rumca-js / django-link-archive Goto Github PK

Provide more robust unit tests

Right now we have only git exports.

Spotify does not allow web scrapping. To obtain links from a page contents we would need to use real browser by selenium.

https://docs.python.org/3/library/urllib.robotparser.html

Right now templates need to be modified in place. There is no 'project' directory where user-specific changes can be done. Right now there is no plan about how to implement this.

import / export should use some sort of generic interface

Something like:
site/add-new-link/https://whatever

It should be done by GET, which is simpler. Maybe later on there should be a browser extension to support dropping the link into database.

If a browser extension supports POST, then maybe we should go with POST.

I think a logged user should be able to add new links.

Disable data range from search form for users that are not logged in.
To prevent bad actors.

Check docker makefile if works

Twitter had its limitation. Every time a tweet was made, the number of characters was limited. I would like for the people to really think through their comments.

Low hanging fruit:

• Most of the search form inputs should be hidden

Goal:

• should be similar to https://confluence.atlassian.com/jirasoftwareserver/advanced-searching-939938733.html
• should use javascript, to be more responsive

1. do not use sudo in install script
2. use poetry, or virtualenv, or even docker
3. requirements should specify versions of packages

Close tasks Admin button, or use different method than python tasks to process messages.

Right now we have BasePage, HtmlPage, RssPage. It is true that we can have separate handlers, it is also true, that most common actions should have same API.

Should it be in BasePage?

• get_title
• get_description
• get_author
• get_language
• get_thumbnail
• get_content_rating
• is_valid

Should be implemented in a different table (different from LinkDataModel).

For example for substack things I would like to add "substack" tag.

Add import mechanism for sources

Export domain list as a plaintext

Entry export - should export ID of source.
LinkDataModel field "source" should be left as is. Some urls are added without having a proper "source".

Domain view should be consistent with entries view

Add import behavior for git repositories (daily data)

Create docker installation