaults to 240.0
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
-        claude_api_key (str, optional): Claude API key. Will look for CLAUDE_API_KEY env variable if not provided
-        xai_api_key (str, optional): XAI API key. Will look for XAI_API_KEY env variable if not provided
         gemini_api_key (str, optional): Gemini API key. Will look for GEMINI_API_KEY env variable if not provided
 
     Attributes:
@@ -179,10 +162,8 @@ class Retab(BaseRetab):
         base_url: Optional[str] = None,
         timeout: float = 240.0,
         max_retries: int = 3,
-        openai_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        gemini_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        # claude_api_key: Optional[str] = PydanticUndefined,   # type: ignore[assignment]
-        # xai_api_key: Optional[str] = PydanticUndefined,   # type: ignore[assignment]
+        openai_api_key: Optional[str] = FieldUnset,
+        gemini_api_key: Optional[str] = FieldUnset,
     ) -> None:
         super().__init__(
             api_key=api_key,
@@ -191,16 +172,12 @@ class Retab(BaseRetab):
             max_retries=max_retries,
             openai_api_key=openai_api_key,
             gemini_api_key=gemini_api_key,
-            # claude_api_key=claude_api_key,
-            # xai_api_key=xai_api_key,
         )
 
         self.client = httpx.Client(timeout=self.timeout)
-        self.evals = evals.Evals(client=self)
         self.evaluations = evaluations.Evaluations(client=self)
         self.files = files.Files(client=self)
         self.fine_tuning = finetuning.FineTuning(client=self)
-        # self.prompt_optimization = prompt_optimization.PromptOptimization(client=self)
         self.documents = documents.Documents(client=self)
         self.models = models.Models(client=self)
         self.schemas = schemas.Schemas(client=self)
@@ -422,7 +399,7 @@ class AsyncRetab(BaseRetab):
 
     Args:
         api_key (str, optional): Retab API key. If not provided, will look for RETAB_API_KEY env variable.
-        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.dev
+        base_url (str, optional): Base URL for API requests. Defaults to https://api.retab.com
         timeout (float): Request timeout in seconds. Defaults to 240.0
         max_retries (int): Maximum number of retries for failed requests. Defaults to 3
         openai_api_key (str, optional): OpenAI API key. Will look for OPENAI_API_KEY env variable if not provided
@@ -447,10 +424,8 @@ class AsyncRetab(BaseRetab):
         base_url: Optional[str] = None,
         timeout: float = 240.0,
         max_retries: int = 3,
-        openai_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        gemini_api_key: Optional[str] = PydanticUndefined,  # type: ignore[assignment]
-        # claude_api_key: Optional[str] = PydanticUndefined,   # type: ignore[assignment]
-        # xai_api_key: Optional[str] = PydanticUndefined,   # type: ignore[assignment]
+        openai_api_key: Optional[str] = FieldUnset,
+        gemini_api_key: Optional[str] = FieldUnset,
     ) -> None:
         super().__init__(
             api_key=api_key,
@@ -459,17 +434,13 @@ class AsyncRetab(BaseRetab):
             max_retries=max_retries,
             openai_api_key=openai_api_key,
             gemini_api_key=gemini_api_key,
-            # claude_api_key=claude_api_key,
-            # xai_api_key=xai_api_key,
         )
 
         self.client = httpx.AsyncClient(timeout=self.timeout)
 
-        self.evals = evals.AsyncEvals(client=self)
         self.evaluations = evaluations.AsyncEvaluations(client=self)
         self.files = files.AsyncFiles(client=self)
         self.fine_tuning = finetuning.AsyncFineTuning(client=self)
-        # self.prompt_optimization = prompt_optimization.AsyncPromptOptimization(client=self)
         self.documents = documents.AsyncDocuments(client=self)
         self.models = models.AsyncModels(client=self)
         self.schemas = schemas.AsyncSchemas(client=self)

{retab-0.0.41 → retab-0.0.43}/retab/resources/consensus/client.py

@@ -21,7 +21,7 @@ class BaseConsensusMixin:
             mode=mode,
         )
 
-        return PreparedRequest(method="POST", url="/v1/consensus/reconcile", data=request.model_dump(), idempotency_key=idempotency_key)
+        return PreparedRequest(method="POST", url="/v1/consensus/reconcile", data=request.model_dump(mode="json", exclude_unset=True), idempotency_key=idempotency_key)
 
 
 class Consensus(SyncAPIResource, BaseConsensusMixin):

{retab-0.0.41 → retab-0.0.43}/retab/resources/consensus/completions_stream.py

@@ -123,7 +123,7 @@ class Completions(SyncAPIResource, BaseCompletionsMixin):
 
         Usage:
         ```python
-        with retab.devpletions.stream(json_schema, messages, model, temperature, reasoning_effort) as stream:
+        with retab.completions.stream(json_schema, messages, model, temperature, reasoning_effort) as stream:
             for response in stream:
                 print(response)
         ```
@@ -210,7 +210,7 @@ class AsyncCompletions(AsyncAPIResource, BaseCompletionsMixin):
 
         Usage:
         ```python
-        async with retab.devpletions.stream(json_schema, messages, model, temperature, reasoning_effort, n_consensus) as stream:
+        async with retab.completions.stream(json_schema, messages, model, temperature, reasoning_effort, n_consensus) as stream:
             async for response in stream:
                 print(response)
         ```

{retab-0.0.41 → retab-0.0.43}/retab/resources/consensus/responses.py

@@ -55,7 +55,7 @@ class BaseResponsesMixin:
             instructions=instructions,
         )
 
-        return PreparedRequest(method="POST", url="/v1/responses", data=request.model_dump(), idempotency_key=idempotency_key)
+        return PreparedRequest(method="POST", url="/v1/responses", data=request.model_dump(mode="json", exclude_unset=True), idempotency_key=idempotency_key)
 
     def prepare_parse(
         self,