OpenOffice to ePub export extension

[Cid]

Hi there, I'm new to the forum and have to admit I didnt read the entire thread, but I do have a small question. But first let me thank you for this amazingly helpful extension!

I have found 0.3.7 exports the Paragraph/heading structure nicely. However I have a lot of trouble as Umlaute (ä, ü, ö and such) are not exported correctly from OpenOffice. Saving directly to PDF or HTML from OO works.
Opening a xhtml page from the epub gives a "XML Parsing Error: not well-formed" for a Umlaut.

I have tried changing the character encoding for export in OO but it didn't change the result.

I have also used the Writer2Latex Extension which also features ePub, however that extension can not properly export the ToC from the paragraphs.

Also, somehow I can't unzip the epub file created by Writer2Latex, so merging your proper writer2epub toc with the properly exported Writer2Latex xhtml does not work.

[eBookLuke]

Hello Cid,

in a couple of days I will release the final version of the extension, that solves many issues you found in 0.3.x

Thanks for yuor testing

Luke

[st_albert]

Originally Posted by Cid

Hi there, I'm new to the forum and have to admit I didnt read the entire thread, but I do have a small question. But first let me thank you for this amazingly helpful extension!

I have found 0.3.7 exports the Paragraph/heading structure nicely. However I have a lot of trouble as Umlaute (ä, ü, ö and such) are not exported correctly from OpenOffice. Saving directly to PDF or HTML from OO works.
Opening a xhtml page from the epub gives a "XML Parsing Error: not well-formed" for a Umlaut.

I have tried changing the character encoding for export in OO but it didn't change the result.

I have also used the Writer2Latex Extension which also features ePub, however that extension can not properly export the ToC from the paragraphs.

Also, somehow I can't unzip the epub file created by Writer2Latex, so merging your proper writer2epub toc with the properly exported Writer2Latex xhtml does not work.

Hola, Cid!

I also have writer2latex installed (it comes by default on the linux distro I use) but I had / have no idea how to use it to export to epub. I really don't think that "writer2latex" and "writer2epub" were intended to work together. But I could easily be wrong about that. If I'm missing something, please enlighten me.

And BTW, for me the same applies to "writer2xhtml", another very useful plugin for OOo, though not nearly as direct a path to an epub as is "writer2epub".

[roger64]

Originally Posted by eBookLuke

in a couple of days I will release the final version of the extension, that solves many issues you found in 0.3.x
Luke

Hi Luke

Congratulations for your nice and useful work. I am available to translate your work in French, preferably with your "final" version

But i'll probably need basic explanations. I'm none too bright.

[Cid]

Originally Posted by st_albert

Hola, Cid!

I also have writer2latex installed (it comes by default on the linux distro I use) but I had / have no idea how to use it to export to epub. I really don't think that "writer2latex" and "writer2epub" were intended to work together. But I could easily be wrong about that. If I'm missing something, please enlighten me.

And BTW, for me the same applies to "writer2xhtml", another very useful plugin for OOo, though not nearly as direct a path to an epub as is "writer2epub".

There is a new version on the sourceforge site for it, where it lists export to epub as functionality. However I have tested both results on my Sony PRS-600, and none worked. The epub exported through writer2latex didn't open (error (!)) and the export by writer2epub worked up to the malformed character, then nothing from there on. The weird thing is that both worked with the epub viewer of Calibre...

[JvdW]

Originally Posted by Cid

There is a new version on the sourceforge site for it, where it lists export to epub as functionality. However I have tested both results on my Sony PRS-600, and none worked. The epub exported through writer2latex didn't open (error (!)) and the export by writer2epub worked up to the malformed character, then nothing from there on. The weird thing is that both worked with the epub viewer of Calibre...

Thats weird. Which version you're talking about?
Last one I tested was the 1.14 development version. Downloaded that and unpacked it, then just double clicked on the writer2xhtml.oxt file and let OOo do its job of installing it. Restarted OOo and loaded up doc/usermanual.odt and exported it to epub. Dropped it into my PRS-505 and read it. Looks OK to me.

The version (1.14) I used didn't have problems with ToC, als not if its a really big file.

Regards,

Joop

[Cid]

Weird indeed. I'm also using the 1.1.4 dev snapshot. Hmmm. I'm a bit at a loss, the text file shouldn't be the source of the problem... Maybe it's because of my system setup, running a german OO in an otherwise EN OS environment (plus the. Anyway, I'm looking foward to writer2epub 1.0.

Will report back with my results.

[eBookLuke]

To Cid:
sometimes two extensions can conflict, when there are subroutines with the same name in different extension. Try to remove all of the extensions and to reinstall one at time.

to JvdW:
Would you translate W2E in Dutch?

To Roger64:
Thanks, but Zelda was faster But I can ask to you for translation of the "user manual", if you can...

Thanks to all

Luke

[st_albert]

Well first of all thanks to Cid and others for the pointer to the updated writer2latex extensions. I installed them, and sure enough, you can export to epub. So I did, using a novel I'm also using to test writer2epub.

One thing I noticed first of all was that it does not split up the text into < 300kb sections. So that may be one reason you have loading problems on your Sony device, if the file is too big. (And I don't think Calibre or Adobe Digital Editions have that size limitation, so it SHOULD be readable there, as you noted.)

I also noticed that all it seems to do is export the document in xhtml format as a single file, and produces a very simple epub wrapper for it. Thus, no separate CSS stylesheet, no splitting into bite-sized text chunks, etc. Also, non-standard names are used, such as "book.opf" instead of the more usual "content.opf."

Granted writer2latex 1.14 is still in alpha, but at this point I think you'd be better off using writer2epub, for the sake of the more refined structure of the epub it creates. BTW, The two extensions don't seem to conflict on my installation, at least. (EN versions of everything).

[roger64]

Originally Posted by eBookLuke

To Roger64:
Thanks, but Zelda was faster But I can ask to you for translation of the "user manual", if you can...
Luke

OK. I just hope it's has not the size of a pocket Bible, if so I would probably review my commitment.

[DoctorOhh]

Originally Posted by st_albert

One thing I noticed first of all was that it does not split up the text into < 300kb sections. So that may be one reason you have loading problems on your Sony device, if the file is too big. (And I don't think Calibre or Adobe Digital Editions have that size limitation, so it SHOULD be readable there, as you noted.)

You are correct that would be a problem for the Sony. The Adobe epub renderer used in Sony and other devices was the original source of the 300k limit. Since readers have limited space and power this limit was implemented as a quality control feature.

[st_albert]

Originally Posted by dwanthny

You are correct that would be a problem for the Sony. The Adobe epub renderer used in Sony and other devices was the original source of the 300k limit. Since readers have limited space and power this limit was implemented as a quality control feature.

And just to be clear, ADE on Windows (desktop) and probably Mac (if there is one) does not have this size limitation. It was the desktop version I was referring to in my post above. I think the Adobe rendering engine for mobile devices has a slightly different name. (Could be wrong, though. I don't follow those details as closely as I should.)

[eBookLuke]

Hi all,

finally I released the 1.0.8 version, that generates a fully validate ePub!!!
Thank to the work of many of you, this extension can do a good job.

There are still some limits (lists, frames, some kind of images) but the extension now works fine

Look at my blog to download it…



Luke

[Cid]

Everything works fine, including ToC and special characters (german Umlaute).

I noticed two things though;
My text is all Arial, but the exported epub is in a serif font as I can see in Calibres epub viewer. I made a reference epub using the Atlantis trial, which displays in the Arial font used in the viewer. (my reader seems to only display serif in both files anyway, so this point is somewhat academic).

The second point is the orientation in tables seems to get lost or reset to left, even though it's "centered" in writer.

Hint: Getting rid of page margins helps greatly in the appearance of the text on the reader.

THANK YOU VERY MUCH, THIS IS AWESOME AND EASY!!!

[eBookLuke]

Originally Posted by Cid

I noticed two things though;

You're right: W2E uses an internal CSS to set the font. Regardless what font you use, W2E always replaces it with Georgia 12.

W3E is not a format converter, but an ePub book creator.
This can be a problem (at the moment) if your original text has mixed fonts, various heights, multiple styles and so on.
But is a big advantage if you start from a bad text (like OCR generated ones) and you want a good epub with just one click. W2E applies all the good tipographic traditional rules, building a well-formatted book regardless of the original formatting. The next versions of W2E will support more text formatting, but still superimpose his own layout.

In the future, I will add an "Expert Mode" that gets out the styles present in the original doc, but this is yet done by Calibre and Atlantis, and if you start from a bad text, you will get a bad ePub.

About the tables misaligned, please send me the original doc, I will check and solve the problem



Luke