mahesh-hegde / rrip Goto Github PK

Some sites like gfycat links are not directly downloadable.

It's a good idea if user can specify something like:

{"gfycat.com": "/head/meta[@property=\"og:video\"]@content"}

Don't remember XPath syntax but that's beside the point.

• Refactor code to allow automated testing

• Unit tests

• Integration tests

Requirements

• Filtering should be possible based on arbitrary attributes of the post
• Can set final file name & link format using arbitrary attributes of the post

The best approach seems to be pass the post details to a Go template, which can be specified by the user.

Post details should be passed as a JSON Map, to avoid deserializing unnecessary details when this feature is not used.

We can also add additional attributes to this map, like quoted_title.

Many tools (docker & related CLI tools) support golang text/template for similar use cases. rrip can use the same.

Rationale: some image hosts seem to randomly throttle unusual user agents.

• Provide user agent presets for popular browsers: [chrome,firefox,edge] x [macos,linux,windows,android]

• Override user agent on regex basis. Something like

  --user-agent-override=*.imgur.com:chrome-macos

It can be run using bind mounts.

It will protect in hypothetical scenario that an RCE is detected in rrip.

It sounds cool to say "Distroless container image with Docker or Podman".

This flag should download all pictures uploaded after last download.