Hi,

today the date at the Personalkantine website is wrong: "026.02.2018"
A copy of todays website can be found in the attachments.
I had quick look at the code and have no idea how to fix this. But maybe someone else has one.
Additionally this maybe leads to an overlap when the menu for the 26.02.2018 is reed.
But I would assume that this will just be overwritten.

Website:
website.zip

Implement some way to inform users about new features, canteens or other updates.

Would be great to have support for the K10 Bistro @ Uni Kassel :-)

https://www.studentenwerk-kassel.de/meta/speiseplan/bistrok10/

EN canteen is a canteen on TU campus. I think it should be possible to use the parser of Personalkantine.

@numbleroot mentioned that the bot claims that Chilli Paneer contains murdered animals. This is indeed wrong.

Currently the bot has exactly zero tests. As the codebase and the userbase grow it would be smart to write tests. 💡

After 18:00 it does not make sense to answer with the menu of the current day. We should send the menu for tomorrow.
This depends on #15.

I don't know whether you are already parsing some PDF menus anywhere, but if so, it would be nice if you could add the Cafe Nero located at TU Universitätsbibliothek as well.

The link to the current day's menu is embedded into an iframe view of the PDF uploaded to Dropbox... Maybe you can work some magic here and extract the meals anyways.
Link to menu: http://cafenero.net/speisekarte/

Thanks for the awesome bot! :)

When a user asks for an unknown canteen I want to check if the request matches some existing canteens and offer them to the user.

/tu_m

Did you mean:
/tu_mensa
/tu_marchstr

This projects needs a readme file to explain what it does and how to set it up.

Another canteen @ Uni Kassel (the main one)

https://www.studentenwerk-kassel.de/meta/speiseplan/zentralmensa-arnold-bode-strasse/

They changed the PDF menu and removed the daily schonkost meal. This breaks all annotations in the parser.

Cafe Nero is broken for some months now.

If there are problems during the update of menus the cache should not get changed.

entrypoint.sh calls warmup_cache.py, but this file is missing in the repository.

Mo-Do 07-17
Fr 07-16

It would be nice to allow to set the ACCESS_TOKEN as environment variable. This is useful to deploy OmNomNom in a docker container.

if env
  use env
else if
  use config.py
else
  throw error
fi

We depend on python-telegram-bot. They released a release candidate for version 4.0 which is not backwards compatible. So we have to change a few things before we use the new version.

Currently the bot is missing mostly everything you would expect from a well documented project.

If you are thinking about contributing code or just want to run the bot yourself, please let me know, so I can prioritize this issue ;-)

It would be nice if the response of the bot would contain the original url that it tried to parse so the user can check it manually in case of any error happening during processing or if the user just wants to see the original thing.

Of course the URL should not be displayed fully. A link on the title of the response would suffice for successful responses and a more detailed description in case of an error.

It would be nice to see if the meals include allergenes like milk or gluten.

Some meals have no pricetag at the website and are shown with a colon in OmNomNom: 🌽 American Dressing:
It should be more clear that this is not a bug in OmNomNom.

Example:
https://www.stw.berlin/mensen/mensa-hfm-charlottenstra%C3%9Fe.html

The Berlin Studentenwerk has many canteens in Berlin. It would be nice to add all of them. It should be really easy.

#!/usr/bin/env python3

import bs4
import requests

HEADERS = {'user-agent': 'User-Agent: Mozilla'}
PARAMS = {'resources_id': 534}
URL = 'https://www.stw.berlin/xhr/speiseplan-und-standortdaten.html'


request = requests.post(URL, headers=HEADERS, data=PARAMS)

if request.status_code == requests.codes.ok:
    soup = bs4.BeautifulSoup(request.text, 'html.parser')
    listbox = soup.find('select', class_='listboxStandorte')
    options = listbox.findAll('option')

    for option in options:
        canteen = '%s : %s' % (option.attrs['value'], option.get_text())
        print(canteen)

534 : Mensa ASH Berlin Hellersdorf
535 : Mensa Beuth Hochschule für Technik Kurfürstenstraße
527 : Mensa Beuth Hochschule für Technik Luxemburger Straße
537 : Mensa Charité Zahnklinik
529 : Mensa EHB Teltower Damm
271 : Mensa FU Herrenhaus Düppel
322 : Mensa FU II Otto-von-Simson-Straße
528 : Mensa FU Lankwitz Malteserstraße
531 : Mensa HfM Charlottenstraße
533 : Mensa HfS Schnellerstraße
320 : Mensa HTW Treskowallee
319 : Mensa HTW Wilhelminenhof
147 : Mensa HU Nord
191 : Mensa HU Oase Adlershof
367 : Mensa HU Süd
270 : Mensa HU Spandauer Straße
526 : Mensa HWR Badensche Straße
532 : Mensa Katholische HS für Sozialwesen
530 : Mensa KHS Weißensee
321 : Mensa TU Hardenbergstraße
323 : Mensa Veggie N° 1 - Die grüne Mensa
368 : Cafeteria FU Ihnestraße
660 : Cafeteria FU Koserstraße
542 : Cafeteria FU Pharmazie
277 : Cafeteria FU Rechtswissenschaft
543 : Cafeteria FU Wirtschaftswissenschaften
723 : Cafeteria HfM „Neuer Marstall“
726 : Cafeteria HTW Treskowallee
659 : Cafeteria HU „Jacob und Wilhelm Grimm Zentrum“
539 : Cafeteria TU Ackerstraße
540 : Cafeteria TU Architektur
657 : Cafeteria TU „Skyline“
631 : Cafeteria TU Hardenbergstraße
541 : Cafeteria TU Hauptgebäude „Wetterleuchten“
538 : Cafeteria TU Marchstraße
722 : Cafeteria UdK „Jazz-Cafe“
658 : Cafeteria UdK Lietzenburger Straße
647 : Coffeebar Beuth Hochschule
648 : Coffeebar Beuth Hochschule Haus Grashof
1407 : Coffeebar EHB Teltower Damm
724 : Coffeebar HfM Charlottenstraße
725 : Coffeebar HTW Wilhelminenhof
661 : Coffeebar HU „c.t“
721 : Coffeebar HU Mensa Nord
720 : Coffeebar HU Oase Adlershof
727 : Coffeebar HWR Alt-Friedrichsfelde
728 : Coffeebar HWR Badensche Straße
649 : Coffeebar Mensa FU II
650 : Coffeebar Mensa Lankwitz
632 : Coffeebar TU Hardenbergstraße

Currently every canteen is a task. Since #15 every canteen parses two weeks of menus. To avoid rate limits there are a lot of sleep($time) statements in the code which leads to long blocking tasks in the queue. (Some tasks have to wait more than twenty minutes...)

Solution: Create a single celery task for every canteen/date combination and apply a global rate limit for this type of task. The new tasks are smaller and need less time. Other tasks without (or with another limit) should be able to run between them.

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Ignored or Blocked

These are blocked by an existing closed PR and will not be recreated unless you click a checkbox below.

• chore(deps): update dependency redis to v5

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

A feature to find other people to enjoy your lunch break together.

Currently the bot only knows the menus of the current day. It would be nice to add support for something like this.

/mar tomorrow
/tel tuesday
/a 10.02.2017

This needs changes in every parser.

Currently the bot is logging some messages into InfluxDB. This should be streamlined to be more useful.
Maybe I can implement it as a custom python logger.

Currently a push to master only triggers a build at hub.docker.com but it needs manual interaction to get deployed. This could and should be automated.

the canteen has it's menu at https://personalkantine.personalabteilung.tu-berlin.de/#wochenkarte

Currently every menu is stored as a single string in redis. Instead they should be stored as a JSON object. This adds some overhead when the bot responds to messages but it enables the possibility to implement other frontends like an Alexa skill or a general API.

F.ex. /march prints a tip message at the bottom:

Tipp: Wenn Sie dieses Fenster schließen, wird es nur erneut durch Leeren des Cache wieder angezeigt.

which has nothing to do with the list of meals. This text should be removed from the bot response.

Users should be able to report invalid answers to bring a broken parser to my attention.

Sing Catering is a canteen on TU campus. We need an additional HTML parser for this.

The Studierendenwerk added a new canteen: https://www.stw.berlin/news/mensavegan.html

On saturday and sunday the parser tries to find a menu for the current week. But most of the time the menu of Personalkantine is updated for the next week on friday night. As a result the parser overwrites the entire menu of the current week with an error.

It would be nice if the bot could monitor the studierendenwerk website to report changes in the available canteens.

Because the bot runs on a server, the GPL copyleft probably doesn't work. If you intended it to work, you should change your license to AGPL.

Can we translate the menu in English any other language as well. I will be happy to work with you on this feature if you needed. Thanks