HTML to ePub stripping out Content text

nimblebooks · 01-25-2012, 02:23 PM

Here is a puzzler. I am running ebook-convert on a HTML toc doc with the following settings:

sudo ebook-convert tmp/temptoc.html $mediatargetpath$sku".epub" --max-levels=1 --toc-threshold=100 --cover=$imagedir$sku$cover_image_extension --book-producer="Nimble Combinatorial Publishing" --publisher="Nimble Combinatorial Publishing" --max-toc-links=100 --preserve-cover-aspect-ratio

the document 1.html referenced by tmp/temptoc.html

http://en.wikipedia.org/w/index.php?...&title=Magento

has a "Contents" section whose html source looks like this:

Quote:

<table id="toc" class="toc">
<tr>
<td>
<div id="toctitle">
<h2>Contents</h2>
</div>
<ul>
<li class="toclevel-1 tocsection-1"><a href="#History"><span class="tocnumber">1</span> <span class="toctext">History</span></a></li>
<li class="toclevel-1 tocsection-2"><a href="#See_also"><span class="tocnumber">2</span> <span class="toctext">See also</span></a></li>
<li class="toclevel-1 tocsection-3"><a href="#References"><span class="tocnumber">3</span> <span class="toctext">References</span></a></li>
<li class="toclevel-1 tocsection-4"><a href="#External_links"><span class="tocnumber">4</span> <span class="toctext">External links</span></a></li>
</ul>
</td>
</tr>
</table>

When Calibre processes this document, it is removing the text from the bullets, so that all that's showing up is four bullets, which looks stupid. I used Sigil to inspect the HTML inside the ePub, and it looks as if Calibre is applying new styles to what it detects as TOC bullets.

Quote:

<body class="calibre">
<table class="toc" id="toc">
<tr class="calibre11">
<td class="calibre15">
<div class="calibre8" id="toctitle">
<h2 class="calibre16" id="calibre_pb_1">Contents</h2>
</div>

<ul class="calibre9">
<li class="toclevel"><a class="calibre5" href="../Text/1_split_000.html#History"></a></li>

<li class="toclevel"><a class="calibre5" href="../Text/1_split_000.html#See_also"></a></li>

<li class="toclevel"><a class="calibre5" href="../Text/1_split_000.html#References"></a></li>

<li class="toclevel"><a class="calibre5" href="../Text/1_split_000.html#External_links"></a></li>
</ul>
</td>
</tr>
</table>
</body>
</html>

Apparently this has something to do with toc detection, but I've been pulling my hair out and haven't gotten anywhere. Can some kind soul speed things along for me?

nimblebooks · 01-27-2012, 02:56 PM

Bump! Maybe I provided too much information. The basic problem is this: I am providing an HTML TOC (doc 0) for a collection of n HTML documents. Calibre is correctly crawling and building that TOC document, but, when it encounters a section llabeled "Contents" in doc #1, it is applying its own classes *AND* stripping out the text from between the anchor tags. The result is that doc 1 has a bunch of blank bullets where it should have an unchanged "Contents" section (the contents are of doc 1, not the whole collection). Please help! This is on my critical path.

kovidgoyal · 01-28-2012, 12:28 AM

You provided too little information

A minimal test case is what is needed. i.e. a small set of thml files that show this behavior.

nimblebooks · 01-30-2012, 07:07 PM

Here are the requested htmls, and the ebook-convert used. there are some other problems, but the one that I'm trying to figure out right now is that the "Contents" in document 1.html are blanked out in the resulting ePub.

sudo ebook-convert tmp/temptoc.html $mediatargetpath$sku".epub" --max-levels=1 --toc-threshold=100 --level1-toc="//h:h1" --level2-toc="//h:h2" --cover=$imagedir$sku$cover_image_extension --book-producer="Nimble Combinatorial Publishing" --publisher="Nimble Combinatorial Publishing" --max-toc-links=100 --preserve-cover-aspect-ratio

nimblebooks · 01-31-2012, 07:38 PM

bump!

theducks · 01-31-2012, 08:23 PM

Moderator Notice
Please don't Bump (it is rude to folke that get Mail Notification).
227 People that did not have an answer viewed this thread.

kovidgoyal · 02-01-2012, 01:50 AM

The HTML in those files is a complete mess. Parsing it is failing, it has nothing to do with table of contents. For a start remove the invalid title and head tags at the beginning of the document.

02-01-2012, 01:50 AM	#7
kovidgoyal creator of calibre Posts: 43,858 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	The HTML in those files is a complete mess. Parsing it is failing, it has nothing to do with table of contents. For a start remove the invalid title and head tags at the beginning of the document. Last edited by kovidgoyal; 02-01-2012 at 01:54 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Overlapping text when converting html to mobi/epub	TopCat	Conversion	4	11-28-2011 06:13 AM
HTML to EPUB Inline Text/Image Issue	HoushaSen	Conversion	2	07-02-2011 08:03 PM
Calibre Recipe HTML content differs from raw html of index.html.	krunk	Calibre	4	09-20-2010 09:48 PM

01-27-2012, 02:56 PM	#2
nimblebooks Enthusiast Posts: 28 Karma: 10 Join Date: May 2010 Device: Kindle	Bump! Maybe I provided too much information. The basic problem is this: I am providing an HTML TOC (doc 0) for a collection of n HTML documents. Calibre is correctly crawling and building that TOC document, but, when it encounters a section llabeled "Contents" in doc #1, it is applying its own classes AND stripping out the text from between the anchor tags. The result is that doc 1 has a bunch of blank bullets where it should have an unchanged "Contents" section (the contents are of doc 1, not the whole collection). Please help! This is on my critical path.

01-28-2012, 12:28 AM	#3
kovidgoyal creator of calibre Posts: 43,858 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	You provided too little information A minimal test case is what is needed. i.e. a small set of thml files that show this behavior.

01-31-2012, 07:38 PM	#5
nimblebooks Enthusiast Posts: 28 Karma: 10 Join Date: May 2010 Device: Kindle	bump!

01-31-2012, 08:23 PM	#6
theducks Well trained by Cats Posts: 29,800 Karma: 54830978 Join Date: Aug 2009 Location: The Central Coast of California Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A	Moderator Notice Please don't Bump (it is rude to folke that get Mail Notification). 227 People that did not have an answer viewed this thread.

Advert

Advert