canteen = LazyBuilder()

# prices are stored on a separate page

document = parse(urlopen(base + '/mensa-preise/').read(), 'lxml')

prices = {}

for tr in document.find('div', 'ce-bodytext').find_all('tr'):

meal = tr.find('th')

if not meal or not meal.text.strip():

continue

if len(tr.find_all('td', 'betrag')) < 3:

continue

if 'titel' in meal.attrs.get('class', []) or 'zeilentitel' in meal.attrs.get('class', []):

continue

meal = meal.text.strip()

prices[meal] = {}

for role, _id in [('student', 0), ('employee', 1), ('other', 2)]:

price_html = tr.find_all('td', 'betrag')[_id].text

price_search = price_regex.search(price_html)

if price_search:

prices[meal][role] = price_search.group('price')

errorCount = 0

date = datetime.date.today()

while errorCount < 7:

try:

document = parse(urlopen(url.format(date)).read(), 'lxml')

errorCount = 0

except HTTPError as e:

if e.code == 404:

errorCount += 1

date += datetime.timedelta(days=1)

continue

else:

raise e

# extract legend

legend = {}

legends = document.find('div', 'tx-stwm-speiseplan')

additions = legends.find('div', 'c-schedule__filter-body')

for table in additions.find_all('div', 'c-schedule__filter-item'):

for ingredient in table.find('ul').find_all('li'):

name = ingredient.find('dt').text.strip()

description = ingredient.find('dd').text.strip()

legend[name] = description

for label in legends.find('ul', 'c-schedule__type-list').find_all('li'):

name = label.find('dt').text.replace('(', '').replace(')', '').strip()

description = label.find('dd').text.strip()

legend[name] = description

# extract meals

mensa_data = document.find('ul', 'c-schedule__list')

category = None

for meal in mensa_data.find_all('li'):

# update category or use previous one if not specified

category_text = meal.find('dt', 'c-schedule__term').text.strip()

if category_text:

category = category_text

data = meal.find('dd').find('p', 'js-schedule-dish-description')

name = data.contents[0].strip() # name is the first text node

if not name:

continue

# notes are contained in 3 boxes (type, additional, allergen) and

# are comma-separated lists enclosed in brackets or parentheses

notes = []

for note in meal.find_all('span', 'c-schedule__marker'):

note_text = note.find('span', 'u-text-sup').text \

.replace('(', '').replace(')', '') \

.replace('[', '').replace(']', '')

notes += [n for n in note_text.split(',') if n]

# some meals contain the GQB label in their name (instead of in notes)

if '(GQB)' in name:

name = name.replace('(GQB)', '').strip()

notes.append('GQB')

# the price for both meals is specified as Bio-/Aktionsgericht

price_category = category \

.replace('Aktionsessen', 'Bio-/Aktionsgericht') \

.replace('Biogericht', 'Bio-/Aktionsgericht') \

.strip()

canteen.addMeal(date, category, name,

[legend.get(n, n) for n in notes],

prices.get(price_category, {})

)

date += datetime.timedelta(days=1)

if today:

break