Templates: various questions not worth their own thread

[davidfor]

Originally Posted by imNOTdwight

I thought about doing that. But it won't work if I want to have two formats of the same book on the device. It's the same “two files one path problem”.

I might just be better of using kepubs exclusively and pdfs where appropriate.

If you are talking about your Kobo device, there is no problem having multiple formats of the same book. I do this all the time. The device uses the full path and file name including the extension to differentiate the books. Just use the "Send specific format to" option on the "Send to device" button.

[ownedbycats]

Originally Posted by chaley

If using this method, which BTW is a good one, then for performance it would be better to move the s and f calculations into the inner if. Something like this (I didn't try to run/compile it):

program:

	if 
		$#chapters != 'Not Set'
		then 
			f = re($#chapters, '(.*)/.*', '\1');
			s = re($#chapters, '.*/(.*)', '\1');
			if and(
				$$#fanficcat,
				check_yes_no('#onkobo', '', 1, 1),
				$$#percentread ==#100,
				(s - f) ==# 1
			) 
			then '98'
			else $$#percentread
			fi
# IS SOMETHING SUPPOSED TO GO HERE?
# In the original template this would return $$#percentread
# The new code returns the empty string
		fi

Yeah, I realized partway through testing that I needed a second else $$#percentread in the commented bit or it would return blank for some books. This is why you should include a couple of control books when testing templates.

thanks

[imNOTdwight]

Originally Posted by davidfor

If you are talking about your Kobo device, there is no problem having multiple formats of the same book. I do this all the time. The device uses the full path and file name including the extension to differentiate the books. Just use the "Send specific format to" option on the "Send to device" button.

But that's the problem. My wife would see duplicated books in Nickel, and I would see them as well in KOReader.

The ideia would be to have "One Device, Two Separate Libraries". One library with "her" kepubs, and one with "his" epubs.

It would not work since there's the small problem of Calibre escaping the "." used for hidden folders.

[davidfor]

Originally Posted by imNOTdwight

But that's the problem. My wife would see duplicated books in Nickel, and I would see them as well in KOReader.

The ideia would be to have "One Device, Two Separate Libraries". One library with "her" kepubs, and one with "his" epubs.

You are sharing devices?

It would not work since there's the small problem of Calibre escaping the "." used for hidden folders.

There is no reason to use a directory with a leading dot in it. You can tell nickel to ignore any directory but adding it to the "ExcludeSyncFolders" in the config file. But, if you have any doubts about how to do this, I would recommend checking the exact format in the Kobo forum. I do not use it at all, so I would need to check the details before doing it. I would suggest posting what you have and describing what you want to do. Someone will show what you need to do.

[imNOTdwight]

Originally Posted by davidfor

You are sharing devices?

Yes, we have a Libra and a Clara we sometimes exchange.

[QUOTE=davidfor;4182824]
There is no reason to use a directory with a leading dot in it. You can tell nickel to ignore any directory but adding it to the "ExcludeSyncFolders" in the config file. But, if you have any doubts about how to do this, I would recommend checking the exact format in the Kobo forum. I do not use it at all, so I would need to check the details before doing it. I would suggest posting what you have and describing what you want to do. Someone will show what you need to do.[/QUOTE]

Forgot about that. So that's one problem solved. I tried exporting the books via the exporting specific format to a single folder, but then I have to always delete the opf and jpg companion files.

I remembered plugboards would let me change things according to format, but that would mess up the metadata.

[davidfor]

Originally Posted by imNOTdwight

Forgot about that. So that's one problem solved. I tried exporting the books via the exporting specific format to a single folder, but then I have to always delete the opf and jpg companion files.

You can turn off creating those files in "Saving books to disk" preferences.

I remembered plugboards would let me change things according to format, but that would mess up the metadata.

Which is the point of the metadata plugboards, to fiddle with the metadata when sending them outside of the calibre library. I add page and word counts and some other stuff to the comments so I can see them in the synopsis on the device. Others play with other fields to get some effect they want.

But, it won't help what you want to do as it the metadata used for the save template is the original version, not the post plugboard version.

[imNOTdwight]

Originally Posted by davidfor

You can turn off creating those files in "Saving books to disk" preferences.

(…)

But, it won't help what you want to do as it the metadata used for the save template is the original version, not the post plugboard version.

Thanks for the help anyways. For now manual "Saving to disk" seems to be the easiest way to achieve what I'm trying to do.

[ownedbycats]

Mainly out of curiosity: Suppose you have a column with a single unmodified function, e.g. virtual_libraries(). Would there be any performance impacts involved in calling the #virtual_library column rather than the function directly?

(I'm also thinking -- if for some reason I were to change the column in the future, calling #virtual_libraries would avoid needing to change the function on all the templates calling it.)

Also, I can't quite remember: In search, it's possible to do inclusive ands/ors with parentheses, e.g.:
#columnA:foobar AND (#columnB:foo OR #columnC:bar)

Would the best way to do this would to with a "IF foobar THEN foo||bar" type thing, or can the checks be done in the first IF?

Thanks and merry Christmas (or whatever holiday you celebrate).

[chaley]

Originally Posted by ownedbycats

Mainly out of curiosity: Suppose you have a column with a single unmodified function, e.g. virtual_libraries(). Would there be any performance impacts involved in calling the #virtual_library column rather than the function directly?

(I'm also thinking -- if for some reason I were to change the column in the future, calling #virtual_libraries would avoid needing to change the function on all the templates calling it.)

This is hard to answer because it depends on context. If the column has been evaluated (displayed, sorted, something) then using the column will be faster because the result is cached. If the column has not been evaluated then calling the function will be faster because it avoids the overhead of handing the column. If you are using the template for search or sort then calling the function should be faster. Personally I would use the function because it is clearer what the template is doing, (as you say) it avoids unintended effects of changes, and the performance difference is zero to small.

Also, I can't quite remember: In search, it's possible to do inclusive ands/ors with parentheses, e.g.:
#columnA:foobar AND (#columnB:foo OR #columnC:bar)

Would the best way to do this would to with a "IF foobar THEN foo||bar" type thing, or can the checks be done in the first IF?

Yes, you can do it in the first if in almost the same way as the search. For example, assuming your search is intended to be 'contains' (as written) you would write the if like this:

if 'foobar' in $#columnA && ('foo' in $#columnB || 'bar' in $#columnC) then

Thanks and merry Christmas (or whatever holiday you celebrate).

Thanks. Same wishes to you.

[ownedbycats]

program:
m = is_marked();

first_non_empty
   (
    contains(m, "invalid_author_sort|invalid_title_sort", 'auto_author_sort.png', ''),
    contains(m, "epub_calibre_bookmarks", 'snippets.png', ''),
    if 'fff' in m && check_yes_no($#onkobo, 0, 0, 1) then 'sync.png' fi,
    contains(m, "fff", 'download-metadata.png', ''),
    contains(m, "search_results", 'search.png', ''),
    if m then 'marked.png' fi
   )

I'm likely going about this the wrong way - line 8 I want to check the book is marked with fff and #onkobo is set to yes. But instead I'm getting download_metadata.png on those ones.

Any idea?

[chaley]

Originally Posted by ownedbycats

program:
m = is_marked();

first_non_empty
   (
    contains(m, "invalid_author_sort|invalid_title_sort", 'auto_author_sort.png', ''),
    contains(m, "epub_calibre_bookmarks", 'snippets.png', ''),
    if 'fff' in m && check_yes_no($#onkobo, 0, 0, 1) then 'sync.png' fi,
    contains(m, "fff", 'download-metadata.png', ''),
    contains(m, "search_results", 'search.png', ''),
    if m then 'marked.png' fi
   )

I'm likely going about this the wrong way - line 8 I want to check the book is marked with fff and #onkobo is set to yes. But instead I'm getting download_metadata.png on those ones.

Any idea?

Your problem is that check_yes_no() takes a field name not a field value. The line should be

    if 'fff' in m && check_yes_no('#onkobo', 0, 0, 1) then 'sync.png' fi,

Also, it might be that you want list_contains() (also named in_list()) instead of contains(). This is true if there is any chance that 'm' can contain items that contain queried terms. For example, contains() will return True if the query term is 'fff' and the value is fff_good. Contains() is fine if no submatches are possible.

EDIT: You can also use str_in_list() to do exact matching if you don't need regular expressions. For example, this

contains(m, "invalid_author_sort|invalid_title_sort", 'auto_author_sort.png', ''),

can be written as

str_in_list(m, ',', 'invalid_author_sort', 'auto_author_sort.png', 'invalid_title_sort', 'auto_author_sort.png', ''),

Also like this:

if '^(invalid_author_sort|invalid_title_sort)$' inlist m then 'auto_author_sort.png' fi,

[ownedbycats]

I was trying to change my 'genres' column like so:

program:
# Current virtual library
	vl = current_virtual_library_name();

## Main three 'categories' of books. fanfic is taglike, manual enumerated
	tags = $tags;
	fanfic = $#fanficcat;
	manual = $#manualtype;

## Stripping sub-tags of fanfic and manual to use their specific columns instead
	stripped_tags = list_re_group(tags, ',', '.',
			'^(Fanfiction|Documentation & Manuals)($|\..*$)',
			'{$}', '');

## Splitting tag hierarchy
	split_tags = re(stripped_tags, '\.', ',');

## Removing a few unwanted tags
cleaned_tags = list_difference(
			split_tags,
			'@cleanup, Cultures & Regions, School, Social Issues',
			',');

## Trying to concenate the three book categories for shared VLs. Otherwise just use the cleaned tags.
	if
		'(Read|Unread|On Device|Cleanup)' in vl 
	then 
		list_union(cleaned_tags, fanfic, manual, ',') 
	else
		cleaned_tags
	fi

1. Line 27's list_union isn't working as long as I have 'manual' in there, which I'm guessing is due to it being an enumerated column instead of taglike. How do I use it in the list_union?

2. How do I add 'All Books' to the three VLs I have listed?

3. Is it expected for the list_union's combined entries to display in the opposite order? (e.g., "Mass Effect Trilogy, Fanfiction" even though the first is from fanfic, which is second in list).

4. Sometimes if I switch between multiple VLs/views too quickly, the column 'sticks' and displays the wrong entries until I press F2 and re-save the template. Any tips to unstick it?

If any help, this mostly happens when I switch between Fanfiction (where the view manager doesn't even have the column displayed) and Read/Unread.

Worst case, I'll just used combined entries for all VLs - the minor nuisance of having fanfic categories shown twice in book details I can deal with.

[chaley]

Originally Posted by ownedbycats

1. Line 27's list_union isn't working as long as I have 'manual' in there, which I'm guessing is due to it being an enumerated column instead of taglike. How do I use it in the list_union?

The error message I get is

EXCEPTION:  Formatter: Incorrect number of arguments for function list_union near ')' on line 27

This error message is correct. The list_union() function takes 3 arguments: the first list, the second list, and the separator. You are passing 4 arguments, 2 of which are lists. You need something like

list_union(list_union(cleaned_tags, fanfic, ','), manual, ',')

2. How do I add 'All Books' to the vl check? "All Books" didn't work.

Thanks

The 'All Books' virtual library isn't a VL, it is the null VL. Its name is the empty string: ''. You would use something like

	if
		vl == '' || '(Read|Unread|On Device)' in vl 
	then

This should work as well:

	if
		'^(|Read|Unread|On Device)$' in vl 
	then

In fact it is better because 'read' is a substring of 'unread' so without the anchors 'read' will match both 'read' and 'unread', which in some cases might not be what you want.

[chaley]

Question 3: There is no guarantee of the order in calibre lists. You can sort the lists if you want that.

Question 4: I don't understand what you are asking. What column sticks and what template?

[ownedbycats]

The #genres columns - main thing is if I switch between, say Fanfiction and Read/Unread too quickly, they start displaying the wrong entries for that VL (combined or only cleaned_tags) -- I'm guessing something to do with the caching, because re-saving the template would recalculate them all. F5 doesn't fix this, nor does re-searching.

If it matters, I'm using the action chain that automatically switches view manager views upon clicking on a VL tab.