initial

.gitignore

@@ -0,0 +1 @@
+*.db

fetch.py

@@ -0,0 +1,71 @@
+"""
+Try to input dishes from the coming week.
+"""
+
+import json
+import sqlite3
+import subprocess
+from datetime import date, timedelta
+
+DB = "mensa.db"
+MENSEN = ["cgin", "mensa_am_turm", "zentralmensa"]
+
+db = sqlite3.connect(DB)
+
+db.execute("""
+CREATE TABLE IF NOT EXISTS dishes (
+    date TEXT NOT NULL,
+    mensa TEXT NOT NULL,
+    type TEXT NOT NULL,
+    title TEXT,
+    description TEXT,
+    category TEXT,
+    allergens TEXT,
+    UNIQUE(date, mensa, type)
+)
+""")
+
+TODAY = date.today()
+
+for i in range(7):
+    day = TODAY + timedelta(days=i)
+
+    for mensa in MENSEN:
+        result = subprocess.run(
+            ["python", "mensa.py", day.isoformat(), mensa],
+            capture_output=True,
+            text=True,
+        )
+
+        if not result.stdout:
+            print(f"no results: {day} {mensa}")
+            continue
+
+        dishes = json.loads(result.stdout)
+
+        for dish in dishes:
+            try:
+                db.execute(
+                    """
+                    INSERT INTO dishes
+                    (date, mensa, type, title, description, category, allergens)
+                    VALUES (?, ?, ?, ?, ?, ?, ?)
+                    """,
+                    (
+                        dish["date"],
+                        dish["mensa"],
+                        dish["type"],
+                        dish["title"],
+                        dish["description"],
+                        json.dumps(dish["category"], ensure_ascii=False),
+                        json.dumps(dish["allergens"], ensure_ascii=False),
+                    ),
+                )
+                print(f"registered: {dish['date']} {dish['mensa']} {dish['type']}")
+            except sqlite3.IntegrityError:
+                print(
+                    f"already registered: {dish['date']} {dish['mensa']} {dish['type']}"
+                )
+
+db.commit()
+db.close()

justfile

@@ -0,0 +1,2 @@
+fetch:
+    python fetch.py

mensa.py

@@ -0,0 +1,69 @@
+"""
+When there is an error there is no output but an error.
+All in german.
+"""
+
+import json
+import re
+import sys
+
+import requests
+from bs4 import BeautifulSoup
+
+DATE, MENSA = sys.argv[1:3]
+URL = f"https://www.studierendenwerk-goettingen.de/fileadmin/templates/php/mensaspeiseplan/cached/de/{DATE}/{MENSA}.html"
+
+r = requests.get(URL)
+if r.status_code == 404:
+    sys.exit(1)
+
+soup = BeautifulSoup(r.text, "html.parser")
+table = soup.select_one("table.sp_tab")
+
+if not table or "kein Angebot" in table.get_text():
+    sys.exit(1)
+
+dishes = []
+
+for row in table.select("tr")[1:]:
+    cells = row.select("td")
+    if len(cells) != 3:
+        continue
+
+    typ, bez, hin = cells
+    strong = bez.select_one("strong")
+    raw_title = strong.get_text(" ", strip=True) if strong else ""
+
+    title = re.sub(r"\s*\([^)]*\)", "", raw_title)
+
+    description = bez.get_text(" ", strip=True).removeprefix(raw_title).strip()
+    angebot = bez.select_one("i.smaller")
+    if angebot:
+        description = description.removesuffix(
+            angebot.get_text(" ", strip=True)
+        ).strip()
+
+    allergens = [
+        x.strip()
+        for group in re.findall(r"\(([^()]*)\)", raw_title + " " + description)
+        for x in group.split(",")
+    ]
+
+    category = [
+        str(img["src"]).rsplit("/", 1)[-1].rsplit(".", 1)[0]
+        for img in hin.select("img")
+    ]
+
+    dishes.append(
+        {
+            "date": DATE,
+            "mensa": MENSA,
+            "type": typ.get_text(strip=True),
+            "title": title,
+            "description": description,
+            "category": category,
+            "allergens": allergens,
+        }
+    )
+
+print(json.dumps(dishes, ensure_ascii=False))

pretty.py

@@ -0,0 +1,21 @@
+import json
+import sqlite3
+
+db = sqlite3.connect("mensa.db")
+
+rows = db.execute("""
+    SELECT date, mensa, type, title, description, category, allergens
+    FROM dishes
+    ORDER BY rowid DESC
+    LIMIT 3
+""").fetchall()
+
+for date, mensa, typ, title, description, category, allergens in rows:
+    print(f"{date} | {mensa} | {typ}")
+    print(f"  {title}")
+    print(f"  {description}")
+    print(f"  category: {', '.join(json.loads(category))}")
+    print(f"  allergens: {', '.join(json.loads(allergens))}")
+    print()
+
+db.close()