[eschwartz]

Originally Posted by chaley

TPM is Template Program Mode.

A template that produces what you want for the example you provided is

{author_sort:'re($, '([, ]+)(\w)\w*', '\1\2.')'}

I don't know it it will produce the right answer in all cases.

It truncates everything after the first word. Including the second author's first name (since it includes "& LN,[...]").

This should work, though:

program:

re(
	field('author_sort'),
	'(\w)\w+\b(?!,)',
	'\1.'
)

At some point I thought I'd try list_re to fix your example.

But I discovered this:

program:

list_re(
	field('author_sort'),
	'&',
	'(.*)',
	'\1'
)

gave me "LN1&FN1&LN2&FN2" -- isn't it supposed to search for "(.*)" and replace with "\1" on each list element where "&" is the separator? I cannot imagine what I might be doing wrong. But this looks completely crazy.