minor bug report - html conversion

[cybmole]

more of a weakness than a bug really:

I started reading Mississippi Blues - (nanoquartet vol 2) which I'd converted from a found html version.

from chapter 2 onwards everything went into CAPITALS, centre justified. AND STAYED THAT WAY UNTIL END OF BOOK - you can see how annoying that would be

So I opened the source in both IE8 and chrome & saw the same thing. in IE8 it was actually RED capitals as red was being used for headers, so clearly there was a missing bit of html code.

tried to edit the source file in word 97 but word refused to open it - said there was a problem with mismatched tags.

so as last resport I ran the source through auto kindle. that converted it OK and got rid of the all caps problem.

if anyone wishes to delve into why auto kindle outperformed calibre in this instance I'd be happy to pass on the source file. I told that I can't post it here because of copyright.

I'm trying to get IE8 to show me the source html but as it's book length it's been contemplating the request for 20 mins now with no signs of completion. If I can extract the buit where it goes pear shaped I'll post it for info.

[cybmole]

OK - thanks to wordpad - it goes wrong in the middle of this block, at start of chapter 1. or in the 2nd block - at start of chapter 2:
[code]
but only another spell of the blindness I’ve been experiencing, I might almost be dead again.</p>
<p>Save for that distant music, beginning again, a bright, tenuous thread to which I cling.</p>
</div>
<hr /><h5><a name="part_1" id="part_1"></a>1<br /><span class="title">Trouble in Mind</span></h5>
</div>
<hr /><h3><a name="chapter_1" id="chapter_1"></a>One<br /><span class="title">Resurrection Blues</span</h3>
<div class="right">
<a href="#prologue"><img src="Previous.gif" class="noborder" alt="Next" title ="Next" /></a>
<a href="#contents"><img src="Top.gif" class="noborder" alt="Top" title="Top" /></a>
<a href="#chapter_2"><img src="Next.gif" class="noborder" alt="Next" title ="Next" /></a>
</div>
<div>
<p class="first">T<em>he first sunset after Cincinnati’s awakening was fiery, flaring incandescent red and yellow sheets of light through transparent intense blue, coloring the coming
[code]

2nd extract
[code]
the rough water above the rapids. Verity had one final glimpse of golden, moving, surging matter…</p>
<p><em>Good-bye, my lovely City.</em></p>
<p>Then they moved beyond that point, ever more swiftly.</p>
<p>Lighting out.</p>
</div>
<hr /><h5><a name="part_2" id="part_2"></a>2<br /><span class="title">Beautiful Ohio</span></h5>
</div>
<hr /><h3><a name="chapter_2" id="chapter_2"></a>Two<br /><span class="title">Free-Fall Blues</span></h5>
<div class="right">
<a href="#chapter_1"><img src="Previous.gif" class="noborder" alt="Previous" title="Previous" /></a>
<a href="#contents"><img src="Top.gif" class="noborder" alt="Top" title="Top" /></a>
<a href="#chapter_3"><img src="Next.gif" class="noborder" alt="Next" title ="Next" /></a>
</div>
<div>
<p class="first">Her own crazed laugh startled Verity. Torn from her, it blended with the whitecapped waves heaving round the raft.
[code]

[ldolse]

You should open a bug on bugs.calibre-ebook.com. Attach the source format of the book. There isn't any way to tell you why from this small snippet of code.

[cybmole]

Originally Posted by ldolse

You should open a bug on bugs.calibre-ebook.com. Attach the source format of the book. There isn't any way to tell you why from this small snippet of code.

damn - took me ages to extract those snippets!

well technically the bug is in the source file, not in calibre.

word97 error message (document cannot be opened - details) is:

end tag "div" does not match the start tag "body"
location line 376 column 3

[Perkin]

In the second block,

<hr /><h3><a name="chapter_2" id="chapter_2"></a>Two<br /><span class="title">Free-Fall Blues</span></h5>

should probable be

<hr /><h3><a name="chapter_2" id="chapter_2"></a>Two<br /><span class="title">Free-Fall Blues</span></h3>

Change the last </h5> tag to </h3> to match it's opening <h3> tag

[ldolse]

That would make the text go bold throughout the book, but it shouldn't make it all capitals..

Unless the CSS style for H3 also says to make it all caps...

[ldolse]

You should probably set Calibre's "Page breaks before" setting in structure detection to break on H3. That might limit the portion of the book that's affected.

[cybmole]

thanks - the source for all 4 books in this quartet is poor, but it seems to be the only source out there. they are not available to buy from Amazon kindle

[Manichean]

A comment on your "extract source"- problems: HTML files are really just text files with a special syntax and a different file extension. Try opening them in Windows' builtin editor, or, for nice things like syntax highlighting, something like Notepad++. That should go way faster to open.

[Dopedangel]

I would recommend running any html you want to convert using calibre through html tidy in notepad++ apart from catching errors it helps removing many of the crap that microsoft word adds to an html file.
Its really improves the conversion speed specially to epub.

[Perkin]

In relation to my previous post (#5), you could also change the start <h3> to <h5> to match the end tag, which would probably be more inline with the rest of the code given and what was wanted.

This should fix the mismatched tag issue (unless there's more of them)

As Idolse says, it would then depend on css for h3/h5 classes.

In the code of the file, Is the text normal, or is that all uppercase. If it's normal, as Idolse says (again) try looking at the css/styleheader for 'text-transform:uppercase;'
and removing it.

[cybmole]

Originally Posted by Manichean

A comment on your "extract source"- problems: HTML files are really just text files with a special syntax and a different file extension. Try opening them in Windows' builtin editor, or, for nice things like syntax highlighting, something like Notepad++. That should go way faster to open.

Hmm - my windows "built in editor" appears to be word 97, which will not open the file.
I guess the install of office alters some original windows 7 default ?

PS I think I've made the book readable enough for my needs now, via the auto kindle route. it's not perfect but I can live with it.

[Manichean]

Originally Posted by cybmole

Hmm - my windows "built in editor" appears to be word 97, which will not open the file.
I guess the install of office alters some original windows 7 default ?

I'm not talking about your word processor, but the built-in text editor. You should be able to find that in the accessories group in the start menu. Although personally, I'd recommend Notepad++.

[theducks]

Originally Posted by cybmole

Hmm - my windows "built in editor" appears to be word 97, which will not open the file.
I guess the install of office alters some original windows 7 default ?

PS I think I've made the book readable enough for my needs now, via the auto kindle route. it's not perfect but I can live with it.

Windows also has the built in TEXT editor: notepad.exe that is safer to use as it adds nothing ON IT'S OWN to your file.

Right click on a file. Select "Open With..." choose Notepad"

Get Notepad++ (free), (also suggested by Manichean) way better if you are doing code tweaking.

[cybmole]

Originally Posted by theducks

Windows also has the built in TEXT editor: notepad.exe that is safer to use as it adds nothing ON IT'S OWN to your file.

Right click on a file. Select "Open With..." choose Notepad"

:

yes, notepad works - I thought it would choke on file size, bur maybe the win7 version has improved since windows XP days

wordpad also works fine.

& to answer an earlier question the body of story text in source is in correct case, it's just the HTML bugs that are forcing the ALL CAPS outcome.