ngpt 2.15.2__tar.gz → 2.16.1__tar.gz

{ngpt-2.15.2 → ngpt-2.16.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ngpt
-Version: 2.15.2
+Version: 2.16.1
 Summary: A lightweight Python CLI and library for interacting with OpenAI-compatible APIs, supporting both official and self-hosted LLM endpoints.
 Project-URL: Homepage, https://github.com/nazdridoy/ngpt
 Project-URL: Repository, https://github.com/nazdridoy/ngpt
@@ -41,7 +41,7 @@ Description-Content-Type: text/markdown
 [![Python Versions](https://img.shields.io/pypi/pyversions/ngpt.svg)](https://pypi.org/project/ngpt/)
 [![Documentation](https://img.shields.io/badge/docs-available-brightgreen.svg)](https://nazdridoy.github.io/ngpt/)
 
-A lightweight Python CLI and library for interacting with OpenAI-compatible APIs, supporting both official and self-hosted LLM endpoints.
+A versatile Python LLM client that functions as a CLI utility, library, and development framework. Supports multiple providers (OpenAI, Ollama, Groq, Claude) with features including interactive chat, code generation, shell command execution, git commit messages generation, and markdown rendering.
 
 ![2025-04-23_16-18-01](https://github.com/user-attachments/assets/b8e58926-5165-4352-b48b-9f4a982da86e)
 
@@ -62,6 +62,9 @@ A lightweight Python CLI and library for interacting with OpenAI-compatible APIs
   - [Interactive Configuration](#interactive-configuration)
   - [Configuration File](#configuration-file)
   - [Configuration Priority](#configuration-priority)
+  - [API Key Setup](#api-key-setup)
+  - [OpenAI API Key](#openai-api-key)
+  - [Google Gemini API Key](#google-gemini-api-key)
 - [Contributing](#contributing)
 - [License](#license)
 
@@ -397,6 +400,35 @@ For detailed information about building CLI tools with nGPT, see the [CLI Framew
 
 ## Configuration
 
+### API Key Setup
+
+#### OpenAI API Key
+1. Create an account at [OpenAI](https://platform.openai.com/)
+2. Navigate to API keys: https://platform.openai.com/api-keys
+3. Click "Create new secret key" and copy your API key
+4. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: OpenAI
+   # Enter API key: your-openai-api-key
+   # Enter base URL: https://api.openai.com/v1/
+   # Enter model: gpt-3.5-turbo (or other model)
+   ```
+
+#### Google Gemini API Key
+1. Create or use an existing Google account
+2. Go to [Google AI Studio](https://aistudio.google.com/)
+3. Navigate to API keys in the left sidebar (or visit https://aistudio.google.com/app/apikeys)
+4. Create an API key and copy it
+5. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: Gemini
+   # Enter API key: your-gemini-api-key
+   # Enter base URL: https://generativelanguage.googleapis.com/v1beta/openai
+   # Enter model: gemini-2.0-flash
+   ```
+
 ### Command Line Options
 
 You can configure nGPT using the following options:

{ngpt-2.15.2 → ngpt-2.16.1}/README.md

@@ -5,7 +5,7 @@
 [![Python Versions](https://img.shields.io/pypi/pyversions/ngpt.svg)](https://pypi.org/project/ngpt/)
 [![Documentation](https://img.shields.io/badge/docs-available-brightgreen.svg)](https://nazdridoy.github.io/ngpt/)
 
-A lightweight Python CLI and library for interacting with OpenAI-compatible APIs, supporting both official and self-hosted LLM endpoints.
+A versatile Python LLM client that functions as a CLI utility, library, and development framework. Supports multiple providers (OpenAI, Ollama, Groq, Claude) with features including interactive chat, code generation, shell command execution, git commit messages generation, and markdown rendering.
 
 ![2025-04-23_16-18-01](https://github.com/user-attachments/assets/b8e58926-5165-4352-b48b-9f4a982da86e)
 
@@ -26,6 +26,9 @@ A lightweight Python CLI and library for interacting with OpenAI-compatible APIs
   - [Interactive Configuration](#interactive-configuration)
   - [Configuration File](#configuration-file)
   - [Configuration Priority](#configuration-priority)
+  - [API Key Setup](#api-key-setup)
+  - [OpenAI API Key](#openai-api-key)
+  - [Google Gemini API Key](#google-gemini-api-key)
 - [Contributing](#contributing)
 - [License](#license)
 
@@ -361,6 +364,35 @@ For detailed information about building CLI tools with nGPT, see the [CLI Framew
 
 ## Configuration
 
+### API Key Setup
+
+#### OpenAI API Key
+1. Create an account at [OpenAI](https://platform.openai.com/)
+2. Navigate to API keys: https://platform.openai.com/api-keys
+3. Click "Create new secret key" and copy your API key
+4. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: OpenAI
+   # Enter API key: your-openai-api-key
+   # Enter base URL: https://api.openai.com/v1/
+   # Enter model: gpt-3.5-turbo (or other model)
+   ```
+
+#### Google Gemini API Key
+1. Create or use an existing Google account
+2. Go to [Google AI Studio](https://aistudio.google.com/)
+3. Navigate to API keys in the left sidebar (or visit https://aistudio.google.com/app/apikeys)
+4. Create an API key and copy it
+5. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: Gemini
+   # Enter API key: your-gemini-api-key
+   # Enter base URL: https://generativelanguage.googleapis.com/v1beta/openai
+   # Enter model: gemini-2.0-flash
+   ```
+
 ### Command Line Options
 
 You can configure nGPT using the following options:

{ngpt-2.15.2 → ngpt-2.16.1}/docs/README.md

@@ -1,6 +1,6 @@
 # nGPT Documentation
 
-Welcome to the documentation for nGPT, a lightweight Python CLI and library for interacting with OpenAI-compatible APIs, supporting both official and self-hosted LLM endpoints. nGPT offers three main modes of use:
+Welcome to the documentation for nGPT, a versatile Python LLM client that functions as a CLI utility, library, and development framework. It supports multiple providers (OpenAI, Ollama, Groq, Claude) with features including interactive chat, code generation, shell command execution, git commit messages generation, and markdown rendering. nGPT offers three main modes of use:
 
 - A command-line interface (CLI) tool for quick AI interactions
 - A Python library for integration into your applications

{ngpt-2.15.2 → ngpt-2.16.1}/docs/configuration.md

@@ -2,6 +2,35 @@
 
 nGPT uses a flexible configuration system that supports multiple profiles for different API providers and models. This guide explains how to configure and manage your nGPT settings.
 
+## API Key Setup
+
+### OpenAI API Key
+1. Create an account at [OpenAI](https://platform.openai.com/)
+2. Navigate to API keys: https://platform.openai.com/api-keys
+3. Click "Create new secret key" and copy your API key
+4. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: OpenAI
+   # Enter API key: your-openai-api-key
+   # Enter base URL: https://api.openai.com/v1/
+   # Enter model: gpt-3.5-turbo (or other model)
+   ```
+
+### Google Gemini API Key
+1. Create or use an existing Google account
+2. Go to [Google AI Studio](https://aistudio.google.com/)
+3. Navigate to API keys in the left sidebar (or visit https://aistudio.google.com/app/apikeys)
+4. Create an API key and copy it
+5. Configure nGPT with your key:
+   ```bash
+   ngpt --config
+   # Enter provider: Gemini
+   # Enter API key: your-gemini-api-key
+   # Enter base URL: https://generativelanguage.googleapis.com/v1beta/openai
+   # Enter model: gemini-2.0-flash
+   ```
+
 ## Configuration File Location
 
 nGPT stores its configuration in a JSON file located at:
@@ -17,7 +46,7 @@ The configuration file uses a JSON list format that allows you to store multiple
 ```json
 [
   {
-    "api_key": "your-openai-api-key-here",
+    "api_key": "your-openai-api-key",
     "base_url": "https://api.openai.com/v1/",
     "provider": "OpenAI",
     "model": "gpt-4o"

{ngpt-2.15.2 → ngpt-2.16.1}/docs/overview.md

@@ -2,7 +2,7 @@
 
 ## What is nGPT?
 
-nGPT is a lightweight Python library and command-line interface (CLI) tool designed for interacting with OpenAI-compatible APIs. It provides a simple, flexible way to communicate with various large language model (LLM) endpoints, including official OpenAI services and self-hosted alternatives.
+nGPT is a versatile Python LLM client that functions as a CLI utility, library, and development framework. It supports multiple providers (OpenAI, Ollama, Groq, Claude) with features including interactive chat, code generation, shell command execution, git commit messages generation, and markdown rendering.
 
 ## Key Features
 

{ngpt-2.15.2 → ngpt-2.16.1}/ngpt/cli/modes/gitcommsg.py

@@ -4,6 +4,7 @@ import sys
 import tempfile
 import time
 import subprocess
+import threading
 from datetime import datetime
 import logging
 from ..formatters import COLORS
@@ -498,12 +499,27 @@ def process_with_chunking(client, diff_content, context, chunk_size=200, recursi
             logger.log_template("DEBUG", "CHUNK", chunk_prompt)
         
         # Process chunk - use technical system prompt for analysis
-        print(f"{COLORS['yellow']}Analyzing changes...{COLORS['reset']}")
+        stop_spinner = threading.Event()
+        spinner_thread = threading.Thread(
+            target=spinner, 
+            args=("Analyzing changes...",), 
+            kwargs={"stop_event": stop_spinner, "color": COLORS['yellow']}
+        )
+        spinner_thread.daemon = True
+        spinner_thread.start()
+        
         try:
             result = handle_api_call(client, chunk_prompt, technical_system_prompt, logger)
-            partial_analyses.append(result)
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
+            # Show success message
             print(f"{COLORS['green']}✓ Chunk {i+1} processed{COLORS['reset']}")
+            partial_analyses.append(result)
         except Exception as e:
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
             print(f"{COLORS['red']}Error processing chunk {i+1}: {str(e)}{COLORS['reset']}")
             if logger:
                 logger.error(f"Error processing chunk {i+1}: {str(e)}")
@@ -511,7 +527,7 @@ def process_with_chunking(client, diff_content, context, chunk_size=200, recursi
         
         # Rate limit protection between chunks
         if i < chunk_count - 1:
-            # Use the spinner function
+            # Use the spinner function with fixed duration
             spinner("Waiting to avoid rate limits...", 5, color=COLORS['yellow'])
     
     # Combine partial analyses
@@ -539,7 +555,16 @@ def process_with_chunking(client, diff_content, context, chunk_size=200, recursi
         )
     else:
         # Combined analysis is under the chunk size limit, generate the commit message
-        print(f"{COLORS['green']}Generating commit message from combined analysis...{COLORS['reset']}")
+        # Start spinner for generating commit message
+        stop_spinner = threading.Event()
+        spinner_thread = threading.Thread(
+            target=spinner, 
+            args=("Generating commit message from combined analysis...",), 
+            kwargs={"stop_event": stop_spinner, "color": COLORS['green']}
+        )
+        spinner_thread.daemon = True
+        spinner_thread.start()
+        
         combine_prompt = create_combine_prompt(partial_analyses)
         
         # Log combine template
@@ -550,6 +575,10 @@ def process_with_chunking(client, diff_content, context, chunk_size=200, recursi
             # Use commit message system prompt for final generation
             commit_message = handle_api_call(client, combine_prompt, commit_system_prompt, logger)
             
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
+            
             # If the commit message is too long, we need to condense it
             if len(commit_message.splitlines()) > max_msg_lines:
                 return condense_commit_message(
@@ -563,6 +592,10 @@ def process_with_chunking(client, diff_content, context, chunk_size=200, recursi
                 )
             return commit_message
         except Exception as e:
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
+            
             print(f"{COLORS['red']}Error combining analyses: {str(e)}{COLORS['reset']}")
             if logger:
                 logger.error(f"Error combining analyses: {str(e)}")
@@ -683,18 +716,33 @@ SECTION OF ANALYSIS TO CONDENSE:
         if logger:
             logger.log_template("DEBUG", f"CONDENSE_ANALYSIS_DEPTH_{current_depth}_CHUNK_{i+1}", condense_prompt)
         
-        print(f"{COLORS['yellow']}Condensing analysis chunk {i+1}/{analysis_chunk_count}...{COLORS['reset']}")
+        # Start spinner for analysis
+        stop_spinner = threading.Event()
+        spinner_thread = threading.Thread(
+            target=spinner, 
+            args=(f"Condensing analysis chunk {i+1}/{analysis_chunk_count}...",), 
+            kwargs={"stop_event": stop_spinner, "color": COLORS['yellow']}
+        )
+        spinner_thread.daemon = True
+        spinner_thread.start()
         
         # Condense this analysis chunk - use technical system prompt for condensing analysis
         try:
             condensed_chunk = handle_api_call(client, condense_prompt, technical_system_prompt, logger)
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
+            
+            print(f"{COLORS['green']}✓ Analysis chunk {i+1}/{analysis_chunk_count} condensed{COLORS['reset']}")
             condensed_chunks.append(condensed_chunk)
             
             if logger:
                 logger.log_content("DEBUG", f"CONDENSED_ANALYSIS_DEPTH_{current_depth}_CHUNK_{i+1}", condensed_chunk)
-                
-            print(f"{COLORS['green']}✓ Analysis chunk {i+1}/{analysis_chunk_count} condensed{COLORS['reset']}")
         except Exception as e:
+            # Stop the spinner
+            stop_spinner.set()
+            spinner_thread.join()
+            
             print(f"{COLORS['red']}Error condensing analysis chunk {i+1}: {str(e)}{COLORS['reset']}")
             if logger:
                 logger.error(f"Error condensing analysis chunk {i+1} at depth {current_depth}: {str(e)}")
@@ -702,7 +750,7 @@ SECTION OF ANALYSIS TO CONDENSE:
         
         # Rate limit protection between chunks
         if i < analysis_chunk_count - 1:
-            # Use the spinner function
+            # Use the spinner function with fixed duration
             spinner("Waiting to avoid rate limits...", 5, color=COLORS['yellow'])
     
     # Combine condensed chunks
@@ -812,10 +860,21 @@ REQUIREMENTS:
     if logger:
         logger.log_template("DEBUG", f"CONDENSE_PROMPT_DEPTH_{current_depth}", condense_prompt)
     
-    print(f"{COLORS['yellow']}Condensing commit message (depth {current_depth}/{max_recursion_depth})...{COLORS['reset']}")
+    # Start spinner for condensing
+    stop_spinner = threading.Event()
+    spinner_thread = threading.Thread(
+        target=spinner, 
+        args=(f"Condensing commit message (depth {current_depth}/{max_recursion_depth})...",), 
+        kwargs={"stop_event": stop_spinner, "color": COLORS['yellow']}
+    )
+    spinner_thread.daemon = True
+    spinner_thread.start()
     
     try:
         condensed_result = handle_api_call(client, condense_prompt, system_prompt, logger)
+        # Stop the spinner
+        stop_spinner.set()
+        spinner_thread.join()
         
         if logger:
             logger.log_content("DEBUG", f"CONDENSED_RESULT_DEPTH_{current_depth}", condensed_result)
@@ -842,6 +901,10 @@ REQUIREMENTS:
         else:
             return condensed_result
     except Exception as e:
+        # Stop the spinner
+        stop_spinner.set()
+        spinner_thread.join()
+        
         print(f"{COLORS['red']}Error condensing commit message: {str(e)}{COLORS['reset']}")
         if logger:
             logger.error(f"Error condensing commit message at depth {current_depth}: {str(e)}")

{ngpt-2.15.2 → ngpt-2.16.1}/ngpt/cli/ui.py

@@ -154,14 +154,16 @@ def get_multiline_input():
             print("\nInput cancelled by user. Exiting gracefully.")
             return None
 
-def spinner(message, duration=5, spinner_chars="⣾⣽⣻⢿⡿⣟⣯⣷", color=None):
+def spinner(message, duration=5, spinner_chars="⣾⣽⣻⢿⡿⣟⣯⣷", color=None, stop_event=None):
     """Display a spinner animation with a message.
     
     Args:
         message: The message to display alongside the spinner
-        duration: Duration in seconds to show the spinner
+        duration: Duration in seconds to show the spinner (used if stop_event is None)
         spinner_chars: Characters to use for the spinner animation
         color: Optional color from COLORS dict to use for the message
+        stop_event: Optional threading.Event to signal when to stop the spinner
+                   If provided, duration is ignored and spinner runs until event is set
     """
     # Default color handling
     color_start = ""
@@ -172,16 +174,24 @@ def spinner(message, duration=5, spinner_chars="⣾⣽⣻⢿⡿⣟⣯⣷", color
     
     # Each character shows for 0.2 seconds
     char_duration = 0.2
-    # Total number of characters to show (not iterations through the entire spinner sequence)
-    total_chars = int(duration / char_duration)
     
-    # Run the spinner
-    for i in range(total_chars):
-        # Get the appropriate character by cycling through the spinner characters
-        char = spinner_chars[i % len(spinner_chars)]
-        sys.stdout.write(f"\r{color_start}{message} {char}{color_end}")
-        sys.stdout.flush()
-        time.sleep(char_duration)
+    if stop_event:
+        # Run until stop_event is set
+        i = 0
+        while not stop_event.is_set():
+            char = spinner_chars[i % len(spinner_chars)]
+            sys.stdout.write(f"\r{color_start}{message} {char}{color_end}")
+            sys.stdout.flush()
+            i += 1
+            time.sleep(char_duration)
+    else:
+        # Run for fixed duration
+        total_chars = int(duration / char_duration)
+        for i in range(total_chars):
+            char = spinner_chars[i % len(spinner_chars)]
+            sys.stdout.write(f"\r{color_start}{message} {char}{color_end}")
+            sys.stdout.flush()
+            time.sleep(char_duration)
     
     # Clear the line when done
     sys.stdout.write("\r" + " " * (len(message) + 10) + "\r")

{ngpt-2.15.2 → ngpt-2.16.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ngpt"
-version = "2.15.2"
+version = "2.16.1"
 description = "A lightweight Python CLI and library for interacting with OpenAI-compatible APIs, supporting both official and self-hosted LLM endpoints."
 authors = [
     {name = "nazDridoy", email = "nazdridoy399@gmail.com"},

{ngpt-2.15.2 → ngpt-2.16.1}/uv.lock

@@ -134,7 +134,7 @@ wheels = [
 
 [[package]]
 name = "ngpt"
-version = "2.15.2"
+version = "2.16.1"
 source = { editable = "." }
 dependencies = [
     { name = "prompt-toolkit" },