How to correct odd, unwanted characters in an epub?

[Skydog]

I've browsed the Sigil forum and Conversion sub-forum looking for a solution without success. I have an epub loaded with the following characters and cannot figure out how to correct the entire book. In conversions I have tried, individually: 'Unsmarten punctuation,' 'Transliterate unicode characters to ASCII,' and also selected utf-8 in 'input character encoding.' Nothing seems to work. Here are the unwanted characters:

â (throughout the text, possibly replacing single quotation marks)
 (between chapter breaks)

and then this - which I think is meant to be the word protégé with accents aigu:

protégé

I cannot do a simple search and replace in Sigil since the characters may represent more than one character and I have no idea what they might be.

Does anyone have a solution or suggestion how to correct this? Is there a conversion setting I am missing?

[JSWolf]

Those are not unwanted characters. Those are actually the characters you do want. They are displaying incorrectly because of the wrong encoding being used.

In the header right after the title (if needed), I put the following...

<title>Cat &amp; Mouse</title>
<meta content="http://www.w3.org/1999/xhtml; charset=utf-8" http-equiv="Content-Type"/>

I cannot speak for the current Sigil, but in a pre-0.7.x version, when it got the code page wrong, it could convert things like your quotes to those sorts of characters and if you saved, your screwed.

[jackie_w]

Using the Modify Epub plugin, with the option Encode HTML in UTF-8 checked usually works for me.

[JSWolf]

Originally Posted by jackie_w

Using the Modify Epub plugin, with the option Encode HTML in UTF-8 checked usually works for me.

In most cases, it's not an encoding issue of the actual file. It's an encoding issue in the header of the XML. That is why I have to add in the line I specified so the software reading the XML gets it right. ADE doesn't have an issue because uses UTF-8 regardless of what the header specifies.

[jackie_w]

Originally Posted by JSWolf

In most cases, it's not an encoding issue of the actual file. It's an encoding issue in the header of the XML. That is why I have to add in the line I specified so the software reading the XML gets it right. ADE doesn't have an issue because uses UTF-8 regardless of what the header specifies.

As far as I know, that's exactly what Modify Epub fixes, i.e the HTML headers.

[Skydog]

Originally Posted by jackie_w

Using the Modify Epub plugin, with the option Encode HTML in UTF-8 checked usually works for me.

Neither this method (Modify Epub plugin) nor converting using 'input character encoding' as utf-8 works for me.

I'll try what Jon suggested and hope it works.

Edit: I applied the edit to a chapter and there is no change.

[jackie_w]

@Skydog,

If you've converted it a few times it's possible the encoding is well and truly messed up by now. If all else fails, try going back to your clean original source before trying to correct.

[Skydog]

Originally Posted by jackie_w

@Skydog,

If you've converted it a few times it's possible the encoding is well and truly messed up by now. If all else fails, try going back to your clean original source before trying to correct.

I am working on a "new" copy. I simply save each one as a "mod 1," mod 2," etc., but I always start with a fresh copy of the original.

In any case, I still have the problem.

[Ripplinger]

When that's happened to me a few times, a simple epub-to-epub conversion in Calibre fixed all those odd symbols. I didn't check anything special, just my usual of "remove blank lines" and it worked every time for me.

[JSWolf]

Originally Posted by Ripplinger

When that's happened to me a few times, a simple epub-to-epub conversion in Calibre fixed all those odd symbols. I didn't check anything special, just my usual of "remove blank lines" and it worked every time for me.

ePub > ePub to fix the encoding is not a good idea. In fact, ePub > ePub is not a good idea if things can be fixed any other way.

[Skydog]

Third time is the charm! I once again copied the original and did nothing but apply the modify epub (utf-8 encoding), as jackie_w suggested. It worked!! Thank you, jackie_w. In this case, it accomplished the shortcut to Jon's suggestion.

@Ripplinger - I originally performed an epub-epub conversion as I mentioned above which did not work for some reason. The utf-8 encoding was indeed the issue -- I just wasn't initially able to get it to stick for some reason.

[Skydog]

Originally Posted by JSWolf

ePub > ePub to fix the encoding is not a good idea. In fact, ePub > ePub is not a good idea if things can be fixed any other way.

Is there a better way to accomplish widows: 0; orphans: 0; ?? I sure wish some talented person here would include it in the Modify Epub plugin.

[JSWolf]

Originally Posted by Skydog

Is there a better way to accomplish widows: 0; orphans: 0; ?? I sure wish some talented person here would include it in the Modify Epub plugin.

Use Tweak eBook (in Calibre unless you are already editing in Sigil) and you will then be able to edit the CSS to add in widows and orphans of 0 to the body style.

[Skydog]

Originally Posted by JSWolf

Use Tweak eBook (in Calibre unless you are already editing in Sigil) and you will then be able to edit the CSS to add in widows and orphans of 0 to the body style.

Thanks - I forgot about Tweak eBook.

[JSWolf]

Originally Posted by Skydog

Thanks - I forgot about Tweak eBook.

Happy to help.