In a blog post a few days back, I was confounded by the problem of converting ALL CAPS into lower case letters. Using a Javascript that I licensed from a Danish programmer, I was able to make the case change, but the GREP search was also converting articles like I and A into lower case, and that was creating a new (but easily soluble) problem.
One of my former students, Jeff Bauer, sent me a fix that he found on a programmer’s web site. It solves a related problem using the same Javascript: not changing Roman Numerals into lower-case letters. Curiously, that solution also solves the articles problem I was having.
Thank you Jeff for submitting the solution to my problem!
Let’s take a look at how it’s done. I had wrestled with this problem, and had failed to find a solution because I do not know how to use “negative look-aheads” in GREP. I tried, believe me, but I failed to solve the problem.
The solution Jeff sent is complex but elegant: \b(?![LXIVCDM]+\b)([A-Z]{2,})\b
Huh? This uses both a negative look-ahead, and a little-documented GREP feature that says “this has to be at least two characters in length. It’s the part in series braces {}.
Here is a diagram of how it works:
My test document is a page of InDesign-generated pseudo-Latin into which I inserted a handful of all-caps examples, as well as a few cap-I and cap-A situations. I also threw-in Roman numerals to test the advertised function of the string.
Above: the text before I ran the search for all-caps.
The search string works fine, except for one exception where it fails with a string of all-caps letters preceded by a hyphen.
Above is the same text after I ran the script with the new GREP string. Notice the single word that is still capitalized, a compound word.
I guess that’s my next challenge.
_____
I’m writing a new book about printing processes and prepress. Click on the link below to give me your e-mail address, and I will let you know when the book is published (target date is March, 2012).
