Pager widget - inconsistency between pageSet and p.page

[iustin]

Hello,

I think the docs are incorrect regarding p.page (the widget). It says that both it, and the pageSet event, use a one-base index. While that is true for the pageSet, it looks like p.page is zero-based. You can see this inconsistency here:

p.page = ( parseInt( num, 10 ) || 1 ) - 1;

(num being the argument to pageSet).

I think it'd be best that zero-based is always used, but if that's not possible (backwards compat), then at least the docs for p.page should be updated.

[Mottie]

Hi @iustin!

Yes, the p.page variable is stored as a zero-based index. All internal functions use a zero-based index. The pager methods (pageAndSize, pageSet and pageSize) all use a one-based index as this was expected to be used along with the user interface – which shows a one-based index to the user.

As a developer, you can change the table.config.pager.page and table.config.pager.size values directly using a zero-based index, then call the $.tablesorter.pager.moveToPage(config, config.pager) function.

Either way, I think it's too late to change this behavior now since it would not be backwards compatible, and I don't plan on creating a major revision with tablesorter. I will make this behavior consistent once I get around to working on Abelt again.

[iustin]

Sounds good to not change the code - but what do you think about fixing the docs? I couldn't find exactly where in the repository are the doc pages to send a pull request, sorry. Basically just fixing the wording on p.page (which incorrectly states one-based indexing).

[Mottie]

Oops! Ok, I'll get that fixed.

[iustin]

Ah, so the main page is simply index.html, wondering why I couldn't find that wording (I blame github search). Thanks for the quick fix!