Merge pull request #39 from renan-siqueira/develop

Develop -> main

Author: renan-siqueira

CHANGELOG.md

@@ -3,6 +3,27 @@
 All notable changes to this project will be documented in this file.
 ___
 
+## [0.6.0] - 2025-04-05
+
+### Added
+- Introduced a fully modular architecture under `src/rsazure_openai_toolkit/`, with well-defined packages: `core/`, `cli/`, `env/`, `model_config/`, `models/`, `session/`, `samples/`, `utils/`
+- Created `ChatCLI` and `ToolsCLI` classes to encapsulate the logic of `rschat` and `rschat-tools` commands
+- Added `ModelConfig` class and enhanced `get_model_config()` function for flexible, reproducible model parameter configuration
+- Introduced `ContextInfo` and `ChatResult` as structured output representations in the `models/` module
+- Session manager now supports full history via `.full.jsonl`, backup of metadata, and safe system prompt/config overrides
+
+### Changed
+- Refactored all CLI logic into object-oriented classes (`cli.py`, `tools.py`)
+- Centralized helper functions in a single `utils/` module
+- Renamed sample templates to a clearer format (`chat_loop_usage.py.j2`, etc.)
+- Updated `pyproject.toml` to reflect new package structure and layout
+- Improved CLI feedback: now prints context load info, system prompt in use, token usage, and fallback logging status
+- Bumped version to 0.6.0
+
+### Removed
+- Removed deprecated and fragmented files: `handler.py`, `integration.py`, `token_utils.py`, `model_config_utils.py`
+___
+
 ## [0.5.2] - 2025-04-03
 
 ### Fixed

README.md

@@ -53,7 +53,17 @@
   </a>
 </h1>
 
-A fast, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture.
+A fast, modular, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture.
+___
+
+## 🚀 What's New in v0.6.0
+
+- Fully modular codebase with dedicated folders for each responsibility
+- Object-oriented CLI (`rschat`, `rschat-tools`) for better maintainability
+- Persistent session context with full/trimmed history and safety validation
+- Unified `utils/` module and improved logging/debug output
+
+> Check the full [CHANGELOG](https://github.com/renan-siqueira/rsazure-openai-toolkit/blob/main/CHANGELOG.md) for details.
 ___
 
 ## 📖 Documentation

RELEASING.md

@@ -69,6 +69,7 @@ git push origin v<new_version>
 
 ## 📦 5. Build Distributable Packages
 ```bash
+find . -type d -name "__pycache__" -exec rm -r {} +
 rm -rf dist/
 python -m build
 ```

SECURITY.md

@@ -49,10 +49,23 @@ This toolkit supports optional **context persistence** via `SessionContext`, whi
 - Includes message history, system prompt, and model configuration
 - All context storage is **opt-in**, triggered via environment variables
 - Full history is saved to a separate `.full.jsonl` file, even if context is trimmed
+- Metadata changes (e.g., prompt overrides) are automatically backed up with timestamped `.bak` files
 
 > ⚠️ Use with care on shared machines. Avoid using sensitive data in prompts or responses if local storage is enabled.
 ___
 
+## 🔐 Security-minded Architecture
+
+The modular design of this toolkit (since v0.6.0) reinforces its commitment to safe, auditable, and professional AI integration:
+
+- No shared state or global mutability across modules
+- Clear separation of concerns (e.g., `env`, `logging`, `model_config`, `session`)
+- CLI and utility components validate inputs explicitly and fail gracefully
+- Configuration and history are saved locally with full user control
+
+You are always in charge of what gets stored, when, and how.
+___
+
 ## 📣 Reporting a Vulnerability
 
 If you discover a potential security issue in this project, please **report it responsibly**:

docs/config.md

@@ -30,7 +30,7 @@ ___
 If you call `get_model_config()` with no arguments, you get:
 
 ```python
-from rsazure_openai_toolkit.utils.model_config_utils import get_model_config
+from rsazure_openai_toolkit.utils import get_model_config
 
 model_config = get_model_config()
 

pyproject.toml

@@ -4,8 +4,8 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "rsazure-openai-toolkit"
-version = "0.5.2"
-description = "A fast, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture."
+version = "0.6.0"
+description = "A fast, modular, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture."
 authors = [{ name = "Renan Siqueira Antonio", email = "[email protected]" }]
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.9"
@@ -37,7 +37,7 @@ classifiers = [
 
 [project.scripts]
 rschat = "rsazure_openai_toolkit.cli:cli"
-rschat-tools = "rsazure_openai_toolkit.tools_cli:main"
+rschat-tools = "rsazure_openai_toolkit.cli.tools:main"
 
 [project.urls]
 "Homepage" = "https://github.com/renan-siqueira/rsazure-openai-toolkit"

requirements.txt

@@ -2,15 +2,97 @@
 rsazure_openai_toolkit
 
 A fast, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture.
+
+Top-level exports available when importing as `import rsazure_openai_toolkit as rschat`:
+
+# Core interaction
+- rschat.main(...)
+- rschat.generate_response(...)
+
+# Configuration & utility
+- rschat.get_model_config(...)
+- rschat.ModelConfig(...)
+- rschat.estimate_input_tokens(...)
+- rschat.load_env()
+- rschat.get_cli_config()
+
+# Session context
+- rschat.SessionContext(...)
+- rschat.get_context_messages(...)
+
+# Logging
+- rschat.InteractionLogger(...)
+- rschat.get_logger(...)
+
+# CLI Entrypoints
+- rschat.cli(...)
+- rschat.main_cli_tools(...)
+
+# Results and metadata
+- rschat.ChatResult(...)
+- rschat.ContextInfo(...)
 """
 
-__version__ = "0.5.2"
+__version__ = "0.6.0"
+__description__ = "A fast, modular, secure, and auditable toolkit to integrate with Azure OpenAI — with a friendly CLI and dev-first architecture."
+__author__ = "Renan Siqueira Antonio"
+__license__ = "MIT"
+__status__ = "Beta"
+__url__ = "https://github.com/renan-siqueira/rsazure-openai-toolkit"
+__docs__ = "https://github.com/renan-siqueira/rsazure-openai-toolkit/tree/main/docs"
+__security_policy_url__ = "https://github.com/renan-siqueira/rsazure-openai-toolkit/security/policy"
+
+
+# Core interaction
+from .core import main, generate_response
+
+# Utility
+from .utils import estimate_input_tokens
+
+# Environment
+from .env import load_env, get_cli_config
+
+# Model configuration
+from .model_config import get_model_config, ModelConfig
+
+# Session management
+from .session import SessionContext, get_context_messages
+
+# Logging
+from .logging import InteractionLogger, get_logger
+
+# CLI entrypoints
+from .cli import cli, tools_main
+
+# Result models
+from .models import ChatResult, ContextInfo
 
-from .handler import call_azure_openai_handler
-from .integration import generate_response, load_azure_client
 
 __all__ = [
-    "call_azure_openai_handler",
+    # Core OpenAI interaction
+    "main",
     "generate_response",
-    "load_azure_client",
+
+    # Config & utils
+    "get_model_config",
+    "ModelConfig",
+    "estimate_input_tokens",
+    "load_env",
+    "get_cli_config",
+
+    # Context management
+    "SessionContext",
+    "get_context_messages",
+
+    # Logging
+    "InteractionLogger",
+    "get_logger",
+
+    # CLI entrypoints
+    "cli",
+    "tools_main",
+
+    # Result and metadata representations
+    "ChatResult",
+    "ContextInfo",
 ]

src/rsazure_openai_toolkit/__init__.py

@@ -0,0 +1,16 @@
+"""
+CLI utilities for running and extending rschat functionality.
+
+Modules:
+- cli: Main CLI entrypoint used in the command-line interface.
+- tools: Developer-oriented tools such as sample generation.
+"""
+
+from .cli import cli
+from .tools import main as tools_main
+
+
+__all__ = [
+    "cli",
+    "tools_main"
+]