agile-computing-gp

Deployment

I recommend using Visual Studio Code to develop the site, and its built-in terminals to run commands.

Setup instructions

So far we have identified two possible ways to run this:

1. On mac
2. On WSL1 Ubuntu 18.04, but with the mongodb server on Windows 10

On WSL1 Ubuntu + Windows:

1. Download Ubuntu 18.04 from Windows store https://www.microsoft.com/en-us/p/ubuntu-1804-lts/9n9tngvndl3q?activetab=pivot:overviewtab, open in VSCode https://code.visualstudio.com/docs/remote/wsl
2. In the WSL terminal, install python and anaconda
3. git clone this repository somewhere
4. Create the conda python environment in agilecompenv, specifically with conda create --prefix="/home/[your username]/agile-computing-gp/agilecompenv" --strict-channel-priority python
5. "Activate" the environment in your terminal, with conda activate /home/[your username]/agile-computing-gp/agilecompenv
6. Install anaconda within the environment: pip install anaconda
7. Install fasttext within the environment: pip install fasttext
8. Add conda-forge channel to conda: call conda config --append channels conda-forge
9. Install scraper requirements: call conda install --file backend/scraper/requirements.txt
10. Install nlp model requirements: call conda install --file backend/nlp_models/requirements.txt
11. Run this: python -m spacy download xx_ent_wiki_sm
12. And this: python -m spacy download en_core_web_sm
13. On Windows, download the necessary files from Model Files as described in the folder's README https://drive.google.com/drive/u/1/folders/1P_-hKTWsxYOPcK0mThp28UQbNbGraGe7
14. Move the downloaded files from Windows into your subsystem's combined_model folder. On Windows, this folder is located at e.g. [your AppData folder]\Local\Packages\CanonicalGroupLimited.Ubuntu18.04onWindows_79rhkp1fndgsc\LocalState\rootfs\home[your unix username]\agile-computing-gp\backend\nlp_models\combined_model
15. Now back in the WSL terminal, free the permissions on files in combined_model using chmod -R 777 backend/nlp_models/combined_model
16. Insert the line fasttext.FastText.eprint = lambda x: None to the top of predict.py and match.py in combined_model to prevent them from throwing warnings as errors.
17. Install npm and node by following these specific instructions: https://docs.microsoft.com/en-us/windows/nodejs/setup-on-wsl2
18. Install node dependencies with npm install in the project root
19. Install node dependencies with npm install in /backend
20. On Windows, install mondodb. https://www.mongodb.com/try/download/community
21. On Windows still, create a folder called "mongodata".

Running the site:

1. On Windows, run mongod --dbpath [path to mongodata folder]
2. Back on WSL, run the frontend server with nodemon server in the project root.
3. Run the backend server with nodemon server in /backend.
4. You should be done :)

This hosts the frontend on localhost:3000, and backend on localhost:5000. The servers will refresh if you change any files.

Database interaction

If you want to interact with the database server through the command line (e.g. manually adding entries, altering schema, listing data), you can open a new terminal and simply use the command mongo to load a shell through which you can send commands.

Examples:

- use task-management-site
      switch to database
      
- show tables
      show all tables in database
      
- db.tasks.insert([{“title”: “Test Task”, “description”: “Lmao test”}])
      insert record to table
      NOTE: creates new table if doesn’t exist, inserts _id field automatically
      
- db.tasks.find().pretty()
      print table contents

Common Errors

Frontend: nodemon.ps1 cannot be loaded because running scripts is disabled on this system
      -> Fix: delete nodemon.ps1, kill all node background processes, retry
          
Frontend: 'react-scripts' is not recognised as an internal or external command
      -> Fix: delete package-lock, yarn.lock, and node-modules, run npm clean cache, then npm install againS