Search results contains no extract

[benoit74]

Search result do not contain any text extract, only the title. Why?

[benoit74]

Guess: I'm pretty sure this is because we have no real HTML content on these pages, we just pass the indexing content to the libzim, but the real page has no content.

Not sure we can do anything for this, since there is no real content to preview

[kelson42]

@benoit74 You are probably right, but I don't really see why it should not be fixable (maybe by extending scraperlib/pylibzim API).

[benoit74]

We should probably discuss this live, because I don't see which libzim API could be used to fix this.

Currently what we have for each page the user "see" is:

• 1 real "textual" content stored in a JSON file, used by the Vue.JS application
• 1 "fake" redirect-only HTML entry, for which we pass the "textual" content to the libzim so that both search index are properly populated ; this entry has no text inside its HTML since it is not supposed to be displayed, only to redirect to proper URL inside the Vue.JS application

And if I'm not mistaken, what has been discussed is that we should find a solution in libzim / ZIM format to not even need this "fake" redirect-only HTML entry, just store an entry without any content but a proper redirect URL (btw, I consider it could even be wiser to attach this directly to the JSON file, but to be discussed). But this is a different topic.

[benoit74]

Discussed live with @kelson42 @Popolechien, probably not something easy to change, out of LibreText 2024 project at least