veripath/implementation/auditing/bing_provider.py

#!/usr/bin/env python3
"""
Bing Places scraper via Bing Web Search (NOT Bing Maps).

Bing Maps blocks headless (WebGL wall), but Bing *web search* renders
full local business data in the DOM: name, address, phone, hours,
rating, reviews, categories, description.

Data is Yelp-sourced (Bing attributes to Yelp). Same as Apple Maps.
"""
import asyncio
import json
import re
import sys
from playwright.async_api import async_playwright


async def scrape_bing(query: str) -> dict | None:
    """Scrape Bing Places data from Bing web search results.

    Args:
        query: "Business Name City, ST"

    Returns:
        dict with name, address, phone, website, rating, reviews,
        hours, category, description, reviews_sample, or None on failure.
    """
    q = re.sub(r'\s+', ' ', query).strip()
    url = f"https://www.bing.com/search?q={q.replace(' ', '+')}"

    async with async_playwright() as pw:
        browser = await pw.firefox.launch(headless=True)
        context = await browser.new_context(
            viewport={"width": 1920, "height": 1080},
            user_agent="Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.6 Safari/605.1.15",
            locale="en-US",
            timezone_id="America/Los_Angeles",
        )
        await context.add_init_script("Object.defineProperty(navigator,'webdriver',{get:()=>false});")
        page = await context.new_page()
        try:
            await page.goto(url, wait_until="domcontentloaded", timeout=30000)
            await page.wait_for_timeout(8000)
            body_text = await page.locator("body").text_content()
        except Exception:
            await browser.close()
            return None
        await browser.close()

    # Find business section by first word of query
    first_word = q.split()[0]
    biz_idx = body_text.find(first_word)
    if biz_idx < 0:
        return None

    # Grab 3000 chars from business mention — this is one long blob
    section = body_text[biz_idx : biz_idx + 3000]
    result = {}

    # NAME: from first word up to the first URL
    name_m = re.match(r"([A-Za-z\s&+,.\'-]{3,60}?)\s*https?://", section)
    if not name_m:
        name_m = re.match(r"([A-Za-z\s&+,.\'-]{3,60}?)\s*Share", section)
    if not name_m:
        name_m = re.match(r"([A-Za-z\s&+,.\'-]{3,60}?)\s*\d", section)  # before rating
    if name_m:
        result["name"] = name_m.group(1).strip()
    else:
        return None

    # WEBSITE
    url_m = re.search(r"(https?://[a-zA-Z0-9._/-]+?)(?:\s|$)", section)
    if url_m:
        result["website"] = url_m.group(1).rstrip("/")

    # RATING + REVIEWS: "5/5 (65 reviews)"
    rating_m = re.search(r"(\d+(?:\.\d+)?)\s*/\s*5.*?\(\s*(\d+)\s*reviews?\)", section)
    if rating_m:
        result["rating"] = float(rating_m.group(1))
        result["reviews"] = int(rating_m.group(2))

    # CATEGORY: "reviews) · Beauty & spa in Cameron Park"
    cat_m = re.search(r"reviews?\)\s*·\s*([A-Za-z\s&]+?)\s*in\s+", section)
    if cat_m:
        result["category"] = cat_m.group(1).strip()

    # ADDRESS: "3460 Robin Ln Ste 4, Cameron Park, CA 95682"
    addr_m = re.search(r"(\d+\s+[A-Za-z\s.,#SteSuite]+?\d{5})", section)
    if addr_m:
        result["address"] = addr_m.group(1).strip().rstrip(",.")

    # PHONE: "(530) 350-9197"
    phone_m = re.search(r"(\(\d{3}\)\s*\d{3}-\d{4})", section)
    if not phone_m:
        phone_m = re.search(r"(\d{3}-\d{3}-\d{4})", section)
    if phone_m:
        result["phone"] = phone_m.group(1)

    # HOURS: "Thursday9 AM - 7 PMFriday9 AM - 6 PM..."
    # Pattern: DayName + hours (with no space between day and time)
    hours = {}
    days_order = ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday", "Sunday"]
    for day in days_order:
        day_m = re.search(
            rf"{day}(\d+(?::\d+)?\s*[AP]M?\s*-\s*\d+(?::\d+)?\s*[AP]M?|Closed|opens\s+\w+\s+\d+[AP]M?)",
            section,
            re.I,
        )
        if day_m:
            hours[day.lower()] = day_m.group(1).strip()
    if hours:
        result["hours"] = hours

    # DESCRIPTION: "AboutPhoenix Salon + Spa offers..."
    about_idx = section.find("About")
    if about_idx >= 0:
        about_snip = section[about_idx:about_idx + 600]
        # Skip "About" then business name
        name = result.get("name", "")
        name_idx = about_snip.find(name)
        if name_idx >= 0:
            desc_text = about_snip[name_idx + len(name):].strip()
            # Grab first sentence
            end = desc_text.find(".")
            if end > 0:
                result["description"] = desc_text[:end + 1].strip()

    # REVIEWS: "Apr 21, 2026I have been a customer...Learn more ·"
    reviews = []
    for rm in re.finditer(
        r"((?:Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)\s+\d{1,2},\s+\d{4})(.*?)Learn more",
        section,
        re.DOTALL,
    ):
        text = rm.group(2).strip().replace("✕", "")
        text = re.sub(r'\s+', ' ', text)[:200]
        if len(text) < 20:
            continue
        reviews.append({"text": text, "date": rm.group(1)})
        if len(reviews) >= 3:
            break
    if reviews:
        result["reviews_sample"] = reviews

    return result if result.get("name") else None


if __name__ == "__main__":
    query = sys.argv[1] if len(sys.argv) > 1 else "Phoenix Salon + Spa Cameron Park, CA"
    result = asyncio.run(scrape_bing(query))
    if result:
        print(json.dumps(result, indent=2, ensure_ascii=False))
    else:
        print("No data extracted")