fastshot 1.0.2

A versatile screen capturing tool with annotation and OCR features

Fastshot

Fastshot is an open-source Python-based screenshot tool for Windows, inspired by Snipaste. It provides a versatile and user-friendly interface for capturing and annotating screenshots, making it ideal for research, reading papers, writing code, and various comparison and demonstration scenarios.

Features

• Always on Top: Screenshots stay on top, allowing easy comparison.
• Smooth Zoom and Drag: Effortlessly zoom and drag screenshots.
• Annotation Tools: Hand-drawing and text mode for annotations.
• Clipboard Export: Easily export screenshots to the clipboard.
• OCR Integration: Extract text from images using the built-in OCR plugin powered by PaddleOCR.

Installation

You can install Fastshot from PyPI:

pip install fastshot

Usage

Once installed, you can start Fastshot from the command line:

fastshot

Hotkeys

Ctrl+F1: Activate screen capturing mode.

Ctrl+P: Activate paint mode.

Ctrl+T: Activate text mode.

Esc: Exit the current mode.

Right-Click Menu

❌ Close: Close the current window.

💾 Save As...: Save the current screenshot.

🖌️ Paint: Activate paint mode.

↩️ Undo: Undo the last action.

🚪 Exit Edit: Exit paint or text mode.

📋 Copy: Copy the current screenshot to the clipboard.

🔤 Text: Activate text mode.

🔍 OCR: Perform OCR on the current screenshot and copy the result to the clipboard.

Plugin Development

Fastshot supports a plugin mechanism that allows developers to extend its functionality. Here is a brief description of how to develop a plugin:

1. Create a Plugin Class: Your plugin should be a Python class with the desired functionality. For example, an OCR plugin might look like this:

   from paddleocr import PaddleOCR
   from PIL import Image
   import win32clipboard
   import tkinter as tk
   
   class PluginOCR:
       def __init__(self):
           self.ocr_engine = PaddleOCR(use_angle_cls=True, lang='en')
   
       def ocr(self, image):
           result = self.ocr_engine.ocr(image, cls=True)
           ocr_text = "\n".join([line[1][0] for res in result for line in res])
           self.copy_to_clipboard(ocr_text)
           return ocr_text
   
       def copy_to_clipboard(self, text):
           win32clipboard.OpenClipboard()
           win32clipboard.EmptyClipboard()
           win32clipboard.SetClipboardText(text, win32clipboard.CF_UNICODETEXT)
           win32clipboard.CloseClipboard()
   
       def show_message(self, message, parent):
           label = tk.Label(parent, text=message, bg="yellow", fg="black", font=("Helvetica", 10))
           label.pack(side="bottom", fill="x")
           parent.after(3000, label.destroy)
   

2. Register the Plugin: In the SnipasteApp class, you can register your plugin by adding it to the plugin list.

   class SnipasteApp:
       def load_plugins(self):
           plugin_modules = ['fastshot.plugin_ocr']  # Add your plugin module here
           for module_name in plugin_modules:
               module = importlib.import_module(module_name)
               plugin_class = getattr(module, 'PluginOCR')
               self.plugins[module_name] = plugin_class()
   

3. Invoke the Plugin: You can invoke the plugin from your application code, such as from a menu item.

   def ocr(self):
       plugin = self.app.plugins.get('fastshot.plugin_ocr')
       if plugin:
           img_path = 'temp.png'
           self.img_label.zoomed_image.save(img_path)
           result = plugin.ocr(img_path)
           plugin.show_message("OCR result updated in clipboard", self.img_window)
   

By following these steps, you can create and integrate custom plugins to extend the functionality of Fastshot.

Development

Setting Up the Development Environment

1. Clone the repository:

git clone https://github.com/yourusername/fastshot.git
cd fastshot

2. Install the dependencies:

pip install -r requirements.txt

Running Tests

You can run the tests using:

pytest tests/

Contributing

We welcome contributions! Please read our Contributing Guidelines for more details.

License

This project is licensed under the Apache License - see the LICENSE file for details.

Todo:

1. tk window force trigger.
2. ppocr[Default]
3. screenpen integration
4. hyder
5. transprent window
6. fixed on top
7. pyinstaller.
8. gdt-4o multimodal task(OCR/QA)AS