default parser user submitted database idea

[ARYAN-9099]

Is your feature request related to a problem? Please describe.
Many websites work fine using default parsers.

Describe the solution you'd like
A option to use other user submitted settings for default parser for that site. This will reduce the add parser related issues and also we can further automate creation of specific site parser using this data if a setting works well.

[gamebeaker]

We have to watch out that it doesn't violate the policies of the extension stores.

chrome: https://developer.chrome.com/docs/webstore/program-policies/policies#additional_requirements_for_manifest_v3

2. Execution of logic from a remote source is permissible only when accomplished through a documented API that explicitly allows this practice and the use is inline with the documented purpose of the API, as detailed in the API Use policy. The permitted APIs for such remote execution are:
   Debugger API
   User Scripts API
   Note that exemptions apply solely to the specific section of code covered by these APIs. Extensions may still be in violation of this policy if they employ alternative methods to execute logic from remote sources elsewhere in their code.

I think it is debatable if it is logic from a remote source as it would just provide css selectors and the logic on how to use them is already in the extension.

I think it should be an active choice from the user before WebToEpub searches the e.g. github project for css selectors. Something like "search github for possible css sselectors".

On the other hand it is pretty easy to write a parser if the css selectors are known. Maybe instead of fetching them from another project there should be an new Parser abstraction that ingests json so that it isn't neccessary to write a whole parser only a short json with the css selectors defined.

[ARYAN-9099]

I don't think it will be a problem, other famous extensions like Sponserblock for yt are using it just fine.

Even if we make a auto parser maker it will need to go through the entire testing phase then be packaged in the next webstore release.

[gamebeaker]

You are right. It shouldn't be a problem.

[ARYAN-9099]

@dteviot thoughts?

[dteviot]

@ARYAN-9099

I'm not opposed to the idea. But I suggest we start small.
With a Wiki page that lists sites and the values to use with the Default Parser. Could be a simple table.
Should also probably note sites that don't work with the default parser.
Can add a note to the default parser about the site.
Then people can look up or add items to it.
We can see how it grows.
If there's significant activity, then having automatic lookup with Default Parser is something we could add.

[ARYAN-9099]

@dteviot I don't think many users will actively go to a different site to contribute.
My idea was to have a small popup appear after someone uses default parser to pack a epub.

[gamebeaker]

@ARYAN-9099 just to clarify.
The user can submit the css selectors without creating an account? (on github etc.)

[ARYAN-9099]

@gamebeaker yes I was thinking of such. We can make the prompt to submit appear only after the book sucessfully packs. No login needed, directly submit from the tab that extension makes.

[dteviot]

@ARYAN-9099

I don't think many users will actively go to a different site to contribute

That depends. They're probably willing to go there to see if it already HAS the settings for the site they want.
And, once they know it exists, a few will contribute back (based on the experience I had with decrypting ChrysanthiumGarden).

We can make the prompt to submit

Submit to where? As Gamebeaker said, I don't think we can post to GitHub without credentials. If it's a 3rd party site, who is going to manage it?

And if users supply credentials to WebToEpub, and we're posting things back, that has Privacy (and Security) implications.

[gamebeaker]

I can provide hosting for a few php, html, js scripts that act as api endpoints and a postgresql database. (i already have the webspace and domain so no extra cost to me)
If someone wants to test it write me an e-mail: Play4freeNew@gmx.de
And i will create credentials for this.
@dteviot only if you are ok with me hosting this.
Another point is on where to host this server code. An extra repository or just a new folder "default_css_server" (I would like to have this code also open source)
I have to read the tos again if we need to get special consent from the user should this be implemented.

[ARYAN-9099]

@gamebeaker i can try making a prototype of it.

Another option for hosting- I don't think it will need much compute so free tier of any serverless and online db should be enough.

What are your thoughts on this? I'm fine either way.

I will go ahead if I get your and @dteviot consent.

I'm also in support of keeping this open source.

As for consent I don't remember agreeing for anything when I started using sponserblock for yt( I can be forgetting tho)

[dteviot]

@gamebeaker

@dteviot only if you are ok with me hosting this.

My only concern is "what if you're hit by a bus?" i.e. What's the plan if we can no longer use the site for some reason? But we can deal with that if/when it happens. Worst case, we remove the feature.

(Now you're going to ask me: what's the plan if I get hit by a bus? I think the answer to that is I should give you commit permissions to Chrome and Firefox stores. I need to look into that.)

a new folder "default_css_server"

Probably best. Then can keep the two together.

@ARYAN-9099

I will go ahead if I get your and @dteviot consent.

You have my consent. You and gamebeaker can decide between you on the details.

[ARYAN-9099]

@gamebeaker what are your thoughts on using cloudflare worker + supabase?