My resolution to this issue, was to create GREP styles for problem character combinations; keeping in mind, the issue is specifically in using all caps formatting; kerning around certain caps (A, X, and W are particularly bad offenders within this font) within a specific font family. Lowercase and caps and lowercase combinations seem to not have too many issues. What the GREPs look like is this;
Image may be NSFW.
Clik here to view.
So you can see, the GREP(s) is looking for Positive Lookahead and Positive Lookbehind occurrences. For example, the first GREP (?<=[a|A])[c|C|o|O] is looking for any occurrence of a lower or capital A that is followed by either lowercase or caps of C or O, and then applies the character style Title track to +50.
The second GREP down from that [a|A](?=[m|M|n|N|t|T]) is looking for any occurrence of a lower or cap A that is preceded by either lowercase or caps of M, N, or T and applies the character style Title track to +25.
Additionally, the character style applies an underscore so that I can see immediately where the kerning has been applied; each tracking value having its own assigned color. After I have checked all the kern pairs in the book, I can then turn them off at a single character style (since I am huge on properly related styles) to turn off all the underscores in the file. The one thing I might do, is not attempt to place all the tracking exceptions in a single GREP command (like I am doing here)...its getting to be a bit of management nightmare trying to remember what values have been applied where when just scrolling through the GREP list.
But, there you go! So far so good. No expensive typographer to build a custom kern table. And, no script to maintain and debug, and, the tracking values are built into the style so they can't be accidentally blown away by re-applying the paragraph style. And, if I organize my paragraph styles properly, I only have to build these overrides once (i.e., any paragraphs that are all caps and use a specific tracking value, all use the same GREP override combinations, so if I base them all off a master paragraph style, then any changes to the master ripple through the entire document...I love the Based On option!).
I just have to wonder if there's an upper limit of how many GREP styles can exist in single paragraph style. Assuming I do one of these GREPs for every potential character combination...what is that?...650 GREPs to define? Made my head swim just figuring out the Positive Lookahead and Lookbehind language.
