Skip to content

Latest commit

 

History

History
114 lines (83 loc) · 4.2 KB

PLUGINS.md

File metadata and controls

114 lines (83 loc) · 4.2 KB
Raw

Search Plugins

Search plugins can be added from the search screen by clicking on the manage repository button. They let the user search for files on a specific website. The extension is .unchained but it is just json.

Install plugins

Plugins can be installed in multiple ways:

  • download a plugin and click on it, select Unchained when asked which app to use
  • share a link to a plugin file to Unchained
  • add a repository link from the plugins screen and then refresh the plugins list

Uninstall plugins

  • a button to remove all the plugins is available in the settings menu
  • using the plugins screen, click on the remove button

Create a new repository

The repository json structure is pretty simple:

{
    // this is the repository engine version used for future development, not the version of the json file
    "repository_version": 1.0,
    "name": "Unchained",
    "description": "Unchained main repository",
    "author": "LivingWithHippos",
    // plugins list
    "plugins": [
    {
        // the id MUST match the one inside the plugin file
        "id": "etree",
        // every plugin can have different versions. The version should match the one in the plugin file
        "versions": [
        {
            // version of the plugin itself, chosen by the author
            "plugin": 2.0,
            // engine version of the plugin, used to check for compatibility with Unchained
            "engine": 2.0,
            // link to the plugin file itself
            "link": "https://gitlab.com/LivingWithHippos/unchained-plugins/-/raw/main/repository/plugins/etree/etree_v2.0.unchained"
        }
        ]
    }
    ]
}

If you don't know where to publish your repository and plugins, just create a project on Github, Gitlab, Sourcehat, Codeberg or whatever and use that.

N.B. the link to the repository to be used/shared must directly open the text file, test it in your browser. On Github you can get it by pressing the "raw" button on the file page.

Create a new Plugin

The structure of a plugin file is more complex to allow for the necessary flexibility for every website layout. Websites requiring login or protected by Cloudflare are not searchable yet.

To create a new plugin you can take inspiration from the "official" ones. The basic skills required are:

  • some knowledge of json
  • some knowledge of html
  • some knowledge of regular expressions

Check the template.json5 file under extra_assets/plugins for a more detailed explanation of the various fields.

Websites are classified by how we can retrieve the download links from the search page. Magnet/Torrents and hosting files are supported.

Internal links

The search page provides a list of links that points to the searched item page, which contains the download links. It's the slowest since a web request must be made for every result.

Search page -> Details page -> download links

Direct links

The search page has the links available directly on it. Every result is separated by an html class common for every search result.

Search page -> html table class/id -> download links

Direct links on table

The search page has the links available directly on it. Every result is a row in an html table.

Search page -> html table class -> download links

Indirect links on table

The search page has the results in an html table but the download links are in the details page. Similar to the "internal links" one.

Search page -> html table class -> Details page -> download links

Every search to be performed on a page for links or details (the result name, seeders etc.) has the same form:

// the field we are searching for. These are fixed, seeders, leechers, size, name...
"field name": {
    // which regex to use if multiple ones are available
    "regex_use": "first",
    // list of regexes
    "regexps": [
    {
        // the regex itself. Java style.
        "regex": "href=\"(/torrent/[^\"]+)",
        // group capturing is supported. Always add one.
        "group": 1,
        // how to treat the result. Is it completed, partial...
        "slug_type": "append_url",
        // optional field to be prefixed to the link found if slug type is append_url
        "other": "https://other_url.com"
    }
    ]
}