upper case to sentence case conversion

[cybmole]

is there a way, with regex, to tidy books where the first few words of each new chapter are in upper case.

I think that looks naff on an ereader, and would look better in sentence case, but coding that seems challenging

[ldolse]

You should apply a smallcaps style or pseudo smallcaps {text-size:80%} to those capitalized words - the reason they're showing up like that is they were smallcaps in the original book, but the sizing clues are often lost during OCR conversion.

You can find them via regex and wrap them with span tags using search and replace a regex something like this would be a good place to start:

([A-Z]{2,})

and replace with something like this:

<span style="smlcps">\1</span>

And then add the style 'smlcps' to you're css with appropriate sizing.

You could also do something like this:

([A-Z])([A-Z]{2,})

and replace with this:

\1<span style="lower">\2</span>

and use this in your css:

.lower {
   text-transform:lowercase;
}

However I'm pretty sure a lot of readers don't support text-transform (it may not even be part of the epub spec), while almost all will follow the font size changes.

[cybmole]

hmm - but I want to restrict changes to those start of chapter instances, not de-capitalise words elsewhere in the text, and not have to do each chapter manually - could be 50-100 chapters!.
maybe I have to look also for the trailing h2 tag leading into the chapter ?

 <h2 id="heading_id_3"><a class="calibre7" id="_Toc61566414">3</a></h2>

  <p class="MsoNormal3">NAOMI PHELPS DID MOST OF THE TALKING WHILE FRANCES SAT THERE AND shivered. Our secretary got her hot coffee and an afghan. Her hands

being fussy - I'd want to preserve the initial capital letter of the 1st word, also.

[ldolse]

That regex I provided looks for more than 2 characters in a row - so only all caps words are affected, not individual capital letters. The second regex/replacement leaves the first letter alone. You can tweak the regexes to include spaces as well.

[cybmole]

PS - thinking it through some more - there are probably too many special cases to solve, for this to be worthwhile.

e.g. a chapter beginning
THEN JED SAID TO FRED I KNOW would need to become
Then Jed said to Fred I know

even a built-in sentence case converter like in MSword , which would handle the capitalised I, would stumble on the proper names

[ldolse]

Which is another reason I recommended smallcaps. I'm sure if you look at the printed book you will find it's using smallcaps as well.

[cybmole]

ok thanks - another gap in my typesetting knowledge revealed!- I googled small caps - will experiment a little with conversions. I'd want something that works for epub & mobi - calibre viewer & Kindle.

[Perkin]

I know this isn't what you'll want to hear, but I think it's easier to do the replace by hand, because of Names or special cases, where capitals in the sentence need to be preserved.

After doing quite a few regex replaces, I found quite often I had to go back and alter some of them back to keep a few capitals.

[cybmole]

Originally Posted by Perkin

I know this isn't what you'll want to hear, but I think it's easier to do the replace by hand, because of Names or special cases, where capitals in the sentence need to be preserved.

After doing quite a few regex replaces, I found quite often I had to go back and alter some of them back to keep a few capitals.

inclined to agree, & to ask to a change case option to go on sigil wish list ( like the change case options in Word)