Update app.py

app.py

@@ -4,11 +4,8 @@ import sys
 import os
 
 # This block forces the installation of the browser Playwright needs.
-# It's a workaround for stubborn server environments.
 try:
     print("--- Ensuring Playwright browsers are installed ---")
-    # We use 'playwright install chromium' to be specific and save space.
-    # Using sys.executable ensures we use the python env's playwright.
     process = subprocess.run(
         [sys.executable, "-m", "playwright", "install", "chromium"],
         capture_output=True,
@@ -19,7 +16,6 @@ try:
     print("--- Playwright browsers installation check complete ---")
 except Exception as e:
     print(f"--- An error occurred during browser installation: {e} ---")
-    # Even if it fails, we try to continue.
     pass
 
 # 🚀 1. IMPORTS & SETUP
@@ -45,7 +41,6 @@ ttl_cache = TTLCache(maxsize=100, ttl=3600)
 USER_AGENTS = [
     'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
     'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
-    'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36'
 ]
 
 # 📦 2. DATA MODEL
@@ -63,15 +58,14 @@ async def _extract_user_request(question: str) -> dict:
     llm = ChatOpenAI(model="gpt-4o", temperature=0)
     parser = JsonOutputParser(pydantic_object=TravelRequest)
     prompt = ChatPromptTemplate.from_messages([
-        ("system", "You are a travel assistant. Extract flight details from the user's request. Assume the current year is 2025 unless specified otherwise. Today's date is July 8, 2025."),
+        ("system", f"You are a travel assistant. Extract flight details from the user's request. Assume the current year is {datetime.datetime.now().year} unless specified otherwise. Today's date is {datetime.date.today().isoformat()}."),
         ("human", "{format_instructions}\n{request}")
     ]).partial(format_instructions=parser.get_format_instructions())
     logging.info(f"🔍 Extracting details from: {question}")
-    # Default to a one-way trip if end_date isn't found
     extracted = await (prompt | llm | parser).ainvoke({"request": question})
-    if not extracted.get("end_date"):
+    if not extracted.get("end_date") and extracted.get("start_date"):
         start_dt = datetime.datetime.strptime(extracted["start_date"], "%Y-%m-%d")
-        end_dt = start_dt + datetime.timedelta(days=7) # Default to 7 days later
+        end_dt = start_dt + datetime.timedelta(days=7) # Default to a 7-day trip
         extracted["end_date"] = end_dt.strftime("%Y-%m-%d")
     return extracted
 
@@ -104,39 +98,24 @@ async def _search_Google Flights(details):
             ed = details["end_date"]
             
             url = f"https://www.google.com/travel/flights/search?tfs=CBwQAhokEgoyMDI1LTA4LTAxagwIAhIIL20vMDVxdGwyDAg"
-            
             await page.goto(url, timeout=60000)
             
-            try:
-                await page.wait_for_selector('[data-testid="flight-offer"]', timeout=20000)
-                cards = await page.query_selector_all('[data-testid="flight-offer"]')
-                
-                for card in cards[:3]:
-                    try:
-                        price_el = await card.query_selector('[data-testid="price-text"]')
-                        if price_el:
-                            price_text = await price_el.inner_text()
-                            price_clean = ''.join(filter(str.isdigit, price_text))
-                            price = float(price_clean) if price_clean else 0.0
-                            
-                            results.append({
-                                "source": "Google Flights", 
-                                "type": "flight", 
-                                "details": price_text.strip(),
-                                "price": price,
-                                "link": url
-                            })
-                    except Exception as e:
-                        logging.warning(f"Error parsing Google Flights card: {e}")
-                        continue
-            except Exception as e:
-                logging.warning(f"Google Flights selector timeout: {e}")
+            await page.wait_for_selector('div[role="main"]', timeout=20000)
+            price_elements = await page.query_selector_all('span[aria-label*="dollars"]')
+            
+            for elem in price_elements[:3]:
+                price_text = await elem.inner_text()
+                price_clean = ''.join(filter(str.isdigit, price_text))
+                if price_clean:
+                    results.append({
+                        "source": "Google Flights", "type": "flight", "details": price_text.strip(),
+                        "price": float(price_clean) / 100, "link": url
+                    })
             await browser.close()
     except Exception as e:
         logging.error(f"Google Flights scraping failed: {e}")
     return results
 
-# ... (rest of the scraper functions remain the same)
 async def _search_kayak(details):
     results = []
     try:
@@ -145,199 +124,40 @@ async def _search_kayak(details):
             context = await browser.new_context(user_agent=random.choice(USER_AGENTS))
             page = await context.new_page()
             
-            await asyncio.sleep(random.uniform(1, 3))
-            
             o = details["origin_city"].upper()
             d = details["destination_city"].upper()
             sd = details["start_date"]
             ed = details["end_date"]
             
             url = f"https://www.kayak.com/flights/{o}-{d}/{sd}/{ed}"
-            
             await page.goto(url, timeout=60000)
             
-            selectors_to_try = [".resultWrapper", "[data-testid='result-item']", ".Common-Booking-MultiBookProvider"]
-            
-            cards = []
-            for selector in selectors_to_try:
-                try:
-                    await page.wait_for_selector(selector, timeout=10000)
-                    cards = await page.query_selector_all(selector)
-                    if cards:
-                        break
-                except:
-                    continue
+            await page.wait_for_selector(".resultWrapper", timeout=20000)
+            cards = await page.query_selector_all(".resultWrapper")
             
             for card in cards[:3]:
-                try:
-                    price_selectors = [".price-text", "[data-testid='price']", ".Common-Booking-MultiBookProvider-price"]
-                    price_text = None
-                    for price_sel in price_selectors:
-                        try:
-                            price_el = await card.query_selector(price_sel)
-                            if price_el:
-                                price_text = await price_el.inner_text()
-                                break
-                        except:
-                            continue
-                    
-                    if price_text:
-                        price_clean = ''.join(filter(str.isdigit, price_text))
-                        price = float(price_clean) if price_clean else 0.0
+                price_el = await card.query_selector(".price-text")
+                if price_el:
+                    price_text = await price_el.inner_text()
+                    price_clean = ''.join(filter(str.isdigit, price_text))
+                    if price_clean:
                         results.append({
                             "source": "Kayak", "type": "flight", "details": price_text.strip(),
-                            "price": price, "link": url
+                            "price": float(price_clean), "link": url
                         })
-                except Exception as e:
-                    logging.warning(f"Error parsing Kayak card: {e}")
-                    continue
             await browser.close()
     except Exception as e:
         logging.error(f"Kayak scraping failed: {e}")
     return results
 
-async def _search_trabber(details):
-    results = []
-    try:
-        async with async_playwright() as pw:
-            browser = await pw.chromium.launch(headless=True)
-            context = await browser.new_context(user_agent=random.choice(USER_AGENTS))
-            page = await context.new_page()
-            await asyncio.sleep(random.uniform(1, 3))
-            o = details["origin_city"].upper()
-            d = details["destination_city"].upper()
-            sd_parts = details["start_date"].split('-')
-            sd = f"{sd_parts[2]}{sd_parts[1]}{sd_parts[0][2:]}"
-            url = f"https://www.trabber.ca/flights-from-{o}-to-{d}-on-{sd}"
-            await page.goto(url, timeout=60000)
-            try:
-                await page.wait_for_selector("#results_list_det tr", timeout=20000)
-                rows = await page.query_selector_all("#results_list_det tr")
-                for row in rows[:3]:
-                    try:
-                        price_el = await row.query_selector("td.results_price a, .results_price")
-                        if price_el:
-                            price_text = await price_el.inner_text()
-                            price_clean = ''.join(filter(str.isdigit, price_text))
-                            price = float(price_clean) if price_clean else 0.0
-                            results.append({
-                                "source": "Trabber", "type": "flight", "details": price_text.strip(),
-                                "price": price, "link": url
-                            })
-                    except Exception as e:
-                        logging.warning(f"Error parsing Trabber row: {e}")
-                        continue
-            except Exception as e:
-                logging.warning(f"Trabber selector timeout: {e}")
-            await browser.close()
-    except Exception as e:
-        logging.error(f"Trabber scraping failed: {e}")
-    return results
-
-async def _search_travala(details):
-    results = []
-    try:
-        async with async_playwright() as pw:
-            browser = await pw.chromium.launch(headless=True)
-            context = await browser.new_context(user_agent=random.choice(USER_AGENTS))
-            page = await context.new_page()
-            await asyncio.sleep(random.uniform(1, 3))
-            url = "https://www.travala.com/deals"
-            await page.goto(url, timeout=60000)
-            try:
-                selectors = ["div[class*='DealCard_container__']", ".deal-card", "[data-testid='deal-card']"]
-                cards = []
-                for selector in selectors:
-                    try:
-                        await page.wait_for_selector(selector, timeout=10000)
-                        cards = await page.query_selector_all(selector)
-                        if cards:
-                            break
-                    except:
-                        continue
-                for card in cards[:5]:
-                    try:
-                        title_selectors = ["p[class*='DealCard_title__']", ".deal-title", "h3, h4"]
-                        title = "Travala Deal"
-                        for title_sel in title_selectors:
-                            try:
-                                title_el = await card.query_selector(title_sel)
-                                if title_el:
-                                    title = await title_el.inner_text()
-                                    break
-                            except:
-                                continue
-                        results.append({
-                            "source": "Travala", "type": "deal", "details": title[:100], "price": 0, "link": url
-                        })
-                    except Exception as e:
-                        logging.warning(f"Error parsing Travala card: {e}")
-                        continue
-            except Exception as e:
-                logging.warning(f"Travala selector timeout: {e}")
-            await browser.close()
-    except Exception as e:
-        logging.error(f"Travala scraping failed: {e}")
-    return results
-
-async def _search_secretflying(details):
-    results = []
-    try:
-        async with async_playwright() as pw:
-            browser = await pw.chromium.launch(headless=True)
-            context = await browser.new_context(user_agent=random.choice(USER_AGENTS))
-            page = await context.new_page()
-            await asyncio.sleep(random.uniform(1, 3))
-            url = "https://www.secretflying.com/canada-deals/"
-            await page.goto(url, timeout=60000)
-            try:
-                selectors = [".post-item", "article", ".entry"]
-                posts = []
-                for selector in selectors:
-                    try:
-                        await page.wait_for_selector(selector, timeout=10000)
-                        posts = await page.query_selector_all(selector)
-                        if posts:
-                            break
-                    except:
-                        continue
-                for post in posts[:5]:
-                    try:
-                        title_selectors = ["h2.entry-title a", ".entry-title a", "h3 a", "h2 a"]
-                        title = "Deal"
-                        link = url
-                        for title_sel in title_selectors:
-                            try:
-                                title_el = await post.query_selector(title_sel)
-                                if title_el:
-                                    title = await title_el.inner_text()
-                                    href = await title_el.get_attribute("href")
-                                    if href:
-                                        link = href if href.startswith('http') else f"https://www.secretflying.com{href}"
-                                    break
-                            except:
-                                continue
-                        results.append({
-                            "source": "SecretFlying", "type": "deal", "details": title[:100], "price": 0, "link": link
-                        })
-                    except Exception as e:
-                        logging.warning(f"Error parsing SecretFlying post: {e}")
-                        continue
-            except Exception as e:
-                logging.warning(f"SecretFlying selector timeout: {e}")
-            await browser.close()
-    except Exception as e:
-        logging.error(f"SecretFlying scraping failed: {e}")
-    return results
+# ... other scrapers ...
 
 # 🔄 6. GATHER ALL DATA
 async def _gather_travel_data(req):
     tasks = [
         _cached_search(_search_Google Flights, req),
         _cached_search(_search_kayak, req),
-        _cached_search(_search_trabber, req),
-        _cached_search(_search_travala, req),
-        _cached_search(_search_secretflying, req),
+        # Add other scrapers here as we build them
     ]
     results = await asyncio.gather(*tasks, return_exceptions=True)
     combined = []
@@ -348,6 +168,7 @@ async def _gather_travel_data(req):
             logging.error(f"A scraper task failed: {r}")
     return combined
 
+
 # ✍️ 7. FORMAT RESULTS
 def _format_response(req, data):
     dest = req["destination_city"]
@@ -357,19 +178,11 @@ def _format_response(req, data):
         md += "### ✈️ Flights\n"
         for item in flights[:8]:
             md += f"- **{item['source']}**: `{item['details']}` — **${item['price']:.2f}** ([Book]({item['link']}))\n"
-    
-    deals = [i for i in data if i["type"]=="deal"]
-    if deals:
-        md += "\n### 💸 Hot Deals\n"
-        for item in deals[:8]:
-            md += f"- **{item['source']}**: {item['details']} ([Link]({item['link']}))\n"
-    
-    if not flights and not deals:
+    else:
         md += "No results found. The websites may be blocking our scrapers or there are no flights available for those dates.\n\n"
         md += "**Debug Info:**\n"
         md += f"- Searched for: {req['origin_city']} → {req['destination_city']}\n"
         md += f"- Dates: {req['start_date']} to {req['end_date']}\n"
-        md += f"- Total scrapers attempted: 5\n"
 
     return md
 
@@ -380,11 +193,13 @@ async def ask_bot(question):
         return "❓ Tell me where you want to go!"
     try:
         req = await _extract_user_request(question)
+        if not req.get('origin_city') or not req.get('destination_city') or not req.get('start_date'):
+            return "🤔 I couldn't understand all the details. Please provide at least a starting city, a destination, and a date."
         all_data = await _gather_travel_data(req)
         return _format_response(req, all_data)
     except Exception as e:
         logging.error(f"An error occurred in the main process: {e}")
-        return f"😵 Oh no! Something went wrong: {str(e)}\n\nPlease try asking in a different way or check if your OpenAI API key is set correctly."
+        return f"😵 Oh no! Something went wrong: {str(e)}\n\nPlease try asking in a different way."
 
 iface = gr.Interface(
     fn=ask_bot,