Quirk: toc.ncx reformats itself

jcsalomon · 05-20-2016, 11:29 AM

I’m using Sigil for EPUB3 books, and using Git to track changes (by unzipping, and re-zipping with a batch file that does the right thing; see elsewhere on this site). I’ve noticed that toc.ncx takes any opportunity to reformat itself from the for generated by “Tools » Epub3 Tools » Generate NCX from Nav”—nothing substantial, just changing the indentation between

Code:

<?xml version="1.0" encoding="utf-8" ?>
<ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/">
  <head>
    <meta content="401e6912-cfe6-4cb7-8e02-f6ddea0d6467" name="dtb:uid"/>
    <meta content="1" name="dtb:depth"/>
    <meta content="0" name="dtb:totalPageCount"/>
    <meta content="0" name="dtb:maxPageNumber"/>
  </head>

and

Code:

<?xml version="1.0" encoding="utf-8" ?>
<ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/">
<head>
  <meta content="401e6912-cfe6-4cb7-8e02-f6ddea0d6467" name="dtb:uid"/>
  <meta content="1" name="dtb:depth"/>
  <meta content="0" name="dtb:totalPageCount"/>
  <meta content="0" name="dtb:maxPageNumber"/>
</head>

—but adding a useless and confusing delta in my repository.

Very low priority, but why is this happening?

KevinH · 05-20-2016, 02:12 PM

Because the code that creates the ncx the first time from the nav is different fron the code that updates the nav/ncx when the toc is edited or changed.

I will look to see if I can track down where it is coming from. You should also look into getting a diff tool that is smart enough to ignore whitspace on xml files.

Plus order of attributes on elements in xml is not sigificant and is not preserved by most xml tools that will also introduce deltas.

KevinH · 05-20-2016, 09:49 PM

The ncxgenerator.py generates your second example (which is what creates it from the Nav). Any time it is changed after initial creation the NCXWriter code which uses Qt QXmlStreamWriter with auto formatting set to true and default indent of 4 not 2. I can change ncxgenerator.py to match the indented head tag and change the QXmlStreamWriter to use indents of 2.

I'll try to get that into the next release when I get back home.

jcsalomon · 05-25-2016, 12:38 PM

Thank you.

05-20-2016, 11:29 AM	#1
jcsalomon Zealot Posts: 100 Karma: 1204 Join Date: Jun 2012 Device: Bookari (née Mantano Reader) on Android; Kindle Fire HD	Quirk: toc.ncx reformats itself I’m using Sigil for EPUB3 books, and using Git to track changes (by unzipping, and re-zipping with a batch file that does the right thing; see elsewhere on this site). I’ve noticed that toc.ncx takes any opportunity to reformat itself from the for generated by “Tools » Epub3 Tools » Generate NCX from Nav”—nothing substantial, just changing the indentation between Code: <?xml version="1.0" encoding="utf-8" ?> <ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/"> <head> <meta content="401e6912-cfe6-4cb7-8e02-f6ddea0d6467" name="dtb:uid"/> <meta content="1" name="dtb:depth"/> <meta content="0" name="dtb:totalPageCount"/> <meta content="0" name="dtb:maxPageNumber"/> </head> and Code: <?xml version="1.0" encoding="utf-8" ?> <ncx version="2005-1" xmlns="http://www.daisy.org/z3986/2005/ncx/"> <head> <meta content="401e6912-cfe6-4cb7-8e02-f6ddea0d6467" name="dtb:uid"/> <meta content="1" name="dtb:depth"/> <meta content="0" name="dtb:totalPageCount"/> <meta content="0" name="dtb:maxPageNumber"/> </head> —but adding a useless and confusing delta in my repository. Very low priority, but why is this happening?

05-20-2016, 02:12 PM	#2
KevinH Sigil Developer Posts: 8,487 Karma: 5703586 Join Date: Nov 2009 Device: many	Because the code that creates the ncx the first time from the nav is different fron the code that updates the nav/ncx when the toc is edited or changed. I will look to see if I can track down where it is coming from. You should also look into getting a diff tool that is smart enough to ignore whitspace on xml files. Plus order of attributes on elements in xml is not sigificant and is not preserved by most xml tools that will also introduce deltas. Last edited by KevinH; 05-20-2016 at 02:15 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Help with toc.ncx file please	AlexBell	ePub	7	09-04-2013 02:11 AM
Kindler previewer not recognizing toc.ncx file, my html toc, or the start point...	petercrowell	Kindle Formats	2	05-01-2012 08:14 AM
Toc.ncx	JDawg	Amazon Kindle	11	07-29-2011 04:05 AM
Inline TOC from toc.ncx	elmago79	Kindle Formats	38	03-25-2011 12:56 PM
Use Regex to Code an Inline TOC, from an External TOC's .ncx File	mostlynovels	ePub	2	03-16-2011 12:15 PM

05-20-2016, 09:49 PM	#3
KevinH Sigil Developer Posts: 8,487 Karma: 5703586 Join Date: Nov 2009 Device: many	The ncxgenerator.py generates your second example (which is what creates it from the Nav). Any time it is changed after initial creation the NCXWriter code which uses Qt QXmlStreamWriter with auto formatting set to true and default indent of 4 not 2. I can change ncxgenerator.py to match the indented head tag and change the QXmlStreamWriter to use indents of 2. I'll try to get that into the next release when I get back home.

05-25-2016, 12:38 PM	#4
jcsalomon Zealot Posts: 100 Karma: 1204 Join Date: Jun 2012 Device: Bookari (née Mantano Reader) on Android; Kindle Fire HD	Thank you.

Advert