fix: Parse Qwerty restaurant from their website

api/restaurants/qwerty.py

@@ -1,98 +1,68 @@
-from datetime import date, timedelta
+# -*- coding: utf-8 -*-
 import logging
-import os
-
-from playwright.sync_api import sync_playwright, expect
+from datetime import date, timedelta
 from unidecode import unidecode
 
+from .utils import fetch_html
+
 NAME = "Qwerty"
-URL = "https://www.facebook.com/QwertyRestaurant"
+URL = "https://qwerty-restaurant--catering3.webnode.cz/menu/"
 GPS = (49.235655201154906, 16.573111921257983)
 
-PLAYWRIGHT_PROXY = os.getenv("PLAYWRIGHT_PROXY", "")
-
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.INFO)
 
 
-def parse_menu():
-    result = {}
-    last_monday = date.today() - timedelta(days=date.today().weekday())
-    with sync_playwright() as playwright:
-        kwargs = {
-            "viewport": {"width": 1080, "height": 1920},
-            "locale": "cs-CZ"
-        }
-        if PLAYWRIGHT_PROXY:
-            kwargs["proxy"] = {"server": PLAYWRIGHT_PROXY}
-        browser = playwright.firefox.launch(headless=True)
-        context = browser.new_context(**kwargs)
-        page = context.new_page()
-        page.goto(URL)
-
-        cookie_button = page.get_by_role("button", name="Odmítnout")
-        expect(cookie_button).to_be_visible()
-        cookie_button.click()
-
-        login_close_button = page.get_by_role("button", name="Zavřít")
-        expect(login_close_button).to_be_visible()
-        login_close_button.click()
-
-        # Wait for initial posts to load
-        page.wait_for_load_state("networkidle")
-
-        posts = page.locator("[data-ad-preview='message']").all()
-        logger.debug("Checking last %d posts.", len(posts))
-        for idx, post in enumerate(posts):
-            show_more_button = post.get_by_role("button", name="Zobrazit víc")
-            show_more_clicked = False
-            if show_more_button.is_visible():
-                show_more_clicked = True
-                show_more_button.click()
-            lines = post.locator("[style='text-align: start;']").all()
-            logger.debug("Post %d, show_more_clicked=%s, lines=%d", idx+1, show_more_clicked, len(lines))
-
-            current_date = None
-            weekly_section = False
-            weekly_offer = []
-            # Strip first line with dates
-            for line in lines:
-                txt = line.text_content()
-                day_identifier = unidecode(txt.lower())
-                if "pondeli" == day_identifier:
-                    current_date = last_monday
-                elif "utery" == day_identifier:
-                    current_date = last_monday + timedelta(days=1)
-                elif "streda" == day_identifier:
-                    current_date = last_monday + timedelta(days=2)
-                elif "ctvrtek" == day_identifier:
-                    current_date = last_monday + timedelta(days=3)
-                elif "patek" == day_identifier:
-                    current_date = last_monday + timedelta(days=4)
-                elif ("tydenni" in day_identifier or "tydne" in day_identifier) and not "polevka" in day_identifier:
-                    current_date = None
-                    weekly_section = True
-                elif current_date and txt:
-                    result.setdefault(current_date, []).append(txt)
-                elif weekly_section and txt:
-                    weekly_offer.append(txt)
-
-            if weekly_offer:
-                for _, offer in result.items():
-                    offer.append("")
-                    offer.append("All week:")
-                    offer.extend(weekly_offer)
+def format_line(line):
+    return (
+        line.text.strip().replace("\n", " ").replace("1)", "\n1)").replace("Kč", "Kč\n")
+    )
 
-            if result:
-                logger.debug("Seems like post %d contains menu, skipping the rest.", idx+1)
-                break
 
-        #page.pause()
-        #page.screenshot(path="example.png")
-        browser.close()
-
-        if not result:
-            logger.debug("No menu found in posts.")
+def parse_menu():
+    current_date = None
+    last_monday = date.today() - timedelta(days=date.today().weekday())
+    result = {}
+    weekly_section = False
+    weekly_offer = []
+    html = fetch_html(URL)
+    if html:
+        for line in html.find_all("div", {"class": "mt-i cf"}):
+            txt = format_line(line)
+            day_identifier = unidecode(txt.lower())
+
+            if "pondeli" in day_identifier:
+                current_date = last_monday
+                txt = txt.replace("PONDĚLÍ", "")
+            elif "utery" in day_identifier:
+                current_date = last_monday + timedelta(days=1)
+                txt = txt.replace("ÚTERÝ", "")
+            elif "streda" in day_identifier:
+                current_date = last_monday + timedelta(days=2)
+                txt = txt.replace("STŘEDA", "")
+            elif "ctvrtek" in day_identifier:
+                current_date = last_monday + timedelta(days=3)
+                txt = txt.replace("ČTVRTEK", "")
+            elif "patek" in day_identifier:
+                current_date = last_monday + timedelta(days=4)
+                txt = txt.replace("PÁTEK", "")
+            elif "tydenni" in day_identifier and not "polevka" in day_identifier:
+                current_date = None
+                weekly_section = True
+                txt = txt.replace("TÝDENNÍ MENU", "")
+
+            if current_date and txt:
+                result.setdefault(current_date, []).append(txt)
+            elif weekly_section and txt:
+                weekly_offer.append(txt)
+
+        if weekly_offer:
+            for _, offer in result.items():
+                offer.append("All week:")
+                offer.extend(weekly_offer)
+
+    if not result:
+        logger.error("No menu found.")
 
     return result