Converting Capitals -> Italics

[Rand Brittain]

I'm looking at a Project Gutenberg text that's used capital letters to stand in for italics, and I'd like to change it to display those italics instead. Is there a way to use regular expressions to change all words in all-caps to lower-case and slap a <i></i> tag around them? (I guess I'd have to remove it from every use of "I", but that's a simple find-replace.)

[Jellby]

I wouldn't use a global search and replace, since words in all-caps can also be legitimate, or roman numbers, or small-caps...

[HarryT]

I'd suggest doing it manually, too.

You can find the capitals by doing a regex search for "[A-Z][A-Z]"; that will find any two consecutive capital letters. This kind of thing is a pretty standard part of converting PG texts. It's complicated by the fact that some of the words in capitals should probably remain capitals; the only way you'll find that is when you carry out the next stage of your eBook conversion - a thorough proof-read against a page scan or printed edition.

[Fabe]

Originally Posted by HarryT

You can find the capitals by doing a regex search for "[A-Z][A-Z]"; that will find any two consecutive capital letters.

Can this kind of search be carried out in Sigil?

[HarryT]

I'm pretty sure that Sigil supports regex searches, yes. Perhaps someone can confirm?

[Valloric]

Originally Posted by Fabe

Can this kind of search be carried out in Sigil?

Originally Posted by HarryT

I'm pretty sure that Sigil supports regex searches, yes. Perhaps someone can confirm?

Yes, in both wildcard and regex modes.

[Pablo]

Sigil supports regex searches. In "book view" the search is restricted to the current html file, in "code view" you can search in all html files, but in this case it would be useless, as html tags are included in the search.
In short, in this case you have to repeat the search for each html file in your ePub.

[Valloric]

Originally Posted by Pablo

Sigil supports regex searches. In "book view" the search is restricted to the current html file, in "code view" you can search in all html files, but in this case it would be useless, as html tags are included in the search.
In short, in this case you have to repeat the search for each html file in your ePub.

All html tags are lowercased in Sigil, so it would still work well in Code View.

[Pablo]

Originally Posted by Valloric

All html tags are lowercased in Sigil, so it would still work well in Code View.

True, only false detections would be in lines 2 and 3 at the beginning of each file (DOCTYPE, etc), and book title/author if in capitals.....

[Carl314]

I'm not a regex guru, but I think this search/replace should at least insert the italics tag around capital letters (I don't know how to convert upper to lower case, however - perhaps set up a css lower case class for your italics tag if there's no other way?)

Search (case sensitive) for: ([ >“‘])([^a-z]+)([ <”’])
Replace: \1<i>\2</i>\3.

It will search for a block of text which contains at least two characters, and no lower case (but punctuation and numbers are okay) book-ended by either spaces, open/close tag signs, or quotation marks, an then insert the <i></i> around the text.

If you know you're only searching for caps, with no numbers or punctuation, then replace the text search term to [A-Z]+

Hope this helps.

Cheers, Carl.

[HarryT]

An automated search and replace is very dangerous for something like this. There are probably capitals that you DON'T want to turn to italics (eg things like "AM" or "PM" in a time). I really would recommend using a search to find the capitals, but deciding on a case by case basis what to do with each one.

[Fabe]

Originally Posted by HarryT

An automated search and replace is very dangerous for something like this. There are probably capitals that you DON'T want to turn to italics (eg things like "AM" or "PM" in a time). I really would recommend using a search to find the capitals, but deciding on a case by case basis what to do with each one.

I agree with the case-by-case notion. I have used Find Next & Replace to search about 100 instances in a file, and while I would have preferred it all happen correctly and automatically, the peace of mind it gave me was worth it.

I was not familiar with regex searches until this thread. I followed the ink from Sigil's documentation page to Regular Expression.Info where I discovered a new world of Find & Replace syntax.

As I develop searches for my own work, I will keep a file of used regex expressions. This should speed things along. Thanks all. - Fabe