Yelp Scraper¶

from sentimatrix import Sentimatrix

async with Sentimatrix() as sm:
    reviews = await sm.scrape_reviews(
        "https://www.yelp.com/biz/restaurant-name-city",
        platform="yelp",
        max_reviews=100
    )

    for review in reviews:
        print(f"Rating: {'*' * review.rating}")
        print(f"Review: {review.text[:200]}...")

reviews = await sm.scrape_reviews(
    url="https://yelp.com/biz/...",
    platform="yelp",
    max_reviews=200,           # Max reviews
    sort_by="date",            # "date", "rating_asc", "rating_desc", "elites"
    filter_rating=None,        # Filter by stars (1-5)
)

# Business page
"https://www.yelp.com/biz/business-name-city"

# With query params
"https://yelp.com/biz/business-name?sort_by=date_desc"

@dataclass
class Review:
    text: str                    # Review content
    author: str                  # Reviewer name
    rating: int                  # 1-5 stars
    posted_date: datetime        # Review date
    helpful_count: int           # Useful votes
    platform: str = "yelp"
    metadata: dict               # Extra data

review.metadata = {
    "business_name": "Restaurant Name",
    "business_id": "restaurant-name-city",
    "location": "San Francisco, CA",
    "is_elite": True,
    "photos_count": 3,
    "funny_count": 2,
    "cool_count": 5
}

async with Sentimatrix(config) as sm:
    # Scrape restaurant reviews
    reviews = await sm.scrape_reviews(
        "https://yelp.com/biz/my-restaurant-sf",
        platform="yelp",
        max_reviews=200
    )

    # Aspect-based analysis
    aspects = ["food", "service", "ambiance", "price"]
    for review in reviews:
        aspect_sentiment = await sm.analyze_aspects(
            review.text,
            aspects=aspects
        )
        print(aspect_sentiment)

    # Generate insights
    insights = await sm.generate_insights(reviews)
    print("PROS:", insights.pros)
    print("CONS:", insights.cons)