Some page titles are not retrieved properly

[ArthurHoaro]

The titles aren't retrieved properly if they're not present using the <title> HTML tag (e.g. populated in JS, or any other exotic way).

We can extend our ways to retrieve titles. A few examples:

• document.title
• OpenGraph tag og:title
• etc.

---

Original issue:

With the new design (I suppose), the <title> tag only contains YouTube, the actual title is populated using JS:

document.title = "Denis Brogniart :  \"AH\" - YouTube";

Usually, I would avoid adding specific code for specific websites, however YouTube is a common website to shaare, so maybe we should add something to retrieve the actual title.

[virtualtam]

YouTube provides a REST API, but it requires having a Google account to create a Developer API token for authentication.

Corresponding PHP library: php-youtube-api

[nodiscc]

With the new design (I suppose), the <title> tag only contains YouTube, the actual title is populated using JS:

Honestly this sucks. Anyway... Shaarli gets the title correctly when either

• sharing a youtube page from Firefox Share
• sharing a youtube page from the bookmarklet

The only case where title fetching fails is when using +Shaare -> Enter URL. I don't know how widely used each workflow is. In the mean time, why not rename this issue to something generic like Some page titles are not retrieved properly, document it in https://github.com/shaarli/Shaarli/blob/master/doc/md/Troubleshooting.md, and see if/how the problem evolves over time?

I'd like to avoid site-specific code, even more so if we have to import whole libraries for a single website.

[ArthurHoaro]

I wasn't talking about using their API, of course, because it requires an account, authentication token management, etc.

Anyway, I agree with your suggestion @nodiscc.