Templates: various questions not worth their own thread

[ownedbycats]

It worked

[ownedbycats]

Originally Posted by ownedbycats

Also, are there any other improvements that can be made to this? I notice there's a slight delay whenever the column updates. I've thought of taking out the #currentlyreading check and instead just having it run its checks directly, but I'm not sure if that would help or if there's a better way.

An update on this: #currentlyreading was originally a bit sloppy coded. (I'd actually improved it a while ago but forgot to actually copy the new template in, oops) I just updated it and it seems to have helped with the other column too.

[chaley]

Originally Posted by ownedbycats

I might've asked this before, but does it do all the checks, or only continue the check when it passes the other ones?

Logical conditional expressions do "shortcutting", which means that they stop as soon as they know the answer. In practice, for "a && b" it will stop if a is false, returning ''. For "a || b" it will stop if a is true, returning '1'.

If the latter, since #currentlyreading is another composite (psuedobool, it returns 'true' if it passes certain checks), would putting it last be better practice for performance?

Probably. It depends on the complexity of the other parts of the &&. In this template, almost certainly.

In general, for a sequence of && clauses put them in the order you expect them to most often fail. For a sequence of || clauses put them in the order you expect them to most often succeed.

Also, are there any other improvements that can be made to this?

Given that the variable 'a' is used in only one place, it is best to remove the assignment and change the line

strcat(format_number(subtract(a, 1), '{0:,d}'), '/',ccount)

to be

strcat(format_number(re(input, '.*\/file(\d+).*', '\1') - 1, '{0:,d}'), '/',ccount)

The expression "a - 1" is faster than the expression "subtract(a, 1)" because it avoids the function call.

[ownedbycats]

That worked. I swapped the order of the bookmark check and the status check. Also I didn't realize it was possible to avoid calling subtract.

[ownedbycats]

For columns specifically, does commented out template code have any affect on performance? e.g. before the virtual_libraries() thing got fixed, would commenting it out just have it skip entirely or calculate the result and then toss it out?

[chaley]

Originally Posted by ownedbycats

For columns specifically, does commented out template code have any affect on performance?

No, it doesn't.

[ownedbycats]

Question:

I have this template in an Action Chains single-field edit for semi-automated updating of my Kobo collection/shelf column:

program:
	g = field('#genres');
	k = field('#kobocoll');
	vl = field('#vls');

	v = first_non_empty(
		list_contains(vl, ',', '^Loans$', 'Loans', ''),
		list_contains(g, ',', '^Fanfiction$', 'Fanfiction', ''),
		list_contains(g, ',', '^Horses$', 'Horses', ''),
		list_contains(g, ',', '^Cozy Mystery$', 'Cozy Mysteries', ''),
		list_contains(g, ',', '^(Fantasy|Science Fiction)$', 'Fantasy & Sci-Fi', ''),
	);

	list_union(k, v, ',')

Performance isn't an issue, it's only run on demand. But since the "Loans" virtual library only criteria is that the book has an .OVERDRIVE format, would it be better practice to check for that rather than the VL?

[chaley]

Originally Posted by ownedbycats

Question:
...
Performance isn't an issue, it's only run on demand. But since the "Loans" virtual library only criteria is that the book has an .OVERDRIVE format, would it be better practice to check for that rather than the VL?

It depends on how you check for the format. If you use approximate_formats() then doing

'OVERDRIVE' inlist approximate_formats()

will be slightly faster unless you are reasonably sure the #vls column is already evaluated (e.g., it is displayed in a column). In any event the difference will be minor.

If you use any of the formats_???() functions then the time needed to touch the disk to check the physical files will swamp any template processing time.

[ownedbycats]

Question: for my composite chapter count column:



Is it possible, with a template, to display icon only if the first number is one less than the second?

1/2 - icon.png
2/2 - no icon
99/100 - icon.png
50/100 - no icon
50 - no icon

If it helps, here's the composite. It pulls the numbers from #chaptercount and #kobobookmark:

program:

input = $#kobobookmark;
status = $#fanficstatus;
ccount = $$#chaptercount;

	if 
		$#fanficcat
	then
		if		
				substr(input, 0, 10) == 'OEBPS/file'
			&& 	
				!status inlist 'Anthology,Oneshot'

		then		
			strcat(format_number(re(input, '.*\/file(\d+).*', '\1') - 1, '{0:,d}'), '/',ccount)
		else
			ccount
		fi

	elif
		ccount == 'None'
	then
		'Not Set'

	elif
		ccount >#1
	then
		ccount

	fi

Thank you.

[chaley]

Originally Posted by ownedbycats

Question: for my composite chapter count column:

Attachment 188857

Is it possible, with a template, to display icon only if the first number is one less than the second?

1/2 - icon.png
2/2 - no icon
99/100 - icon.png
50/100 - no icon
50 - no icon

If it helps, here's the composite. It pulls the numbers from #chaptercount and #kobobookmark:

Yes. Use an advanced rule that examines the values and generates an icon if needed. Something like this:

program:
	t = $#tool;
	if '/' in t then
		f = re(t, '(.*)/.*', '\1');
		s = re(t, '.*/(.*)', '\1');
		if f <# s then
			'icon.png'
		fi
	fi

[ownedbycats]

Thank you. That also seems to get the 50/100 example though. I'll see if math functions might help here.

[chaley]

Originally Posted by ownedbycats

Thank you. That also seems to get the 50/100 example though. I'll see if math functions might help here.

Sorry, I read "less than" not "one less than". The prototype template would be:

program:
	t = $#tool;
	if '/' in t then
		f = re(t, '(.*)/.*', '\1');
		s = re(t, '.*/(.*)', '\1');
		if (s - f) ==# 1 then
			'icon.png'
		fi
	fi

[ownedbycats]

I actually managed to come up with pretty much the same thing before seeing your post. Thank you.

[kevn57]

This is a repost as I posted in the wrong thread

I'm a complete nube in calibre template language.

I'm trying to take the output of the plugin Pagecount which is the number of words in a book and dividing that by 320 which is my words per minute reading rate and output the total number of minutes to read the book into my custom integer col "#minutes"

I don't know how to assign the product of divide(vals, 320); to my custom column #minutes.

program:
col = '#pagecount';
vals = from_selection(col);
'#minutes' = divide(vals, 320);

But I get this result EXCEPTION: Formatter: Expected end of program, found '=' near '#minutes' on line 4

[capink]

from_selection() is used in Action Chains Formula action to perform mathematical operations that needs input from more than one book.

Templates for composite columns only work for mathematical operations that pulls the data from one book at at time.

If I understand what you want to do correctly, your template should look like this:

program:
    $$#pagecount / 320