TODO

chore: initial commit
Deploy first staging version (v1.0.0-staging.1)
Wikipedia Database Dump
- Download SQL files
- Extract SQL files
- Tables structure CREATE TABLE
  - page.sql (pages table)
  - pagelinks.sql (internal_links table)
- Adapt downloaded SQL files
  - page.sql (pages table)
  - pagelinks.sql (internal_links table)
- Import SQL files
- Try SELECT count(*) FROM internal_links il WHERE il.from_page_id = (SELECT p.id FROM pages p WHERE p.title = 'Linux'); -- Count of internal links for 'Linux' page
- Try:
```
SELECT il.to_page_id, pl.title
FROM internal_links il
JOIN pages pl ON pl.id = il.to_page_id
WHERE il.from_page_id = (
  SELECT p.id FROM pages p WHERE p.title = 'Node.js'
);
```
- Move from POC (Proof of concept) in data folder to apps/cli folder
- Documentation how to use + Last execution date
- Rewrite bash script to download and extract SQL files from Wikipedia Database Dump to Node.js for better cross-platform support and easier maintenance + automation, preferably one Node.js script to generate everything to create the database
- Verify file content up to before inserts, to check if it matches last version, and diff with last version
- Update logic to create custom internal_links table to make it work with latest wikipedia dumps (notably concerning the change in pagelinks.sql where the title is not included anymore, but instead it uses pl_target_id, foreign key to linktarget), last tested dumb working 20240420
- Handle redirects
Implement REST API (api) with JSON responses (AdonisJS) to get shortest paths between 2 pages
- Init AdonisJS project
- Create Lucid models and migrations for Wikipedia Database Dump: pages and internal_links tables
- Implement GET /wikipedia/pages?title=Node.js to search a page by title (not necessarily with the title sanitized, search with input by user to check if page exists)
- Implement GET /wikipedia/pages/[id] to get a page and all its internal links with the pageId
- Implement GET /wikipedia/internal-links/paths?fromPageId=id&toPageId=id to get all the possible paths between 2 pages
- Setup tests with database + add coverage
- Setup Health checks
- Setup Rate limiting
- Share VineJS validators between website and api
Implement Wikipedia Game Solver (website)
- Init Next.js project
- Try to use https://www.npmjs.com/package/@tuyau/client for API calls
- Hard code 2 pages to test if it works with console.log in the browser
- Implement a form with inputs, button to submit, and list all pages to go from one to another, or none if it is not possible
- Add images, links to the pages + good UI/UX
- Autocompletion page titles
- Implement toast notifications for errors, warnings, and success messages
Implement CLI (cli)
- Init Clipanion project
- Implement wikipedia-game-solver internal-links --from="Node.js" --to="Linux" command to get all the possible paths between 2 pages.
Add docs to add locale/edit translations, create component, install a dependency in a package, create a new package, technology used, architecture, links where it's deployed, how to use/install for end users, how to update dependencies with npx taze -l etc.
GitHub Mirror
Delete TODO.md file and instead use issues for the remaining tasks

4.2 KiB Raw Blame History

TODO

Links

4.2 KiB

Raw Blame History