Change in Series Index when saving books to disc

[w00dent0p]

Hi,

I've recently changed laptops, so copied my Calibre library from one machine to the other. Now when using 'Save to disc' I've noticed a difference in the handling of series_index, which is now causing duplicates in my exported library. For example ...

Previous behaviour:
...\Books\Sanderson, Brandon\Stormlight Archive, The\02 Words of Radiance - Brandon Sanderson.epub
...\Books\Sanderson, Brandon\Stormlight Archive, The\2.50 Edgedancer - Brandon Sanderson.epub

Current behaviour:
...\Books\Sanderson, Brandon\Stormlight Archive, The\02 Words of Radiance - Brandon Sanderson.epub
...\Books\Sanderson, Brandon\Stormlight Archive, The\2.5 Edgedancer - Brandon Sanderson.epub

So the issue only affects books with a non-integer series_index. I don't know if this is a change in Calibre behaviour, or a change in laptop (was Win10, now Win11).

My save template is unchanged:

{author_sort}/{series}/{series_index:0>2s} {title} - {authors}

Any thoughts?

Thanks,
Nigel

[chaley]

There was a change on 19/Feb to strip trailing zeros from series indices. The change appeared in calibre V5.38. I strongly doubt that it will be changed back.

You have three choices:

1. Revert back to calibre 5.37 or earlier. You will be stuck there.
2. Rename the existing books to get rid of the trailing zero.
3. Change to this template that formats the series_index differently depending on whether it is a float or an integer

4. {author_sort}/{series}/{series_index:'if floor($) != $ then format_number($, '0.2f') else $ fi'} {title} - {authors}

[w00dent0p]

Thanks, that explains it. I already went with choice 2.

Can you please clarify the template meaning a little more? Is there a way to achieve a consistent format across both float and integer? i.e. xx.x for both?

[chaley]

Originally Posted by w00dent0p

Thanks, that explains it. I already went with choice 2.

Can you please clarify the template meaning a little more? Is there a way to achieve a consistent format across both float and integer? i.e. xx.x for both?

Sure. If you *know* that you want xx.x and never xx.xx or xx then this does it.

{author_sort}/{series}/{series_index:format_number(0.1f)} {title} - {authors}

The original template checked to see if the index was a float by comparing its value to its 'floor' (integer part). If they are not the same then if used format_number() to convert it using 2 digits after the decimal, otherwise it used it unchanged. The new template (above) always converts it to use 1 digit after the decimal point regardless of whether it is integer or float.

Also and FWIW: the first template uses Template Program Mode. The one in this post uses Single Function Mode. My personal preference is General Program Mode, but I didn't provide an example in that mode.

[w00dent0p]

Thanks again.

That seems to result in xx.x, except for x.x when the index is less than 10.

Any way to always get xx.x ? i.e. to pad with the leading zero when needed.

[JSWolf]

I wish people would stop requesting these changes. This change and the search are two changes that would be best (IMHO) to go back to the way they were.

[w00dent0p]

I've removed {series} and {series_index} from my template.

Since this stuff is subject to change, I decided it's a bit too much of a faff to deal with the resulting duplicate files.

[chaley]

Originally Posted by w00dent0p

Thanks again.

That seems to result in xx.x, except for x.x when the index is less than 10.

Any way to always get xx.x ? i.e. to pad with the leading zero when needed.

Like this:

{author_sort}/{series}/{series_index:format_number(04.1f)} {title} - {authors}

Originally Posted by w00dent0p

I've removed {series} and {series_index} from my template.

Since this stuff is subject to change, I decided it's a bit too much of a faff to deal with the resulting duplicate files.

I don't see why it would change again, and in any event the template isolates you from such changes.

[w00dent0p]

Thanks very much !!