Template: various questions not worth their own thread

[ownedbycats]

I'll try to use this thread for smaller questions to avoid spamming up this section even more.

Is there any semantic difference between

"foobar" in #column

and

#column in "foobar"

They seem to have the same results when I used them in a GPM template.

[chaley]

Originally Posted by ownedbycats

I'll try to use this thread for smaller questions to avoid spamming up this section even more.

Is there any semantic difference between

"foobar" in #column

and

#column in "foobar"

They seem to have the same results when I used them in a GPM template.

They are very different. Consider what happens if

#column == 'foobar is a strange word'

The first test would succeed. The second test would not.

If in your situation the two always both succeed then you probably should be using '==' instead of 'in'. EDIT: Or one of the list operators.

[ownedbycats]

In my case, I was doing this:

&& publisher == '(Archive of Our Own|FanFiction.net)'

"that doesn't work, i'll try an in instead," so:

&& publisher in '(Archive of Our Own|FanFiction.net)'

"that gets the right result but looks backwards, will it work the other way?"

&& '(Archive of Our Own|FanFiction.net)' in publisher

[chaley]

Originally Posted by ownedbycats

In my case, I was doing this:

&& publisher == '(Archive of Our Own|FanFiction.net)'

"that doesn't work, i'll try an in instead," so:

&& publisher in '(Archive of Our Own|FanFiction.net)'

"that gets the right result but looks strange, will it work the other way?"

&& '(Archive of Our Own|FanFiction.net)' in publisher

The "other way" is the right one. A regexp pattern works only if it is the left-hand argument. The "second way" might work but only by accident.

It seems like you are checking if some string matches some item in a list. In that case you probably should use list_contains() instead of 'in'. Something like:

list_contains($publisher, ',', '^(Archive of Our Own|FanFiction.net)$', 1, '')

I will look at whether a "in_list" operator makes sense given everything else that is going on inside the interpreter.

[ownedbycats]

I had assigned publisher = $publisher;

Also, what is the name used for these things used to avoid typing the same thing over and over again? e.g.

	status = $#fanficstatus;
	publisher = $publisher;
	ids = $identifiers;
	u = select(ids, 'url');
 	a = select(ids, 'ao3');
	f = select(ids, 'ffnet');

[chaley]

Originally Posted by ownedbycats

I had assigned publisher = $publisher;

I should have noticed that.

Also, what is the name used for these things used to avoid typing the same thing over and over again? e.g.

	status = $#fanficstatus;
	publisher = $publisher;
	ids = $identifiers;
	u = select(ids, 'url');
 	a = select(ids, 'ao3');
	f = select(ids, 'ffnet');

I don't think the "technique" of assigning to local variables to save typing (and usually processing time) has a name. If it does then I don't know it.

[ownedbycats]

"local variables" would probably work well for this.

[BetterRed]

Grace Hopper would have called it: WORKING STORAGE

BR

[ownedbycats]

Question:

strcat(a [, b]*) – can take any number of arguments. Returns a string formed by concatenating all the arguments.

What is B used for? In the templates I have that use it, it's just null ''s.

[PeterT]

Originally Posted by ownedbycats

Question:



What is B used for? In the templates I have that use it, it's just null ''s.

I think you code strcat(a), or strcat (a,b) or strcat (a,b,c) adfinitum. All the arguments will be concatenated together.

Sent from my Pixel 4a using Tapatalk

[chaley]

Originally Posted by ownedbycats

Question:
What is B used for? In the templates I have that use it, it's just null ''s.

I am probably giving you more than you want to know.

That notation describes the "form" of valid calls of the function. It uses 'placeholders' to distinguish between different values. The notation "[, something]" says that the ", something" is optional: zero or one occurrence of what is between the brackets. The notation "[, something]*" says that it is optional but you can have zero to "a lot" of occurrences. Often "somethings" with the same name are the same thing, while "somethings" with different names are not necessarily the same thing.

The notation is an approximation of a formal grammar. I tend to use approximations of EBNF grammars. In this formal system, things that are literal would be in quotes. In that case, what you referenced should be formally written

'strcat' '(' expression [ ',' expression ]* ')'

where the word 'expression' refers to another grammar element describing what it can be. I often use approximations because I am lazy.

This template language grammar is intended to be formal. I think it is correct but I haven't put it through a grammar verifier to be sure.

Taking all the above together, "strcat(a, [, b]*)" says that the literal 'strcat' is followed by:

• a literal '('
• a value, whatever that is. I used 'a' for 'expression'
• a sequence of zero or more
  • a literal comma
  • a value. I used 'b' to say it need not be the same as 'a'.
• a literal ');

[ownedbycats]

Just curious: is the term "composite column" used anywhere in the Calibre UI? In the column dropdown itself, it's "columns built from other columns."

[chaley]

Originally Posted by ownedbycats

Just curious: is the term "composite column" used anywhere in the Calibre UI? In the column dropdown itself, it's "columns built from other columns."

The term is used in a few places in tool tips. It is used frequently in the template language documentation. And FWIW, 'composite' is the internal datatype of these columns.

[ownedbycats]

Originally Posted by chaley

And FWIW, 'composite' is the internal datatype of these columns.

I saw that in the custom_columns table of the database.

[ownedbycats]

When using a template in Action Chains to set a boolean column, should I use "Yes/No" or "True/False"?