shells

I created a simple AI detector that uses the features as described in the previous devlog. For now it works well for what I need it to, but I might change train a model if I need to in the future. There is a new script called getDevlogs which basically fetches the text from each devlog in a project (in my database, each project has an array of devlogs which I iterate over in the script). Then it runs my classifier which basically uses math to calculate the AI probability. Then the mean of the scores are calculated and that is the final percentage which you can see near the end of the video

I created a python script that extracts the important features of the input text (either text or code). I will then later use this to calculate the probability of it being AI without using AI to calculate. So far it returns the number of words, number of sentences, mean sentence length (using numpy), buzzword ratio (using a list of buzzwords provided in a txt file) and the punctuation ratio

I created a program that searches the MongoDB database from my previous devlogs and prints out the results. It uses fuzzy search (the $search operator that can be seen in the code in the video) which basically also returns matches even if they aren't exactly the same as the query. It then sorts it by relevance and includes pagination using this line:

skipcount = (page - 1) * pagesize

which is then passed to the project_collection.find() method

I changed the script to use async which fetches data concurrently and is much faster! (30 minutes down to around 3 minutes!!!) It also now uses bulk_write with mongodb, which is much more efficient and reduces the amount of connections needed.

I also got a nest account, which is where I will be hosting the backend (my next step is to make the front end)

I added devlogs to the data that the script updates

I created a python script that uses the SOM projects api (https://summer.hackclub.com/api/v1/projects) to update a MongoDB collection with information about every project in Summer of Making. This data in the data base will then be used to calculate the estimated shell profit