feat: video.clip implementation

[omgate234]

Pull Request

Description:
Add video.clip() method to generate video clips based on text prompts using AI analysis of video
content (transcript, scenes, or both).

Changes:

• Add ClipContentType class with spoken, visual, multimodal options in _constants.py
• Add clip to ApiPath in _constants.py
• Add clip() method to Video class in video.py
• Export ClipContentType in __init__.py

Related Issues:

• Implements SDK support for POST /video/{video_id}/clip endpoint

Testing:

from videodb import connect, ClipContentType                                                            
                                                                                                        
conn = connect()
coll = conn.get_collection()                                                                                        
video = coll.get_video("m-xxxxx")                                                                       
                                                                                                        
# Test spoken content (default)                                                                         
clips = video.clip("find the introduction")                                                             
                                                                                                        
# Test visual content                                                                                   
clips = video.clip("find action scenes", content_type=ClipContentType.visual)                           
                                                                                                        
# Test multimodal                                                                                       
clips = video.clip("find funny moments", content_type=ClipContentType.multimodal)                       
                                                                                                        
# Verify Shot objects are returned with correct attributes                                              
for shot in clips:                                                                                      
    print(shot.start, shot.end, shot.text)