Looking for a flat in big cities can be a pain in the eye (ass) due to many different reasons, e.g. you are a student and many landlords don't like students; or you don't speak German so well and landlords, most of who are in their old ages, cannot speak English so well; or there are just simply too many people looking for a nice flat, whenever a new offer apears online the landlord will receive tons of applications and they have nothing to do but to choose the first ones who sent them the applications. So making sure that your application always apear among the first emails in the landlords' inbox is crucial to get you a chance to visit the apartment. I learned the lesson in a hard way by spending a lot of my time online monitoring the updates of apartment offers on immobilienscout, wg-gesucht or ebay everyday, but in the end almost all of my inquiries went straight into the vacuum and never got an echo back. After somedays, I realised that maybe it is true that there are too many people online looking for a flat, and my messages are just among the tons of messages the landlords receive, so to increase my chance, I need to send messages as soon as the offer is put online. So I wrote a Bot in Python to monitor the website immobilienscout every minute, once it finds new offers it will send my message, including some information about ourselves, to the landlord automatically. Basically the whole process won't take more than 2 minutes since the offer is put online.
I think immobilienscout has the most number of new offers everyday, that's why I chose to monitor it. In principle we can make the Bot to monitor other websites, too.
Update 22.04.2019: Now it is also working on wg-gesucht.de, too. Actually, that's where I found my current apartment using the script. On immobilienscout, there are many offers, yes. But unless you are one of the guys with a high income job or you are married, your chance of getting a offer is diminishingly low. From my experience, wg-gesucht is more friendly to students and they responded to my report of a fraud on their website very quickly.
The way to use it is almost the same as the old script. Just the names of the scripts are different. Things to pay attention to are:
- the corresponding files are
- immo.py --> wg-gesucht.py
- submit.py --> submit_wg.py
- immo_spider.py --> wg-gesucht-spider.py
- wg-gesucht.de asks for your username and password, which you need to write in the places in the submit_wg.py script.
- immobilienscout now implements a premium feature which will break down the code but I did not take care to handle, due to reasons already stated in above.
- speaking German increases your chance of getting a offer. Try to force yourself speaking German. :)
- please don't abuse the script by sending requests too frequently to the website, otherwise they could take some anti-measures to rule out the bot, which is bad for everyone who wants to look for a flat using the bot.
- The script can be run on raspberry pi 3b. Some more questions please see (closed) issues before you open a new issue. Thanks.
Good luck to your wohnung hunt.
I have only tested the Bot on Linux (Debian, Ubuntu) and Mac. I am not sure about whether you can do some tricks and make it work on Windows, too. But worth trying.
-
chromedriver. I take chrome as an example since it is my favorite browser, but you can also use other popular browsers because you can find their drivers. You need to have chrome installed.
On Linux use the followling command to check if you have chromedriver installed
which chromedriverif it is installed, then something like the following which tells you the path to the driver should appear
/usr/local/bin/chromedriverif nothing appears, then you need to install chromedriver, using the following command on Linux (Debian or Ubuntu),
sudo apt-get install chromedriverand use the following command on Mac (if you use homebrew),
brew install chromedriver -
Tested only with Python 2.7, there might be some syntax errors with Python 3 in the print function. The solution is just to write all print functions with parentheses.
-
Scrapy, which is a package based on Python for writing web spiders. After/if you have python installed, then the following command should install Scrapy for you
pip install scrapy
There are just 3 python scripts.
immo.py
immo_spider.py
submit.py
Get the scripts by
git clone https://github.com/nickirk/immo.git
Go into the directory:
cd immo
- In
submit.pyfile, you will see
last_name.send_keys("last name")
first_name.send_keys("first name")
street.send_keys("your current living street")
and etc. Replace the text with your own information. Especially in
text_area.send_keys(u"Hallo,\n\n your message to the landlord. keep the 'u' before the message to make showing German in the message available. use \n as newline in your message.")
write your message to the landlord. Please keep the 'u' in front of the message so that the special German letters will show in the message on the browser side. After this line, there are no more things needed to be modified.
-
In
immo_spider.pyfile, you need to replace the website links that fits your need. For example,start_urls = [ 'https://www.immobilienscout24.de/Suche/S-2/Wohnung-Miete/Umkreissuche/Stuttgart/70569/-64516/2093406/-/-/5/1,50-/-/EURO--850,00', 'https://www.immobilienscout24.de/Suche/S-2/Wohnung-Miete/Fahrzeitsuche/Stuttgart/70569/-64516/2093406/-/-/30/2,00-/-/EURO--850,00' ]
You can go to immobilienscout and enter your filter, e.g. around 70569 Stuttgart within 5 km or 30 mins away, price until 850 Euros. Click search, you will arrive at the page showing you the results. However, you need to choose the realtime (Aktualität) sorting so that the results you see are always the latest offers. Then copy the link address and paste it into the start_urls. You can put more than one links to it separated by comma.
First, we need to create a new scrapy project called immobot:
scrapy startproject immobot
then you will have the following structure of directories and files:
immobot/ #Working directory
scrapy.cfg # deploy configuration file
immobot/ # project's Python module, you'll import your code from here
__init__.py
items.py # project items definition file
pipelines.py # project pipelines file
settings.py # project settings file
spiders/ # a directory where you'll later put your spiders
__init__.py
Now let's go to the working directory called immobot which contains the file scrapy.cfg and another direcory which is also called immobot:
cd immobot
Now copy the 3 files into the following directories:
cp ../immo.py ../submit.py .
cp ../immo_spider.py ./immobot/spiders/
after this you will have the following structure of directories and files
immobot/ #Working directory
immo.py
submit.py
scrapy.cfg # deploy configuration file
immobot/ # project's Python module, you'll import your code from here
__init__.py
items.py # project items definition file
pipelines.py # project pipelines file
settings.py # project settings file
spiders/ # a directory where you'll later put your spiders
__init__.py
immo_spider.py
In the end, just run
python immo.py
under you working directory immobot, the Bot will be running and doing everything for you.
Further simpilfications of the scripts will be done to make it a blackbox tool.
You may run into issues, hopefully these tips can help:
chromedriver may be installed somewhere else then asumed by the script. You can check this by running which chromedriver, it's result should be: /usr/local/bin/chromedriver. If it's not, then change this value in submit.py in the line with webdriver.Chrome('<pass in your value here>').
you've made changes, but nothing changed. Remember to copy the .py files into the immobot folder.
testing can be easily done by removing one of the id's from the diff.dat file. During the next check, the script will just consider this specific advertisement as a new one.