Open House AI API Test

Open House coding test

Installation

Download python, can be done here.

Next use pip to install flask with this command:

pip install flask

To Run

In the project work directory run on the cmd:

python api.py

Navigate to default page http://127.0.0.1:5000/

API End Points

GET

First one is http://127.0.0.1:5000/api/v1/resources/user/all

This endpoint retrieves all the user entities of the database. It takes no additional params and is primarily for sanity testing the POST request to input user entities into the database.

Second one is http://127.0.0.1:5000/api/v1/resources/action/all

This endpoint retrieves all the action entities of the database. It takes no additional params and is primarily for sanity testing the POST request to input action entities into the database.

Third one is http://127.0.0.1:5000/api/v1/resources/logdump

This endpoint is the solution to the whole coding test you guys gave out. It serves to extract logs based on specified param(s) and outputted in the desired format. It takes in four optional params can be of any combination:

userId : specify a userId you want to retrieve logs of
logType : specify the log type you want to retrieve logs of
lowerTimeRange : specify the lower time range of log time stamps that are equal to or greater than the specified time (note time format is of YYYY-MM-DD HH:MM:SS , example: "2020-08-15 21:56:54")
upperTimeRange : specify the upper time range of log time stamps that are strictly lower than the specified time (note time format is of YYYY-MM-DD HH:MM:SS , example: "2020-08-15 21:56:54")

POST

First one is http://127.0.0.1:5000/api/v1/resources/user

This endpoint is designed for the front end to dump the user id and sessionId into the database. It takes in two mandatory params:

id - Mandatory. This is the user id
sessionId - Mandatory. This is the session id

Second one is http://127.0.0.1:5000/api/v1/resources/action

This endpoint is designed for the front end to dump the action logs into the database. It takes in five params but only one is mandatory (user_sessionId):

user_sessionId : Mandatory. This is the session id for the action and is used to relate back to the user entity when doing queries
type : this is the type of log
locationX : this is the location of the mouse cursor at point x
locationY : this is the location of the mouse cursor at point y
viewedId : this is the viewed id

Example

This example will go over how to generate the output of the sample log provided for this coding project from zero.

Note the database you guys will have when you pull down the gitrepo will have this example already done for you. If you would like to follow along just delete the "sqlite.db" file in the database folder and you should be able to recreate this example.

First we will create the user entity using this post request with these id and sessionId params: http://127.0.0.1:5000/api/v1/resources/user?id=ABC123XYZ&sessionId=XYZ456ABC

You can use this get request to verify http://127.0.0.1:5000/api/v1/resources/user/all

Next we will start adding the action entities using this post request with the specified params:

You can use this get request to verify http://127.0.0.1:5000/api/v1/resources/action/all

Lastly we can use this get request to retrieve all log files from the user id "ABC123XYZ": http://127.0.0.1:5000/api/v1/resources/logdump?userId=ABC123XYZ

This will return us the JSON formatted as the sample log you gave for this coding test. You can also play around with the other params to get a more accurate I've also included a screenshot of the final output for this last get request labeled "Solution.PNG" that can be found in the gitrepo.

Follow Up Question

I would probably look at the documentation for saas offerings, see whether I need a library or anything extra for creating the python api. Would also follow the simple example they usually have in the documentation.

I would also use a different database. Research if there's any cloud based database resources that I could use such as Azure data storage and try and configure how I would create tables, blobs, etc to make this solution work. Of course the database creds would ideally be stored in the pipeline and from there I would research how I would connect to the database and if I needed anything else to execute the sql queries.

STEjovanovicFAN/OpenHouseAi