Convert podcast description to text

[brian-doherty]

The Subsonic players I've used for podcasts (DSub, Airsonic) don't handle HTML in the podcast description well and so it doesn't render properly. Purpose instead to convert to text.

[sentriz]

i think something like bluemonday could be a good fit for that

[archekb]

I was add cleaner from html tags to Airsonic-refix (tamland/airsonic-refix#110) I don't know when owner will merge this commit, but you can take my version with latest commits https://github.com/archekb/airsonic-refix

[brian-doherty]

Don't get me wrong -- it would be way cooler if the client would either strip or (preferably) render the HTML. But other clients don't parse it either. Does the parser you suggest add in \n\n for the paragraph breaks? And should it?

…

On Mon, Aug 14, 2023, 6:00 AM archekb ***@***.***> wrote: I was add cleaner from html tags to Airsonic-refix ( tamland/airsonic-refix#110 <tamland/airsonic-refix#110>) I don't know when owner will merge this commit, but you can take my version with latest commits https://github.com/archekb/airsonic-refix — Reply to this email directly, view it on GitHub <#349 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASFD42NW5NPYZOKM3VM2RBLXVIAODANCNFSM6AAAAAA3O4MEGA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[archekb]

it would be way cooler if the client would either strip or (preferably) render the HTML

No problem, I'll add this feature when will make Airsonic-refix client settings. It can be turned on or off by your chose on the settings.

Preview current:

with html render

With html render it's a little bit more pretty, then we should stay tags in the description on the server.

[brian-doherty]

Yeah, the issue is that other clients don't render the HTML either. I'm asking whether we should just strip at the server because of that. For example, DSub.

[archekb]

I suppose it's not server issue (because server get description from podcast and sent it to client as raw), this is client issue, maybe you need to open issue for client?

Because after your comment I'll add support html tags to Airsonic-refix, because this is a little more pretty, for me. ;)

[brian-doherty]

So I checked and while the Subsonic API definition is ambiguous on this, the actual Subsonic implementation converts HTML to text on the way into the database. The magic question is -- do we keep the HTML and give clients the opportunity to go above and beyond, or do we match so that clients used to the way that Subsonic does things render correctly?

[sentriz]

i think keep it simple for now - if we sanitise the html on the way out of the db and into the rest view, we keep all clients happy, while still keeping the data intact if any clients decide to support it later

[brian-doherty]

So, bluemonday does OK but doesn't add carriage returns for paragraph breaks. See attached photo. I wonder if https://github.com/k3a/html2text or https://github.com/jaytaylor/html2text would be better.

[brian-doherty]

I used jaytaylor/html2text because it was more popular and I think it's way better. I'll send that in as a PR for you to see what you think.

[archekb]

Maybe we can makes it optional? Description without tags looks boring. (after adding rendering xD) Something like GONIC_PODCAST_RAW_DESCRIPTION

PS Maybe we will Sanitaize and podcast description, not only Episodes description?

[betapictoris]

Maybe we can makes it optional? Description without tags looks boring. (after adding rendering xD) Something like GONIC_PODCAST_RAW_DESCRIPTION

PS Maybe we will Sanitaize and podcast description, not only Episodes description?

What about adding it as an option depending on the client parameter, similar to transcoding profiles? That way it'd be possible to use it for clients that support it, all, or none.

[brian-doherty]

Making it a config parameter AND adding sanitization to the podcast description are OK by me if they're OK by Senan. I don't think we should add to the Subsonic API but if we did it would be a parameter to getPodcasts and getNewestPodcasts.

…

On Mon, Aug 14, 2023 at 6:37 PM Daniel ***@***.***> wrote: Maybe we can makes it optional? Description without tags looks boring. (after adding rendering xD) Something like GONIC_PODCAST_RAW_DESCRIPTION PS Maybe we will Sanitaize and podcast description, not only Episodes description? What about adding it as an option depending on the client parameter, similar to transcoding profiles? That way it'd be possible to use it for clients that support it, all, or none. — Reply to this email directly, view it on GitHub <#349 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASFD42JCTK43PJTRUTE55ADXVKZEHANCNFSM6AAAAAA3O4MEGA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[sentriz]

oh, i just noticed we can close this one. we can chat about the option over on #352 i suppose. thanks!

[betapictoris]

Making it a config parameter AND adding sanitization to the podcast description are OK by me if they're OK by Senan. I don't think we should add to the Subsonic API but if we did it would be a parameter to getPodcasts and getNewestPodcasts.

I was thinking more of an option in the admin page like transcoding profiles but for podcast description sanitization, kinda like this:

I don't think doing it this way would require moving away from the Subsonic API....