Duplicate graph

A tool for building a visual graph of all the duplicate issues for a repo

Create a personal access token on github, and slam it in a file named secrets.json, like so:

{
    "GITHUB_API_TOKEN" : "your_token_here",

    "REPO" : "microsoft/terminal"
}

Run these scripts to build the graph

# outdated steps
# python main.py > duplicate-issues.json
# python filter_issues.py duplicate-issues.json > short-dupe-issues.json

# Gather all the "Resolution-Duplicate" issues, and all their comments
python main.py > long-duplicate-issues.json
# Build a graph.json that includes all the issues referenced by the duplicate issues
python build_graph.py long-duplicate-issues.json > graph.json

Then,

run a local python webserver to be able to deliver the graph.json to the wepbage.

In python 3:

python -m http.server

then navigate to localhost:8000/graphs.html

Attempt 2

• Use gather_all.cmd to download all the issues into the database locally
• Use collect_dupes.py to process issues to find all the dupes and mentions in issues
• Use list_dupes.py to just parse an issue's body to find dupes and mentions. Relatively unimportant.
• Use dump_issues.py to list info about an issue
• Use process_raw_data.py to parse the issue's json into the database (body, title, state)
• Use database_to_graph.py to dump the data from the database to json in the json format we want.

Contributing

After cloning the repo, run pip install -r requirements.txt

Code formatting

This project uses black for code formatting. To format the code, run black .. You might need to add your python scripts directory to your path. For me, this was done with:

set PATH=%PATH%;%localappdata%\Packages\PythonSoftwareFoundation.Python.3.7_qbz5n2kfra8p0\LocalCache\local-packages\Python37\Scripts

On a newer machine:

set PATH=%PATH%;%localappdata%\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\Scripts