gaik 0.2.6__tar.gz

gaik-0.2.6/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 GAIK - GenAI for knowledge mgt
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

gaik-0.2.6/PKG-INFO

@@ -0,0 +1,275 @@
+Metadata-Version: 2.4
+Name: gaik
+Version: 0.2.6
+Summary: General AI Kit - Reusable AI/ML components for Python
+Author: GAIK Project
+License: MIT License
+        
+        Copyright (c) 2025 GAIK - GenAI for knowledge mgt
+        
+        Permission is hereby granted, free of charge, to any person obtaining a copy
+        of this software and associated documentation files (the "Software"), to deal
+        in the Software without restriction, including without limitation the rights
+        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+        copies of the Software, and to permit persons to whom the Software is
+        furnished to do so, subject to the following conditions:
+        
+        The above copyright notice and this permission notice shall be included in all
+        copies or substantial portions of the Software.
+        
+        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+        SOFTWARE.
+        
+Project-URL: Homepage, https://gaik.ai/
+Project-URL: Repository, https://github.com/GAIK-project/toolkit-shared-components
+Project-URL: Documentation, https://github.com/GAIK-project/toolkit-shared-components/tree/main/gaik-py
+Project-URL: Issues, https://github.com/GAIK-project/toolkit-shared-components/issues
+Keywords: ai,ml,langchain,openai,anthropic,google,structured-outputs,pydantic,schema,extraction
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pydantic>=2.12.3
+Requires-Dist: langchain-core>=1.0.3
+Requires-Dist: langchain-openai>=1.0.2
+Requires-Dist: langchain-anthropic>=1.0.1
+Requires-Dist: langchain-google-genai>=3.0.1
+Provides-Extra: dev
+Requires-Dist: ruff>=0.14.1; extra == "dev"
+Requires-Dist: build>=1.0; extra == "dev"
+Requires-Dist: twine>=4.0; extra == "dev"
+Provides-Extra: vision
+Requires-Dist: openai>=1.40.0; extra == "vision"
+Requires-Dist: pdf2image>=1.17.0; extra == "vision"
+Requires-Dist: pillow>=10.0.0; extra == "vision"
+Requires-Dist: python-dotenv>=1.0.0; extra == "vision"
+Dynamic: license-file
+
+# GAIK - General AI Kit
+
+**Reusable AI/ML components for Python**
+
+Multi-provider AI toolkit for structured data extraction. Supports OpenAI, Anthropic Claude, Google Gemini, and Azure OpenAI.
+
+## Features
+
+### 🔍 Dynamic Data Extraction (`gaik.extract`)
+
+Extract structured data from unstructured text using LangChain's structured outputs:
+
+- ✅ **Multi-provider** - OpenAI, Anthropic, Azure, Google - easy switching
+- ✅ **Guaranteed structure** - API-enforced schema compliance
+- ✅ **Type-safe** - Full Pydantic validation
+- ✅ **No code generation** - Uses Pydantic's `create_model()`, no `eval()`
+- ✅ **Cost-effective** - Minimal API calls
+- ✅ **Simple & clean** - Easy to understand, minimal dependencies
+
+### 🖼️ Vision PDF Parsing (`gaik.parsers`)
+
+Convert PDF pages to Markdown with OpenAI or Azure OpenAI vision models:
+
+- ✅ **Single API surface** - Works with standard OpenAI or Azure deployments
+- ✅ **Optional extras** - Install with `pip install gaik[vision]`
+- ✅ **CLI ready** - See `examples/demo_vision_parser.py` for quick conversions
+- ✅ **Table-aware** - Keeps multi-page tables intact with optional cleanup
+
+## Installation
+
+```bash
+# Install from Test PyPI
+pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ gaik
+```
+
+## Quick Start
+
+### 1. Set up your provider API key
+
+**OpenAI (default):**
+
+```bash
+export OPENAI_API_KEY='sk-...'  # Get from: https://platform.openai.com/api-keys
+```
+
+**Anthropic:**
+
+```bash
+export ANTHROPIC_API_KEY='sk-ant-...'  # Get from: https://console.anthropic.com
+```
+
+**Google:**
+
+```bash
+export GOOGLE_API_KEY='...'  # Get from: https://ai.google.dev
+```
+
+**Azure OpenAI:**
+
+```bash
+export AZURE_OPENAI_API_KEY='...'
+export AZURE_OPENAI_ENDPOINT='https://your-resource.openai.azure.com/'
+```
+
+### 2. Simple Extraction
+
+```python
+from gaik.extract import SchemaExtractor
+
+# Using default OpenAI provider
+extractor = SchemaExtractor("Extract name and age from text")
+result = extractor.extract_one("Alice is 25 years old")
+print(result)
+# {'name': 'Alice', 'age': 25}
+
+# Switch to Anthropic Claude
+extractor = SchemaExtractor(
+    "Extract name and age from text",
+    provider="anthropic"
+)
+
+# Use Google Gemini
+extractor = SchemaExtractor(
+    "Extract name and age from text",
+    provider="google"
+)
+```
+
+### 3. Batch Extraction
+
+```python
+from gaik.extract import dynamic_extraction_workflow
+
+description = """
+Extract from invoices:
+- Invoice number
+- Total amount in USD
+- Vendor name
+"""
+
+documents = [
+    "Invoice #12345 from Acme Corp. Total: $1,500",
+    "INV-67890, Supplier: TechCo, Amount: $2,750"
+]
+
+# Use any provider
+results = dynamic_extraction_workflow(
+    description,
+    documents,
+    provider="openai"  # or "anthropic", "google", "azure"
+)
+
+for result in results:
+    print(f"Invoice: {result['invoice_number']}, Amount: ${result['total_amount']}")
+```
+
+### 4. Reusable Extractor (Recommended)
+
+```python
+from gaik.extract import SchemaExtractor
+
+# Create extractor once
+extractor = SchemaExtractor("""
+Extract from project reports:
+- Project title
+- Lead institution
+- Total funding in euros
+- List of partner countries
+""")
+
+# Reuse for multiple batches
+batch1_results = extractor.extract(documents_batch1)
+batch2_results = extractor.extract(documents_batch2)
+
+# Inspect the schema
+print(f"Fields: {extractor.field_names}")
+# ['project_title', 'lead_institution', 'total_funding', 'partner_countries']
+```
+
+### 5. Schema-Only Generation
+
+Generate Pydantic schemas without extraction:
+
+```python
+from gaik.extract import FieldSpec, ExtractionRequirements, create_extraction_model
+
+requirements = ExtractionRequirements(
+    use_case_name="Invoice",
+    fields=[
+        FieldSpec(
+            field_name="invoice_number",
+            field_type="str",
+            description="Invoice identifier",
+            required=True
+        ),
+        FieldSpec(
+            field_name="amount",
+            field_type="float",
+            description="Total amount",
+            required=True
+        )
+    ]
+)
+
+# Create Pydantic model
+InvoiceModel = create_extraction_model(requirements)
+schema = InvoiceModel.model_json_schema()
+```
+
+## API Reference
+
+| Function/Class                  | Purpose                                           |
+| ------------------------------- | ------------------------------------------------- |
+| `SchemaExtractor`               | Reusable extractor with provider selection        |
+| `dynamic_extraction_workflow()` | One-shot extraction from natural language         |
+| `create_extraction_model()`     | Generate Pydantic model from field specifications |
+| `FieldSpec`                     | Define a single extraction field                  |
+| `ExtractionRequirements`        | Collection of field specifications                |
+
+### Provider Parameters
+
+```python
+SchemaExtractor(
+    user_description: str | None = None,  # Optional if requirements provided
+    provider: Literal["openai", "anthropic", "google", "azure"] = "openai",
+    model: str | None = None,             # Optional: override default model
+    api_key: str | None = None,           # Optional: override env variable
+    client: BaseChatModel | None = None,  # Optional: custom LangChain client
+    requirements: ExtractionRequirements | None = None  # Optional: pre-defined schema
+)
+```
+
+**Note:**
+
+- IDEs with type checking (VS Code, PyCharm) will show autocomplete for `provider` parameter
+- Either `user_description` or `requirements` must be provided
+- Using `requirements` skips LLM parsing step (faster & cheaper)
+
+## Default Models
+
+- OpenAI: `gpt-4.1`
+- Anthropic: `claude-sonnet-4-5-20250929`
+- Google: `gemini-2.5-flash`
+- Azure: `gpt-4.1`
+
+## Resources
+
+- [GitHub Repository](https://github.com/GAIK-project/toolkit-shared-components)
+- [Examples Directory](https://github.com/GAIK-project/toolkit-shared-components/tree/main/examples)
+- [LangChain Documentation](https://python.langchain.com/docs/how_to/structured_output/)
+- [Pydantic Documentation](https://docs.pydantic.dev/)
+
+## License
+
+MIT License - see [LICENSE](LICENSE) file for details.

gaik-0.2.6/README.md

@@ -0,0 +1,215 @@
+# GAIK - General AI Kit
+
+**Reusable AI/ML components for Python**
+
+Multi-provider AI toolkit for structured data extraction. Supports OpenAI, Anthropic Claude, Google Gemini, and Azure OpenAI.
+
+## Features
+
+### 🔍 Dynamic Data Extraction (`gaik.extract`)
+
+Extract structured data from unstructured text using LangChain's structured outputs:
+
+- ✅ **Multi-provider** - OpenAI, Anthropic, Azure, Google - easy switching
+- ✅ **Guaranteed structure** - API-enforced schema compliance
+- ✅ **Type-safe** - Full Pydantic validation
+- ✅ **No code generation** - Uses Pydantic's `create_model()`, no `eval()`
+- ✅ **Cost-effective** - Minimal API calls
+- ✅ **Simple & clean** - Easy to understand, minimal dependencies
+
+### 🖼️ Vision PDF Parsing (`gaik.parsers`)
+
+Convert PDF pages to Markdown with OpenAI or Azure OpenAI vision models:
+
+- ✅ **Single API surface** - Works with standard OpenAI or Azure deployments
+- ✅ **Optional extras** - Install with `pip install gaik[vision]`
+- ✅ **CLI ready** - See `examples/demo_vision_parser.py` for quick conversions
+- ✅ **Table-aware** - Keeps multi-page tables intact with optional cleanup
+
+## Installation
+
+```bash
+# Install from Test PyPI
+pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ gaik
+```
+
+## Quick Start
+
+### 1. Set up your provider API key
+
+**OpenAI (default):**
+
+```bash
+export OPENAI_API_KEY='sk-...'  # Get from: https://platform.openai.com/api-keys
+```
+
+**Anthropic:**
+
+```bash
+export ANTHROPIC_API_KEY='sk-ant-...'  # Get from: https://console.anthropic.com
+```
+
+**Google:**
+
+```bash
+export GOOGLE_API_KEY='...'  # Get from: https://ai.google.dev
+```
+
+**Azure OpenAI:**
+
+```bash
+export AZURE_OPENAI_API_KEY='...'
+export AZURE_OPENAI_ENDPOINT='https://your-resource.openai.azure.com/'
+```
+
+### 2. Simple Extraction
+
+```python
+from gaik.extract import SchemaExtractor
+
+# Using default OpenAI provider
+extractor = SchemaExtractor("Extract name and age from text")
+result = extractor.extract_one("Alice is 25 years old")
+print(result)
+# {'name': 'Alice', 'age': 25}
+
+# Switch to Anthropic Claude
+extractor = SchemaExtractor(
+    "Extract name and age from text",
+    provider="anthropic"
+)
+
+# Use Google Gemini
+extractor = SchemaExtractor(
+    "Extract name and age from text",
+    provider="google"
+)
+```
+
+### 3. Batch Extraction
+
+```python
+from gaik.extract import dynamic_extraction_workflow
+
+description = """
+Extract from invoices:
+- Invoice number
+- Total amount in USD
+- Vendor name
+"""
+
+documents = [
+    "Invoice #12345 from Acme Corp. Total: $1,500",
+    "INV-67890, Supplier: TechCo, Amount: $2,750"
+]
+
+# Use any provider
+results = dynamic_extraction_workflow(
+    description,
+    documents,
+    provider="openai"  # or "anthropic", "google", "azure"
+)
+
+for result in results:
+    print(f"Invoice: {result['invoice_number']}, Amount: ${result['total_amount']}")
+```
+
+### 4. Reusable Extractor (Recommended)
+
+```python
+from gaik.extract import SchemaExtractor
+
+# Create extractor once
+extractor = SchemaExtractor("""
+Extract from project reports:
+- Project title
+- Lead institution
+- Total funding in euros
+- List of partner countries
+""")
+
+# Reuse for multiple batches
+batch1_results = extractor.extract(documents_batch1)
+batch2_results = extractor.extract(documents_batch2)
+
+# Inspect the schema
+print(f"Fields: {extractor.field_names}")
+# ['project_title', 'lead_institution', 'total_funding', 'partner_countries']
+```
+
+### 5. Schema-Only Generation
+
+Generate Pydantic schemas without extraction:
+
+```python
+from gaik.extract import FieldSpec, ExtractionRequirements, create_extraction_model
+
+requirements = ExtractionRequirements(
+    use_case_name="Invoice",
+    fields=[
+        FieldSpec(
+            field_name="invoice_number",
+            field_type="str",
+            description="Invoice identifier",
+            required=True
+        ),
+        FieldSpec(
+            field_name="amount",
+            field_type="float",
+            description="Total amount",
+            required=True
+        )
+    ]
+)
+
+# Create Pydantic model
+InvoiceModel = create_extraction_model(requirements)
+schema = InvoiceModel.model_json_schema()
+```
+
+## API Reference
+
+| Function/Class                  | Purpose                                           |
+| ------------------------------- | ------------------------------------------------- |
+| `SchemaExtractor`               | Reusable extractor with provider selection        |
+| `dynamic_extraction_workflow()` | One-shot extraction from natural language         |
+| `create_extraction_model()`     | Generate Pydantic model from field specifications |
+| `FieldSpec`                     | Define a single extraction field                  |
+| `ExtractionRequirements`        | Collection of field specifications                |
+
+### Provider Parameters
+
+```python
+SchemaExtractor(
+    user_description: str | None = None,  # Optional if requirements provided
+    provider: Literal["openai", "anthropic", "google", "azure"] = "openai",
+    model: str | None = None,             # Optional: override default model
+    api_key: str | None = None,           # Optional: override env variable
+    client: BaseChatModel | None = None,  # Optional: custom LangChain client
+    requirements: ExtractionRequirements | None = None  # Optional: pre-defined schema
+)
+```
+
+**Note:**
+
+- IDEs with type checking (VS Code, PyCharm) will show autocomplete for `provider` parameter
+- Either `user_description` or `requirements` must be provided
+- Using `requirements` skips LLM parsing step (faster & cheaper)
+
+## Default Models
+
+- OpenAI: `gpt-4.1`
+- Anthropic: `claude-sonnet-4-5-20250929`
+- Google: `gemini-2.5-flash`
+- Azure: `gpt-4.1`
+
+## Resources
+
+- [GitHub Repository](https://github.com/GAIK-project/toolkit-shared-components)
+- [Examples Directory](https://github.com/GAIK-project/toolkit-shared-components/tree/main/examples)
+- [LangChain Documentation](https://python.langchain.com/docs/how_to/structured_output/)
+- [Pydantic Documentation](https://docs.pydantic.dev/)
+
+## License
+
+MIT License - see [LICENSE](LICENSE) file for details.

gaik-0.2.6/pyproject.toml

@@ -0,0 +1,72 @@
+[project]
+name = "gaik"
+version = "0.2.6"
+description = "General AI Kit - Reusable AI/ML components for Python"
+readme = "README.md"
+requires-python = ">=3.10"
+license = { file = "LICENSE" }
+authors = [{ name = "GAIK Project" }]
+keywords = [
+    "ai",
+    "ml",
+    "langchain",
+    "openai",
+    "anthropic",
+    "google",
+    "structured-outputs",
+    "pydantic",
+    "schema",
+    "extraction",
+]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Software Development :: Libraries :: Python Modules",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]
+
+dependencies = [
+    "pydantic>=2.12.3",
+    "langchain-core>=1.0.3",
+    "langchain-openai>=1.0.2",
+    "langchain-anthropic>=1.0.1",
+    "langchain-google-genai>=3.0.1",
+]
+
+[project.optional-dependencies]
+dev = ["ruff>=0.14.1", "build>=1.0", "twine>=4.0"]
+vision = [
+    "openai>=1.40.0",
+    "pdf2image>=1.17.0",
+    "pillow>=10.0.0",
+    "python-dotenv>=1.0.0",
+]
+
+[project.urls]
+Homepage = "https://gaik.ai/"
+Repository = "https://github.com/GAIK-project/toolkit-shared-components"
+Documentation = "https://github.com/GAIK-project/toolkit-shared-components/tree/main/gaik-py"
+Issues = "https://github.com/GAIK-project/toolkit-shared-components/issues"
+
+[build-system]
+requires = ["setuptools>=61.0", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.setuptools.package-data]
+gaik = ["py.typed"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py310"
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "N", "W", "UP"]
+ignore = []

gaik-0.2.6/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

gaik-0.2.6/src/gaik/__init__.py

@@ -0,0 +1,35 @@
+"""General AI Kit (GAIK) - Reusable AI/ML components for Python.
+
+GAIK provides modular, production-ready tools for common AI/ML tasks including:
+- Dynamic data extraction with structured outputs
+- Multi-provider LLM support (OpenAI, Anthropic, Azure, Google)
+- And more modules coming soon...
+
+Available modules:
+    - gaik.extract: Dynamic data extraction with LangChain structured outputs
+    - gaik.providers: Multi-provider LLM interface (OpenAI, Anthropic, Azure, Google)
+    - gaik.parsers: Vision-enabled PDF to Markdown parsing utilities
+
+Example:
+    >>> from gaik.extract import SchemaExtractor
+    >>>
+    >>> # Using default OpenAI provider
+    >>> extractor = SchemaExtractor("Extract title and date from articles")
+    >>> results = extractor.extract(documents)
+    >>>
+    >>> # Using Anthropic Claude
+    >>> # IDE autocomplete shows: "openai" | "anthropic" | "google" | "azure"
+    >>> extractor = SchemaExtractor(
+    ...     "Extract name and age",
+    ...     provider="anthropic"
+    ... )
+"""
+
+import importlib.metadata
+
+try:
+    __version__ = importlib.metadata.version("gaik")
+except importlib.metadata.PackageNotFoundError:
+    __version__ = "0.0.0.dev"
+
+__all__ = ["__version__"]

gaik-0.2.6/src/gaik/extract/__init__.py

@@ -0,0 +1,69 @@
+"""Dynamic data extraction with OpenAI structured outputs.
+
+This module provides tools for extracting structured data from unstructured text
+using dynamically created Pydantic schemas and OpenAI's structured outputs API.
+
+Benefits of this approach:
+- Type-safe and guaranteed structure (enforced by the API)
+- Cost-effective (fewer tokens, no code generation)
+- Secure (no eval/exec needed)
+- Simple and maintainable
+- Reliable results with automatic retries
+
+Quick Start:
+    >>> from gaik.extract import dynamic_extraction_workflow
+    >>>
+    >>> results = dynamic_extraction_workflow(
+    ...     user_description="Extract title, date, and author from articles",
+    ...     documents=[doc1, doc2, doc3]
+    ... )
+
+Advanced Usage:
+    >>> from gaik.extract import SchemaExtractor
+    >>>
+    >>> # Reuse the same schema for multiple batches
+    >>> extractor = SchemaExtractor("Extract invoice number and amount")
+    >>> batch1 = extractor.extract(documents1)
+    >>> batch2 = extractor.extract(documents2)
+    >>>
+    >>> # Access the generated Pydantic model
+    >>> schema = extractor.model.model_json_schema()
+    >>> print(schema)
+
+Custom Field Specifications:
+    >>> from gaik.extract import (
+    ...     FieldSpec,
+    ...     ExtractionRequirements,
+    ...     create_extraction_model,
+    ... )
+    >>>
+    >>> fields = [
+    ...     FieldSpec(
+    ...         field_name="invoice_number",
+    ...         field_type="str",
+    ...         description="Extract invoice ID",
+    ...         required=True
+    ...     )
+    ... ]
+    >>> requirements = ExtractionRequirements(
+    ...     use_case_name="Invoice",
+    ...     fields=fields
+    ... )
+    >>> model = create_extraction_model(requirements)
+"""
+
+from gaik.extract.extractor import SchemaExtractor, dynamic_extraction_workflow
+from gaik.extract.models import ExtractionRequirements, FieldSpec
+from gaik.extract.utils import create_extraction_model, sanitize_model_name
+
+__all__ = [
+    # Main API
+    "SchemaExtractor",
+    "dynamic_extraction_workflow",
+    # Models
+    "FieldSpec",
+    "ExtractionRequirements",
+    # Utilities
+    "create_extraction_model",
+    "sanitize_model_name",
+]