A tiny test tornado webserver that serves a single endpoint answering queries about dogs in New York City

(As an aside, I've been working as a Python Developer in Data Science in a closed-source environment for 3.5 years, not that I've been completely inactive on GitHub during that time, but it's nice to be able to contribute even a small project that's properly engineered instead of the run-once data science "cowboy code" I've tended to quickly commit in the past.)

Please keep in mind that this is after two days of study of Tornado, and one previous project with asyncio. I have extensive experience with MongoDB, but this is the first time I've used the asynchronous client motor.

Very simple, usage is:

python3 doggo.py <port number>

make sure your port is open and you have permissions, etc. The requirements.txt file lists the packages you need; requests is only for the test program, apitest.py, which also needs the environment variable READY_TEST_BASE_URL set to the URL with port number.

The dataset is in MongoDB; its downloading and installation instructions are in the /source_data folder, as is the downloaded CSV itself (a bit more than 5 MB) so you can just use that if you want.

The static directory only contains a favicon, which is kind of ridiculous because this is a REST API not designed to be accessed through a browser, but what the heck, it's a cute doggo! Also robots.txt because I don't like to pollute webcrawlers with my ephemera.

Here's an example of the first row of the table, with the field names

Field names have to be lower-case, but values are case-insensitive. The total number of dogs is 81542

So if you want to know how many male dogs are named buddy, for example, you could curl a GET request to /count with params={"dog_name": "bUdDy", "gender": "m"} Or if you really want to see the cute favicon, goto http://ipaddress:port/count?dog_name=bUdDy&gender=m