No images

[MalcolmRM]

I have a novel perfectly formatted in Sigil and saved as [novel].epub. It has images for Cover, title page, and The End; they show in Sigil and in Calibre when I import the .epub file there. Calibre makes a seemingly perfect conversion to mobi and the images show up in Calibre's ebook reader for both formats. However, ADE shows the epub text perfectly formatted BUT no images. Just blanks spaces of the appropriate size. And the Kindle ebook reader does the same: text perfect but correctly sized blanks for images. The technique I'm using has already converted two other novels perfectly.
I suspect this is because the images in this latest conversion are referenced thus:

src="/C:%5CUsers%5CMalcolm%5CAppData%5CLocal%5CTemp%5Cms ohtml1%5C01%5Cclip_image005.jpg" width="151" /></h1>

when the link to the images should be to the folder [novel_name]_files in the same folder as contains the original html file. But when I try to reload the ebook back into Sigil, all I get is a brief flash of Section 000001 of the epub that then defaults to the original [novel].html file.

QUESTION: How do I get to the html files for sections that contain images, which would be sections 1, 2, and 13? Is there a utility somewhere out there that lets one access and edit all the html files into which an ebook is divided?

[Toxaris]

The links in your epub to the images is not correct. You need to add the images in the epub and reference to the added images. Since you are already using Sigil, you can upload the images into the ePUB with Sigil. Then, adjust the image tags so they would look like:

<img src="../Images/clip_image005.jpg" />

[theducks]

I think the other question was how do I find where (embeded, not external) images were Used

Tools:reports: Select Images on the right: Double click the Name

Note: this only opens a SINGLE location (of all) that used the image.
You need to use that IMG tag for a standard Search for additional locations

[Notjohn]

Well, my images are referenced like this:

<img alt="" src="../Images/logo.gif" width="10%" />

Which seems very different from what you are showing.

To add an image, I right-click on the Images folder and choose Add Existing Files. Then I browse around my computer and click on the desired image. It doesn't have to be in the same folder as the epub.

Note that if you are actually publishing a Kindle edition, there is no need to convert to mobi format. (Indeed that may be counterproductive.) The KDP (Kindle Direct Publishing) platform has no problem with epubs.

[theducks]

Sigil has 2.5 tools that make placing images a breeze.

1)File: Add existing (Image) EMBEDS the image into the Book

2)The Image Placement (Black screen) tool to place the image into a books page


2.5) The cover tool. Tools:cover: Choose image)
Both Makes and FLAGS (in the opf) the cover File and the cover image file, all in one click

[MalcolmRM]

Thanks again to everyone. There are some fixes there that I'm sure will help. I'm having much less trouble now that I'm opening the Word2003 doc file in OpenOffice and saving in its version of filtered html. It's the only version that correctly converts those characters that require &#xxxx; entities; Word2003 leaves them as-is; Word14 seems to turn them into Wingdings.
My present problems still center round the fact that once an ebook is segmented into sections (Cover | title page | Contents links | prelims | Chapter 1 | ... etc) Sigil will only reload Section00001.html for later editing, not the entire ebook it itself created.

[Toxaris]

You could try my Word add-in to create the ePUB... It will create an ePUB from your Word document that does contain clean HTML. Images should be contained, but I recommend adding the original images, as images in Word are not that well contained.
It can be splitted already and a cover will be added if you want.

[Hitch]

Originally Posted by MalcolmRM

Thanks again to everyone. There are some fixes there that I'm sure will help. I'm having much less trouble now that I'm opening the Word2003 doc file in OpenOffice and saving in its version of filtered html. It's the only version that correctly converts those characters that require &#xxxx; entities; Word2003 leaves them as-is; Word14 seems to turn them into Wingdings.
My present problems still center round the fact that once an ebook is segmented into sections (Cover | title page | Contents links | prelims | Chapter 1 | ... etc) Sigil will only reload Section00001.html for later editing, not the entire ebook it itself created.

Sigil will open whatever sections (files) you've created. It doesn't open the whole thing like a word-processor with a single file. An ePUB is a collection of files, if it's done correctly. If you have "section0005.html," then you need to open it for editing by clicking on it, in Sigil. Then look at CodeView, if you want to work in the HTML (which is where you should be doing the work).

We export Word to HTML, and then import it to Sigil, all the time. Personally, I wouldn't waste the time with the additional OO step, but whatever floats your boat.) We tend to use regex to s&r and replace named entities (e.g., "&hellips" and the like) because that's our preference. You are likely encountering encoding issues--Word uses Win-1252, and Sigil assumes utf-8, if memory serves. You should open the HTML in a dedicated HTML editor, like NoteTab Pro, NotePad, whatever you have, and replace any questionable entities with named characters, to make your life easier.

You might also consider the Sigil Tutorial. That might save you some headache. If, for some unknown reason, you think you need to have everything open at once, for editing, you can use ePUBTweak, the utility, open all the files in File Manager, and then open them all in an HTML editor of your choice, but Sigil also gives you complete access to all your files. You could, I suppose, open all the internal files IN Sigil at once, and work from the tabs across the top, making Sigil quite wide, but I'm not sure that you and I are saying the same thing(s), or that I'm answering quite what you're asking.

Hitch

[theducks]

Sigil will open a book with 175 files (the most I have worked upon, not a hard limit)

Sigil AUTO opens the first file (typically the cover) in the OPF spine

You have to open the rest, be it select each one or have them open a part of a Search (next) interactive operation

PS I don't recommend having anywhere close to 175 tabs open at any one time. Frequent saves and using 'Close other tabs' during a interactive S & R

[MalcolmRM]

Wow! Lots to try there but - with a lot of help from my friends (that's you-all) I think I've got there. One (more) idiot question: I do all the editing and tidying up and standardising in an undivided epub, which gives me an undivided html to search/replace. When satisfied, I save it as [novel name]undivided.epub, around 1,000KB, and keep on doing that until it looks as-good-as-it-gets in ADE. Only then do I divide it into cover, title page, contents, prelims, chapter1 ... etc. Unsurprisingly, this also works as I want in ADE (and makes a good mobi in Calibre). So the question is: Why do I need to divide the book at all? The html links from Contents to each chapter and to some individual elements within chapters all work perfectly. Are there ebook readers that can't handle 1,000KB?

[theducks]

Originally Posted by MalcolmRM

Wow! Lots to try there but - with a lot of help from my friends (that's you-all) I think I've got there. One (more) idiot question: I do all the editing and tidying up and standardising in an undivided epub, which gives me an undivided html to search/replace. When satisfied, I save it as [novel name]undivided.epub, around 1,000KB, and keep on doing that until it looks as-good-as-it-gets in ADE. Only then do I divide it into cover, title page, contents, prelims, chapter1 ... etc. Unsurprisingly, this also works as I want in ADE (and makes a good mobi in Calibre). So the question is: Why do I need to divide the book at all? The html links from Contents to each chapter and to some individual elements within chapters all work perfectly. Are there ebook readers that can't handle 1,000KB?

Because performance will suck on many older, dedicated devices if it runs at all.
260K max file (split) size is the rule of thumb.

IMHO, there is no solid reason not to start with a split up book. That is what the 'All HTML files' search is for

I work on a single split (current file mode) and if that S&R is solid, I loosen up the limits to 'Selected' or 'All'

[mrmikel]

On my PRS 300 your single block file will show up as 1 page, it won't open, that is, if it doesn't lock up the reader.

You will find, depending on the material, that actions in the editor will work faster if the chunks are smaller. As theducks said, you can control the damage from search and replace gone wrong if you work on a chapter size chunk at a time.

Your experience may vary. I work on PDF conversions which are full of errors and require endless, so it seems, search and replaces and therefore many chances to go wrong. If you work on original material which is not endlessly buggy, this may be less of an issue. All it will take is one search and replace gone invisibly wrong and finding it out several hours later to convince you that confining your search and replace can be a very useful stress reduction strategy.

Case in point from me yesterday. The book has lots of sentences where the wordsruntogether and even at the end of the sentence it goes from period .To the next word.

An easy search and replace case, right? Only if there are no images because image.png becomes image. png and will not be found or displayed!

[eschwartz]

@mrmikel, there are regexes to catch that.

Try only finding ".[A-Z]" in caps, or else negative lookahead to avoid file extensions.

[mrmikel]

I guess I don't have any faith in my ability to foresee the possible pitfalls.I seem to find them only after the fact. But thanks for pointing it out.

[eschwartz]

That's why the calibre Editor has global undo plus checkpoint diffs.