Added Parser for most of the Mensas from Studentenwerk Thueringen #79

Anarchoschnitzel · 2018-04-26T06:55:36Z

Added Parser for

Mensa Ehrenberg
Cafeteria Ehrenberg
Nanoteria
Roentgenmensa
from TU Ilmenau.
http://www.stw-thueringen.de/deutsch/mensen/einrichtungen/ilmenau/mensa-ehrenberg.html
http://www.stw-thueringen.de/deutsch/mensen/einrichtungen/ilmenau/cafeteria-mensa-ehrenberg.html
http://www.stw-thueringen.de/deutsch/mensen/einrichtungen/ilmenau/cafeteria-nanoteria.html
http://www.stw-thueringen.de/deutsch/mensen/einrichtungen/ilmenau/cafeteria-roentgenbau.html
My current implementation marks days in the past as closed. Please let me know if that is not the desired behavior and I will change it :)

…sa Roentgen)

j-maas · 2018-04-28T12:44:49Z

Hey, thanks for your PR!

My current implementation marks days in the past as closed.

I'm not fully sure what you mean, and I understand your code enough to verify if you did it correctly. OpenMensa keeps a history of all parsed days. I assume you mean when today's menu (say Wednesday, 2018-05-02) contains entries for the past week or the day before (2018-04-23, or 2018-05-01), then the correct behavior is to just output them as well. So the feed will simply contain all the days from the menu, and OpenMensa will update the "old" days accordingly.

I will just quickly go through your code and point out a few general things that you could improve. ;)

Anarchoschnitzel · 2018-04-28T12:48:44Z

Hi Y0hy0h, what I mean by past days as closed is the following:
The websites I pull the information from only shows the menus from the current day on and not from days in the past. These days are currently marked as closed.
Example:
The parser is called on a Wednesday and returns alle meals from Wednesday, Thursday and Friday.
It then determines Saturday and Sunday as closed.
As no meals were parsed for Monday and Tuesday (which are in the past from the current day) these days are marked as closed.
Do you think it would be better to return no info about these day in the past at all?

j-maas

There is some very simple things you can do to break down your code, make it more readable. Also you make it more robust and debuggable by improving your exception handling.

j-maas · 2018-04-28T12:49:16Z

parsers/thueringen.py

+					closed_date = mensa_start_date + timedelta(days=i-2)
+					canteen.setDayClosed(closed_date.date())
+		except Exception:
+			continue


This is dangerous, because you're swallowing all exceptions. Prefer making the try block as short as possible and only excepting exactly the errors you are interested in. Or even better, replace it with a clear check (using e. g. if). That way it becomes clear what you are checking, and you do not accidentally swallow other exceptions.

I reworked that part to get rid of the long and nested try block.
Thanks for your recommendations 👍

j-maas · 2018-04-28T12:51:15Z

parsers/thueringen.py

+from datetime import datetime, timedelta
+from pyopenmensa.feed import LazyBuilder
+
+calendar_week_regex = re.compile('\d{4}-\d{1,2}$')


Prefer declaring variables close to where they are used. Readability trumps performance, since you can always optimize readable code, when you've determined that the performance is important.

I moved the variables to the functions they are used at except one which is used at multiple places

j-maas · 2018-04-28T12:51:57Z

parsers/thueringen.py

+	for fee_candidate in fees:
+		fee_string = fees_regex.findall(fee_candidate.text)
+		for s in fee_string:
+			amount_strings = amount_regex.findall(s)


This is deeply nested. Prefer extracting a function and name it after what happens here.

I extracted and split the parsing into 3 different functions

j-maas · 2018-04-28T12:53:42Z

parsers/thueringen.py

+				current_date = mensa_start_date + timedelta(days=day_id-2)
+				meals_data = day.find_all('tr')
+				for meal_data in meals_data:
+					meals = meal_data.find_all('td')


Same here: Extract a function. Small functions are awesome, because you can quickly understand them, whereas such a big function is difficult to read.

Anywhere you can reasonably extract a function, do it. :)

I extracted and split the parsing into 3 different functions

Changed default behaviour of non-available days in the past so that they are ignored instead of treated as closed. Removed try-catch clause

…ed (could happen at some canteens). Added error handling if the start date of a canteens current week could not be parsed

Anarchoschnitzel · 2018-04-28T14:36:00Z

I took care of the changes you requested.
Thanks for your feedback 👍

j-maas · 2018-04-28T16:29:02Z

Do you think it would be better to return no info about these day in the past at all?

Yes, that's what I would do. ;)

Anarchoschnitzel · 2018-04-30T10:26:25Z

Yes, that's what I would do.

This is already changed in the commits that also addressed the changes you requested.

j-maas · 2018-04-30T13:23:44Z

Sorry, I forgot to say that I'd like to wait for @klemens's (or another maintainer's) approval. I'm quite new to this repo so I have no experience merging new parsers.

@klemens hinted that he might take some time to do that in a while. I hope it's ok for you to wait until that happens.

klemens

Thank you for writing this parser!

I just had a little bit of time and so this is only a first round of review. I have not yet looked at the parse_fees and parse_meals functions.

There are some problems that need to be fixed and I also left a couple of improvement suggestions. Also python doesn't mind if you put an empty line here and there and it will improve readability quite a lot. 😉

klemens · 2018-05-04T17:53:52Z

build_scripts/test-urls.txt

+http://localhost:9090/thueringen/je-zeiss/full.xml
+http://localhost:9090/thueringen/nh-mensa/full.xml
+http://localhost:9090/thueringen/sk-mensa/full.xml
+http://localhost:9090/thueringen/we-horn/full.xml


Only testing one should be enough.

klemens · 2018-05-04T18:03:32Z

parsers/thueringen.py

+	fees = parse_fees(document)
+	if len(fees) is 2:
+		employees_fee = fees[0]
+		guests_fee = fees[1]


You could just define the defaults in the parse_fees function or as "constants" in the global scope. Also len(fees) is always 2 and you should check using == instead of is (see below).

klemens · 2018-05-04T18:08:26Z

parsers/thueringen.py

+def parse_start_date(document):
+	week_start_end_date_regex = re.compile('\d{1,2}\.\d{1,2}\.\d{4}')
+	calendar_week_regex = re.compile('\d{4}-\d{1,2}$')
+	options = document.find_all('option', value=calendar_week_regex)


The select box has name="selWeek", so you don't need to search the whole document.

klemens · 2018-05-04T18:17:02Z

parsers/thueringen.py

+	calendar_week_regex = re.compile('\d{4}-\d{1,2}$')
+	options = document.find_all('option', value=calendar_week_regex)
+	assert len(options) > 0
+	options_filtered = list(filter(lambda x: x.has_attr('selected'), options))


BS can search for multiple attributes at once:

doc.find("option", value=calendar_week_regex, selected=True)

You could also remove the regex if you just generate the correct value for the current week directly.

Good to know, I am fairly new to BS and had my problems with that one ;-)

klemens · 2018-05-04T18:32:52Z

parsers/thueringen.py

+	try:
+		mensa_start_date = parse_start_date(document)
+	except AssertionError:
+		return canteen.toXMLFeed()


Does this happen on weekends? I would prefer just returning None when the error is expected (like on a weekend) and otherwise just let the exceptions terminate the parser, because otherwise there won't be any notification if the website changes and the parsers fails.

With your feedback above, I changed the parse_start_date function so it should not throw an exception (unless something goes really wrong and as you stated in that case we want the parser to fail).
As a result the parse_date function is also a lot shorter than before :)

klemens · 2018-05-04T18:34:13Z

parsers/thueringen.py

+			else:
+				canteen.addMeal(current_date.date(), 'Hauptgerichte', main_dish, notes, costs,
+								None)
+	for i in range(min(days_open), 9):


This will fail if days_open is empty. Also i should be day_id like above.

I modified this line into for day_id in range(min(days_open or [9]), 9): so if days_open is empty the loop will be skipped

klemens · 2018-05-04T18:49:42Z

parsers/thueringen.py

+	for i in range(min(days_open), 9):
+		if i not in days_open:
+			closed_date = mensa_start_date + timedelta(days=i-2)
+			if not today or closed_date.day is datetime.today().day:


Same as above regarding is and the date comparison.

klemens · 2018-05-04T18:53:35Z

parsers/thueringen.py

+		for meal in meals:
+			main_dish = meal[0]
+			notes = meal[1]
+			costs = meal[2]


You can destruct this in one line:

main_dish, notes, costs = meal

klemens · 2018-05-04T18:58:36Z

parsers/thueringen.py

+			if guests_fee is not -1:
+				costs['other'] = costs['student'] + guests_fee
+			if today:
+				if current_date.day is datetime.today().day:


The is operator compares object identity. While this "works" in this case due to an implementation detail, you should really just do a regular compare: == Also you should compare the date objects directly and not the day property.

klemens · 2018-05-04T19:02:20Z

parsers/thueringen.py

+									None)
+			else:
+				canteen.addMeal(current_date.date(), 'Hauptgerichte', main_dish, notes, costs,
+								None)


I think you should check if today and current_date != datetime.today() before parsing the day and just continue if it is the case.

- removed additional testing urls - reworked parse_start_date function - global constants for default fees - using != and == instead of is and is not - comparing date objects directly - handling empty days_open list - some general stuff he mentioned

Anarchoschnitzel · 2018-05-05T07:20:33Z

I took care of the changes you requested.
Thanks for your feedback 👍

Anarchoschnitzel · 2018-06-14T08:57:15Z

Hi there, I know you are pretty busy atm, but I just wanted to check if there is any progress regarding the PullRequest :)

klemens

Looks pretty good already!

The only missing larger issue is the use of readable ingredient notes. It would also be nice to extend the feed to the following two weeks, but we can also implement this later.

The rest are just some minor improvement suggestions.

klemens · 2018-06-17T13:11:06Z

parsers/thueringen.py

+	fees = parse_fees(document)
+
+	employees_fee = fees[0]
+	guests_fee = fees[1]


employees_fee, guests_fee = parse_fees(document)

klemens · 2018-06-17T13:13:54Z

parsers/thueringen.py

+			if employees_fee != employees_fee_default:
+				costs['employee'] = costs['student'] + employees_fee
+			if guests_fee != guests_fee_default:
+				costs['other'] = costs['student'] + guests_fee


You should use None as the missing values, then you don't need the "constants" and can just check using is not None here.

klemens · 2018-06-17T13:38:09Z

parsers/thueringen.py

+def parse_fees(document):
+	fees_regex = re.compile('Bedienstete.*')
+
+	fees_p = document.find_all('p', {'class': 'MsoNormal'})


You can use class_ instead of the dictionary.

Also, you can directly search for the text using the regex:

fees = document.find_all('p', class_='MsoNormal', string=fees_regex)

klemens · 2018-06-17T13:46:49Z

parsers/thueringen.py

+				continue
+
+			canteen.addMeal(current_date.date(), 'Hauptgerichte', main_dish, notes, costs,
+								None)


Coudn't we use Mittag Ausgabe A, Mittag Ausgabe B, etc as categories instead of Hauptgerichte for everything?

klemens · 2018-06-17T14:15:25Z

parsers/thueringen.py

+
+	for meal_data in meals_t_rows:
+		meal_t_datas = meal_data.find_all('td')
+		if len(meal_t_datas) is not 3:


Please use != instead of is not.

klemens · 2018-06-17T14:25:30Z

parsers/thueringen.py

+
+		ingredients = ingredients_regex.findall(meal_t_datas[1].text)
+		if len(ingredients) > 0:
+			notes.append(ingredients[0].strip())


You should add a single note for every ingredient. Also please parse the table at the bottom of the page to write out the abbreviations in full. There are some helpers for this in https://github.com/mswart/pyopenmensa/blob/7b6e7a70ab63530af27b985019f28863f8b8f368/feed.py#L182-L246 that can parse ingredient tables and automatically replace the abbreviations. I can also help you out with this if you want.

klemens · 2018-06-17T14:30:39Z

parsers/thueringen.py

+		students_fee = float(students_fee_string[0].replace(',', '.'))
+		costs = {'student': students_fee}
+
+		meals.append((main_dish, notes, costs))


You can just yield the tuples instead of storing them in a list and then returning the list.

klemens · 2018-06-17T14:41:20Z

parsers/thueringen.py

+	canteen = LazyBuilder()
+
+	content = urlopen(url).read()
+	document = parse(content, 'lxml')


You are only parsing the current week currently, which means that the full fead does not contain the meals for Monday on Sunday, but last Friday instead, which is not very useful.

To get the next week, you have to send a request with the selWeek=2018-25 parameter, where 25 is the calender week. It seems you can also get a list view using selView=liste which contains the explicit date, which might be easier to parse that the tabbed view.

The website uses POST-requests, but it seems GET also works: https://www.stw-thueringen.de/deutsch/mensen/einrichtungen/ilmenau/mensa-ehrenberg.html?selWeek=2018-25&selView=liste

- implemented the syntactic changes klemens requested - categories are now named by Ausgabe A,B,C etc. according to the studentenwerk website - ingredients are parsed according to the table on the bottom of the studentenwerk website, and appended in full as notes (one note for each ingredient / allergen) to the meals - meals are fetched for the current week + 2 as klemens suggested via GET request (selWek=...)

Anarchoschnitzel · 2018-06-20T12:41:41Z

First of all, thanks for your feedback! I haven't worked much with python so far so every bit of feedback that helps me to enhance my knowledge of the language is great 👍
I have only one small question left regarding the changes I implemented: as I am now fetching the meals for the canteens for the current week + 2 - as you suggested - it sometimes happens that the meals for 2 weeks in advance are not fully present on the Studentenwerks pages. This results in some week tabs not being there which my parser interprets as "canteen closed on that day" (which is in theory possible during the week etc. for holidays). Now my question is: if the OpenMensa server calls my parser later again and the data for the week is there so days that were previously returned as closed are returned as open and with meals, will it overwrite the existing information for "closed on that day" or ignore it?

Thanks for your help in advance :)

-btw I referenced the wrong issue in my commit message by accident, sorry for that ;)

klemens · 2018-06-20T15:30:43Z

if the OpenMensa server calls my parser later again and the data for the week is there so days that were previously returned as closed are returned as open and with meals, will it overwrite the existing information for "closed on that day" or ignore it?

I am not 100% sure, but I think that newer data always overwrites older data. Alternatively, you could just return no data for those days and not bother with the "canteen closed" status at all.

I referenced the wrong issue in my commit message by accident, sorry for that ;)

Don't worry, we can clean up the commit history before merging!

- days that dont appear on a tab on the studierendenwerk webpage are now ommited instead of returned as closed - now with correct PR link ;-)

Anarchoschnitzel · 2018-06-20T15:50:39Z

Allrighty, days that don't appear on the websites tabs are now being omitted 👍

klemens

Thank you for the update! We are almost there, just some minor fixes remain. I also added some improvement suggestions, as you seem to like those. 😉

klemens · 2018-06-20T15:50:19Z

parsers/thueringen.py

+
+def parse_url(url, today=False):
+	year = datetime.now().isocalendar()[0]
+	week = datetime.now().isocalendar()[1]


year, week, _ = datetime.now().isocalendar()

klemens · 2018-06-20T15:51:43Z

parsers/thueringen.py

+
+	canteen = LazyBuilder()
+
+	for i in range(0, 3):


NIT: Zero is the default, so you can just use range(3).

klemens · 2018-06-20T15:57:46Z

parsers/thueringen.py

+
+	for i in range(0, 3):
+
+		content = urlopen(url+'?selWeek='+str(year)+'-'+str(week + i)).read()


This formats the week number without a leading zero if it is less then 10. Now I have no idea what this number should look like (and I guess we have to wait until the end of the year to find out), but if I were to implement the website, I would use a leading zero. 😉 You can use a format string to format the number (and the other arguments):

"{}?selWeek={}-{:02}".format(…)

klemens · 2018-06-20T16:02:11Z

parsers/thueringen.py

+			day_id = int(day['id'][-1:])
+
+			current_date = mensa_start_date + timedelta(days=day_id-2)
+			digits, chars = parse_ingredients(document)


This is the same for all days, so you only need to parse it once.

klemens · 2018-06-20T16:06:19Z

parsers/thueringen.py

+	option = document.find("option", value=calendar_week_regex, selected=True)
+
+	if option is None:
+		return None


Why do you return None here? You don't check for None after calling the function and I think the parser should fail if it can't extract the date.

klemens · 2018-06-20T17:01:45Z

parsers/thueringen.py

+		if chars.get(ingredient) is not None:
+			ingredients.append(chars[ingredient][1:])
+
+	return ingredients


You can do this whole function in one line using list comprehensions if you want to:

notes = [legend[note] for note in notes if note in legend]

klemens · 2018-06-20T17:05:28Z

parsers/thueringen.py

+		if len(category) == 1:
+			category = category[0]
+		else:
+			category = 'Hauptgerichte'


len(category) == 1 will always be true, because you just created a list with length 1 above? Did you mean to check if the string was empty?

klemens · 2018-06-20T17:10:01Z

parsers/thueringen.py

+		ingredients = ingredients_regex.findall(meal_t_datas[1].text)
+
+		if len(ingredients) > 0:
+			ingredients_list = ingredients[0].strip('Inhalt: ').strip('\r').split(',')


Please use strip() instead of strip('\r'). Also you should strip the resulting list items (maybe combined with the list comprehension from above?), because there may be some spaces in there sometimes.

klemens · 2018-06-20T17:16:21Z

parsers/thueringen.py

+
+	canteen = LazyBuilder()
+
+	for i in range(0, 3):


You should probably only fetch the current week when the today parser is requested.

klemens · 2018-06-20T17:32:44Z

parsers/thueringen.py

+					costs['employee'] = costs['student'] + employees_fee
+				if guests_fee is not None:
+					costs['other'] = costs['student'] + guests_fee
+				if today and current_date != datetime.today():


./parse.py thueringen il-ehrenberg today.xml currently returns an empty result for me, because this comparison between dates is not actually between dates, but between datetimes. It should be current_date != date.today() and parse_start_date should return a date instead of a datetime (just use .date() before returning).

- implemented the proposed syntactic changes - weeks to be fetched now determined from <select> elements options in the document so only weeks that are available from the website are being fetched. As a bonus we don't need to care about formatting with leading zeroes etc. - variables that don't change are only being parsed once - fess should be parsed even if more paragraphs with Bedienstete etc. exist - reworked the ingredients parsing as proposed which ultimately led to the drop of a whole function as this could be done in a one-liner - dates should now be correctly compared without time so the today feed should work again

Anarchoschnitzel · 2018-06-21T08:40:53Z

I implemented the proposed changes.
I also found a ( I think better / more reliable) way to fetch meals for upcoming weeks, as I now parse the selection element from the document which has all available weeks for the future as options. This should also remove the necessary of bothering with leading zeroes in the ?selWeek parameter

klemens

You are right, parsing the actual values is much better!

The code looks really good now, there are just some optimizations remaining. 👍

klemens · 2018-06-27T18:19:57Z

parsers/thueringen.py

+		for s in table.stripped_strings:
+			if groups_regex.search(s):
+				split = s.split(':')
+				groups[split[0]] = split[1]


The object returned by search already contains the matched groups (the parenthesis in the regex), which can be accesses using group(1) and group(2), so you don't need split. You should also strip the description here instead of in the list comprehension in line 84.

klemens · 2018-06-27T18:23:17Z

parsers/thueringen.py

+		ingredients = ingredients_regex.findall(meal_t_datas[1].text)
+
+		if len(ingredients) > 0:
+			ingredients_list = ingredients[0].strip('Inhalt: ').strip().split(',')


Sorry, I missed this in the last review. strip('Inhalt: ') probably does not do what you think it does. (I suggest just using [8:] or a regex)

klemens · 2018-06-27T18:29:41Z

parsers/thueringen.py

+	content = urlopen(url).read()
+	document = parse(content, 'lxml')
+
+	available_weeks = parse_available_weeks(document)


This fetches the current week's site twice. I suggest skipping the current week and using the already available document for adding the meals. It probably makes sense to split the parse_url function into two functions, one which fetches the sites and checks which are available and the other does the actual parsing (what's currently in the for loop).

klemens · 2018-06-27T18:31:53Z

parsers/thueringen.py

+					costs['employee'] = costs['student'] + employees_fee
+				if guests_fee is not None:
+					costs['other'] = costs['student'] + guests_fee
+				if today and current_date != date.today():


This still fetches all sites, even when just the first one is needed. You need another check that only fetches the current site for today feeds (see also the comment above).

klemens · 2018-06-27T18:42:03Z

parsers/thueringen.py

+
+
+def parse_available_weeks(document):
+


NIT: I find empty lines after function definitions of for loops (see below) a little bit confusing. They also seem to confuse the git diff algorithm. 😉

- implemented the proposed syntactic changes - removed blank lines under function definitions - using matched groups instead of string splitting for ingredients parsing - sites should only be fetched when necessary (first page only once, following pages only if not today etc.) - parsing of meals for canteen now separated in new function parse_meals_for_canteen so parse_url only responsible for fetching necessary information (html documents, weeks, groups, fees etc.) and returning the xml feed

Anarchoschnitzel · 2018-06-29T07:21:42Z

I implemented the proposed changes - thanks for your feedback 👍

Anarchoschnitzel · 2018-07-01T13:53:26Z

I just found out my solution might not work properly yet. If accessed on weekends / Sundays the Studentenwerk page does not auto select a week which currently crashes the parser on those days. I am working on a fix.

If the start date of the initial loaded page cannot be parsed (can happen on weekeends when the seleciton box only shows ---Auswahl Treffen---) the first load of the page by ?selWeek isn't skipped.

Anarchoschnitzel · 2018-07-01T14:07:37Z

Should be fixed now

klemens · 2018-07-28T16:39:24Z

Sorry for – once again – taking so long. The changes look good, I just pushed a small change to replace the assert with a normal check for Null, as exceptions should (normally) not be used for control flow.

I will now merge this (sqashed into one commit) into the master branch, which means that it will become available at http://omfeeds.devtation.de/thueringen/*location*/{full,today}.xml within a day or so. Then you can request an update or the transfer of the parser to your account, if you want. Otherwise, I can also do this.

Thanks again for writing the parser and following my (sometimes nitpicky) review! 😁

Anarchoschnitzel · 2018-07-28T19:37:50Z

Thanks for the update!
If I find the time I will head over to the openmensa project page and request an update in the upcoming days.

Again thanks for the continuous feedback, was really fun writing this parser and learning a few new things and best practices about python 👍

klemens · 2019-02-06T22:48:15Z

Soooo, I just moved all of the already existing canteens into one parser (they were separated by city) and adapted their names to your new naming scheme. After adding the new index-url of my server, they started to work immediately. 😉

I did not add the other canteens yet, mainly because I wanted to automate it (see the halle parser and my changes to it today). My plan is as follows: The city overview page (eg Jena) contains a list of canteens including their addresses. It also contains a hidden form, which contains geo-coordinates for all canteens (used for the map display). This way we should be able to generate the complete metadata info and avoid copying it by hand into openmensa (which of course would be much faster, but missing all the fun 😁).

klemens · 2019-02-09T16:12:21Z

I implemented the metadata extraction in 90b95c9 and fixed a small issue, where an empty meal name would lead to an exception, because they are not allowed. Now, empty meals are just skipped: fddcca4

The extraction worked great for all but three canteens:

These three were missing in the map display, so I created them manually:

I did some quick sanity checks (the coordinates are in the right city), but it would be great if you could also take a look!

Anarchoschnitzel added 6 commits April 25, 2018 21:58

Add parser for Mensa Ilmenau (Ehrenberg, Cafeteria, Nanoteria and Men…

f588de1

…sa Roentgen)

adjusted week_start_end_date_regex

a5cdb69

added test url for ilmenau / ehrenberg mensa

0b1b521

added parsing of employee and guest fees for meals

a10ace3

fixed cut-off string literal

a27d3d1

Anarchoschnitzel changed the title ~~Added Parser for the mensas in Ilmenau~~ Added Parser for most of the Mensas from Studentenwerk Thueringen Apr 28, 2018

j-maas requested changes Apr 28, 2018

View reviewed changes

j-maas added the new parser label Apr 28, 2018

Anarchoschnitzel added 2 commits April 28, 2018 16:02

Extraced parsing functionality into functions to avoid deep nested code.

c8bac58

Changed default behaviour of non-available days in the past so that they are ignored instead of treated as closed. Removed try-catch clause

Fixed crash when no separate fees for employees and guests are provid…

77e20c8

…ed (could happen at some canteens). Added error handling if the start date of a canteens current week could not be parsed

j-maas approved these changes Apr 28, 2018

View reviewed changes

Anarchoschnitzel mentioned this pull request May 4, 2018

Change of Parsers for Canteens in Thuringia #81

Closed

klemens requested changes May 4, 2018

View reviewed changes

Merge branch 'master' of git://github.com/mswart/openmensa-parsers

18f72a4

klemens requested changes Jun 17, 2018

View reviewed changes

Anarchoschnitzel added 2 commits June 17, 2018 18:54

Merge remote-tracking branch 'upstream/master'

b6e204b

mswart#79:

68538c5

- days that dont appear on a tab on the studierendenwerk webpage are now ommited instead of returned as closed - now with correct PR link ;-)

klemens requested changes Jun 20, 2018

View reviewed changes

klemens requested changes Jun 27, 2018

View reviewed changes

mswart#79:

030fa72

If the start date of the initial loaded page cannot be parsed (can happen on weekeends when the seleciton box only shows ---Auswahl Treffen---) the first load of the page by ?selWeek isn't skipped.

thueringen: check if start date exists instead of asserting

b62e11f

klemens approved these changes Jul 28, 2018

View reviewed changes

klemens merged commit 9691720 into mswart:master Jul 28, 2018

klemens mentioned this pull request Feb 4, 2019

Upload of new packages from travis to apt.dxtt.de fails with 406 Not Acceptable #89

Open

klemens mentioned this pull request Feb 9, 2019

Updated Parsers for Studentenwerk Thüringen openmensa/openmensa#37

Closed


		for i in range(0, 3):

		content = urlopen(url+'?selWeek='+str(year)+'-'+str(week + i)).read()

Added Parser for most of the Mensas from Studentenwerk Thueringen #79

Added Parser for most of the Mensas from Studentenwerk Thueringen #79

Conversation

Anarchoschnitzel commented Apr 26, 2018

j-maas commented Apr 28, 2018

Anarchoschnitzel commented Apr 28, 2018

j-maas left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Anarchoschnitzel commented Apr 28, 2018

j-maas commented Apr 28, 2018

Anarchoschnitzel commented Apr 30, 2018 • edited Loading

j-maas commented Apr 30, 2018

klemens left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Anarchoschnitzel commented May 5, 2018

Anarchoschnitzel commented Jun 14, 2018

klemens left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Anarchoschnitzel commented Jun 20, 2018 • edited Loading

klemens commented Jun 20, 2018

Anarchoschnitzel commented Jun 20, 2018

klemens left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Anarchoschnitzel commented Jun 21, 2018 • edited Loading

klemens left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Anarchoschnitzel commented Jun 29, 2018

Anarchoschnitzel commented Jul 1, 2018

Anarchoschnitzel commented Jul 1, 2018

klemens commented Jul 28, 2018

Anarchoschnitzel commented Jul 28, 2018

klemens commented Feb 6, 2019

klemens commented Feb 9, 2019

Anarchoschnitzel commented Apr 30, 2018 •

edited

Loading

Anarchoschnitzel commented Jun 20, 2018 •

edited

Loading

Anarchoschnitzel commented Jun 21, 2018 •

edited

Loading