v0.0.2

Minimum Viable Product of Web-Raider

ChangeLog (from v0.0.1)

• Fixed NoneType does not have iterlinks() method error which occurs when identified CodeArticle URLs do not have embedded links.

v0.0.1

Minimum Viable Product of Web-Raider

ChangeLog (from v0.0.0)

• Module imports changed to relative import for backend integration
• Configure script to be executed using poetry under [tools.poetry.scripts]

v.0.0.0

Minimum Viable Product of Web-Raider

What is Web-Raider?

A web browsing agent which takes in a user query and returns a list of codebase URLs that were deemed relevant to the user query.

Current Features

1. Extract relevant information from urls that are directed to codebases (GitHub, GitLab, BitBucket and Gitee)
2. Comb through articles for urls that could direct to codebases listed above
3. Find relevant information on retrieved codebases and determine whether they are relevant to the query

To implement

1. Use Classification Models to better classify URLs instead of hard-coding logic
2. Handle Forum URLs
3. Scrape Code directly from Articles and Forum URLs
4. Proper implementation of user query decomposition (to find subqueries)