YoutubeCommentsAnalyzer

import json
import requests

def get_all_comments(video_id, api_key):
    comments = []
    url = f"https://www.googleapis.com/youtube/v3/commentThreads?part=snippet&videoId={video_id}&key={api_key}&maxResults=100"
    while url:
        response = requests.get(url)
        data = response.json()
        for item in data["items"]:
            comment = {
                "text": item["snippet"]["topLevelComment"]["snippet"]["textDisplay"],
                "likeCount": item["snippet"]["topLevelComment"]["snippet"]["likeCount"],
                "publishedAt": item["snippet"]["topLevelComment"]["snippet"][
                    "publishedAt"
                ],
                "authorDisplayName": item["snippet"]["topLevelComment"]["snippet"][
                    "authorDisplayName"
                ],
                "replies": [],
            }
            if "replies" in item:
                for reply in item["replies"]["comments"]:
                    comment["replies"].append(
                        {
                            "text": reply["snippet"]["textDisplay"],
                            "likeCount": reply["snippet"]["likeCount"],
                            "publishedAt": reply["snippet"]["publishedAt"],
                            "authorDisplayName": reply["snippet"]["authorDisplayName"],
                        }
                    )
            comments.append(comment)
        url = data.get("nextPageToken")
        if url:
            url = f"https://www.googleapis.com/youtube/v3/commentThreads?part=snippet&videoId={video_id}&key={api_key}&maxResults=100&pageToken={url}"
    return comments

def save_comments_to_file(comments, filename="comments.json"):
    with open(filename, "w", encoding="utf-8") as f:
        json.dump(comments, f, ensure_ascii=False, indent=4)
    print(f"Comments have been saved to {filename}")

def generate_text_file(comments, filename="comments.txt"):
    with open(filename, "w", encoding="utf-8") as f:
        for comment in comments:
            f.write(f"Comment: {comment['text']}\n")
            f.write(f"Likes: {comment['likeCount']}\n")
            f.write(f"Date: {comment['publishedAt']}\n")
            f.write(f"Author: {comment['authorDisplayName']}\n")
            if comment["replies"]:
                f.write(f"Replies ({len(comment['replies'])}):\n")
                for reply in comment["replies"]:
                    f.write(f"- Reply: {reply['text']}\n")
            f.write("\n---\n\n")
    print(f"Comments have been saved to {filename}")

# Example usage
if __name__ == "__main__":
    # Replace these with your actual values
    VIDEO_ID = "YouTube movie ID"                       # ID filmu np. X-i1oBrh2LI
    API_KEY = "YouTube_API_KEY"                         # <-------- miejsce na YouTube API KEY

    # Get all comments
    comments = get_all_comments(VIDEO_ID, API_KEY)

    # Save to file
    save_comments_to_file(comments)

    # Generate text file
    generate_text_file(comments)

    # Print summary
    print(f"Total comments fetched: {len(comments)}")
    total_replies = sum(len(comment["replies"]) for comment in comments)
    print(f"Total replies fetched: {total_replies}")

"""
Example prompt for ChatGPT:

You are a data analyst specializing in social media content and sentiment analysis. I will provide you with YouTube comments data. Please analyze these comments and provide comprehensive insights in the following areas:
1. Sentiment and Emotional Analysis

Analyze the overall sentiment of comments (positive, negative, neutral)
Identify emotional patterns and themes
Point out particularly notable emotional responses
Highlight any interesting patterns in how viewers express their feelings

2. Content and Topics Analysis

Identify main discussion themes and topics
Find recurring subjects or patterns in discussions
Note any interesting debates or disagreements
Analyze references to external topics, events, or other content
Identify inside jokes or community-specific references

3. Engagement Patterns

Analyze patterns in likes and replies
Identify what types of comments generate the most engagement
Look for conversation chains and discussion patterns
Note any viral or highly-referenced comments
Examine user interaction patterns

4. Community Behavior

Identify community-specific language or jargon
Analyze how users interact with each other
Note any community roles or recurring user types
Identify community consensus or disagreements
Examine response patterns to different topics

5. Technical Language Analysis

Identify commonly used terms and phrases
Analyze language complexity and style
Note any multilingual patterns or language mixing
Identify unique expressions or terminology

6. Temporal Analysis

Note any changes in comment patterns over time
Identify peak engagement periods
Analyze how discussions evolved
Note any event-related commenting patterns

Additional Instructions:

Support your analysis with specific examples from the comments
Provide numerical data when relevant
Highlight unexpected or particularly interesting findings
Note any limitations in your analysis
Suggest potential interpretations of patterns you identify
If relevant, compare patterns to typical social media behavior

Please structure your response with clear headings and bullet points for readability. Prioritize insights that would be most valuable for understanding audience engagement and community dynamics.
For each observation, try to provide:

The specific finding
Supporting evidence/examples
Possible interpretation
Relevance or implications

Note: If you notice any other significant patterns or insights not covered by these categories, please include them in your analysis.
End your analysis with a summary of the most significant findings and any recommendations for future analysis or engagement strategies.

Answer in Polish language:

"""