AlexPGP's Corner

The OCR spit out for the document I'm working on right now is behaving very strangely in terms of font formatting.

For reasons best known to my computer's CPUs, in certain places, the OCR detected bold face font where there is clearly no bolding used in the original document. Editing the OCR output would normally not be a problem in Microsoft Word, because it's a simple enough matter to do a Select-All, then hit Ctrl-B a couple of times, where the first keychord takes a mixed string, such as

The rain in Spain falls mainly in the plain.

and makes it all bold, like this:

The rain in Spain falls mainly in the plain.

The second Ctrl-B just turns off the bold-face, and with that, the desired goal is reached, i.e.,

The rain in Spain falls mainly in the plain

.

However, what I'm running into is some kind of additional weirdness. Starting with that same selection of mixed bold and non-bold characters:

The rain in Spain falls mainly in the plain.

pressing Ctrl-B the first time results in the following:

The rain in Spain falls mainly in the plain.

In effect, it would appear that the bold property is being toggled for each individual character. Pressing Ctrl-B again gets me the original bold/not-bold selection (in effect, again toggling bold at a character, and not selection level). Apparently, the only way to make the whole selection not bold is to highlight each individual string of bold characters and do whatever it takes (e.g., press Ctrl-B) to unbold it.

This is very strange. Anyone else ever experience behavior like this? (More to the point, has anyone figured out a way around it?)

Cheers...