<Codebase_Information>
Python-based sentiment analysis tool for Digital research and social media monitoring using OpenAI GPT models with optional Brandwatch integration - only used by internal digital team.
<Logic_Overview>
Broadly, there are two main processing paths:
1. Sentiment Analysis → (optional) Brandwatch Upload
main.py → input_config.py → connector_functions.py → file_operations.py →
[optional: multi_company_analysis.py] → async_core_logic.py (routing and batch processing) →
file_operations.py → [optional: bw_api_handling.py (Brandwatch API)]
2. Direct Brandwatch Upload
main.py → bw_upload_only.py → file_operations.py →
bw_api_handling.py (Brandwatch API)
Core modules:
- main.py: GUI interface and initialization
- input_config.py: Configuration state management and validation
- connector_functions.py: Main workflow orchestration and thread management
- file_operations.py: File I/O and format handling (used throughout)
- multi_company_analysis.py: Optional pre-processing for company-specific analysis
- async_core_logic.py: Routing and async batch processing and batch handling
- bw_api_handling.py: Brandwatch API integration
- bw_upload_only.py: Standalone Brandwatch upload workflow
Uses threading for GUI non-blocking operations and asyncio for OpenAI API calls (using aiohttp requests, not the OpenAI Python SDK).
log_message and update_progress_gui are threadsafe callback functions used to display messages and update the progress bar in the Tkinter GUI, respectively.
</Logic_Overview>
<Usage>
- This tool is packaged with pyinstaller and --onedir and saved in a Sharepoint folder.
- The executable is run by team members directly from the synced Sharepoint folder in file explorer, without them having to install anything.
</Usage>
</Codebase_Information>
openai
python
