Incorrect conversion of bigger odt files

[schubi]

Hello Developer,

Just take a bigger odt file with lets say 300 sides and watch the position of
the headlines and the text between origin and converted and you will see
that headlines and text position change in the sides after about 100 pages.

If you have a bigger odt file it breaks the file in several single parts of html-code.
They are called index_split_000.xhtml, index_split_001.xhtml aso.
But the single files are not the same way converted as one complete xhtml file
converted from the odt source would be.
The start of the second xhtml is not at the point where the first xhtml stops.
So you get a partly blank side, because the second xhtml starts with the top
of the documents side and disreguards the origin place of the lines in your
source document.

This error is repeating with every generated single part so you have a big
difference in the format and headlines (which you usually prefer to be at
the start of the side in your original file) in comparison to the new generated
file throughout the converted document.

I also looked for a possibility to switch off the cutting in single parts, but could not find one.

Best regards
Kurt Schubert
(schubi)

[schubi (schubert-hd)]

sides = pages

sorry ;-)

[Kovid Goyal (kovid)]

Size of file is irrelevant to how splitting works. If you have a
particular file for which splitting is not working correctly, attach it
to this bug report.

If you want to turn off splitting look under the EPUB Output section of
the conversion dialog.

 status invalid

[schubi (schubert-hd)]

Dear Kovid,

I see this effect each time when converting so I am sure it happens.
The dividing in several xhtml Files is riping the documents structure
in the described way during the change from one to the next xhtml.split.
This happens basically and is not depending on the number of
xhtml files.
The one stops for example in the middle of the page and the
next starts at the top of the page.

For your hint:
I only found to change the size for splitting which is set to 260kb,
but no "turning off" - possibility and the given size is higher than my
file is.

And with my files size of 215kb I get 4 xhtml splits...

Cannot attach it, will not send my book anywhere, maybe you check
my description of the bug again.

Thanks
Kurt Schubert

SoftwareEngineer VB6, Java, ABAP

Kovid Goyal schrieb:
> Size of file is irrelevant to how splitting works. If you have a
> particular file for which splitting is not working correctly, attach it
> to this bug report.
>
> If you want to turn off splitting look under the EPUB Output section of
> the conversion dialog.
>
> status invalid
>
> ** Changed in: calibre
> Status: New => Invalid
>

[schubi (schubert-hd)]

Dear Kovid,

I made some test file for you including the result.
(to be honest, something like that would be my first action
if I would develop a converter...)

Regards
Kurt

Dear Kovid,

I see this effect each time when converting so I am sure it happens.
The dividing in several xhtml Files is riping the documents structure
in the described way during the change from one to the next xhtml.split.
This happens basically and is not depending on the number of
xhtml files.
The one stops for example in the middle of the page and the
next starts at the top of the page.

For your hint:
I only found to change the size for splitting which is set to 260kb,
but no "turning off" - possibility and the given size is higher than my
file is.

And with my files size of 215kb I get 4 xhtml splits...

Cannot attach it, will not send my book anywhere, maybe you check
my description of the bug again.

Thanks
Kurt Schubert

SoftwareEngineer VB6, Java, ABAP

Kovid Goyal schrieb:
> Size of file is irrelevant to how splitting works. If you have a
> particular file for which splitting is not working correctly, attach it
> to this bug report.
>
> If you want to turn off splitting look under the EPUB Output section of
> the conversion dialog.
>
> status invalid
>
> ** Changed in: calibre
> Status: New => Invalid
>

[Kovid Goyal (kovid)]

The options you are looking for are

"Do not split on page breaks"

and

"Split files larger than" -- set it to 0 to disable

[Kovid Goyal (kovid)]

And I can see absolutely nothing wrong in your examples.

In the unsplit html file at the location of the fisrt split we have

<p class="P-P2 c_odt4">Ich hatte mir zur Feier des Samstagnachmittags ein bisschen </p>
<p class="P-P2 c_odt4">von dem ganz besonders guten Lawendelduft rein gekippt und </p>
<p class="P-P2 c_odt4">so saß ich hinterher ziemlich entspannt und gut gelaunt vor </p>
<p class="P-P2 c_odt4">dem Fernseher, gerade dabei, die Welt mal wieder ein bisschen </p>
<p class="P-P2 c_odt4">rosiger zu betrachten.</p>
<p class="P-P2 c_odt4"> </p>
<p class="P-P2 c_odt4">Die Tür geht auf, die Frau kommt rein, sagt ärgerlich, was ich </p>
<p class="P-P2 c_odt4">denn mit der Lampe gemacht hätte und das fände sie nicht </p>
<p class="P-P2 c_odt4">sonderlich lustig.</p>
<p class="P-P2 c_odt4"> </p>

In the split files, we have

index_split_000.xhtml
<p class="p-p">Ich hatte mir zur Feier des Samstagnachmittags ein bisschen </p>
<p class="p-p">von dem ganz besonders guten Lawendelduft rein gekippt und </p>
<p class="p-p">so saß ich hinterher ziemlich entspannt und gut gelaunt vor </p>
<p class="p-p">dem Fernseher, gerade dabei, die Welt mal wieder ein bisschen </p>
<p class="p-p">rosiger zu betrachten.</p>
<p class="p-p"> </p>
<p class="p-p">Die Tür geht auf, die Frau kommt rein, sagt ärgerlich, was ich </p>
<p class="p-p">denn mit der Lampe gemacht hätte und das fände sie nicht </p>

----------

index_split_001.xhtml
<p class="p-p">sonderlich lustig.</p>
<p class="p-p"> </p>

So the split is perfectly correct.

If you are complaining that the split causes there to be blank space at the end of the page where there was no blank space to start with -- that cannot be helped, that is the nature of epub.

Either turn off splitting as I told you to, or insert some page breaks in your document so that the splitter can break at those locations instead of at randomly chosen paragraphs.

[schubi (schubert-hd)]

Hello Kovid,

Thanks, I see you try to help!

"If you are complaining that the split causes there to be blank space at
the end of
the page where there was no blank space to start with -- that cannot be
helped,
  that is the nature of epub.

"

Yes, I have been talking about this.
It seems no way to get this behaviour in a correct way during the splitting.
So I have to turn off this.
For declaration, I have a german translated version, where its written:
Bei Seitenwechseln nicht aufteilen
which means:
When changing page do not divide (split)
so I had no relation to my problem, because it was not associated to
pages...
Sorry for that.

I could set the hook there now, but I cannot put the size down to 0 it
stays at 25 KB
and so I have still the splitting.

To put the text the way to have no blank space on that place will not help
because when I add more text, the splitting point moves and so the affected
text, too.

Ok, but if we get the splitting off, this would be fine.

I actually have win7/64bit and calibre 2.44.1/64bit.

Regards
Kurt

Kovid Goyal schrieb:
> And I can see absolutely nothing wrong in your examples.
>
> In the unsplit html file at the location of the fisrt split we have
>
> <p class="P-P2 c_odt4">Ich hatte mir zur Feier des Samstagnachmittags ein bisschen </p>
> <p class="P-P2 c_odt4">von dem ganz besonders guten Lawendelduft rein gekippt und </p>
> <p class="P-P2 c_odt4">so saß ich hinterher ziemlich entspannt und gut gelaunt vor </p>
> <p class="P-P2 c_odt4">dem Fernseher, gerade dabei, die Welt mal wieder ein bisschen </p>
> <p class="P-P2 c_odt4">rosiger zu betrachten.</p>
> <p class="P-P2 c_odt4"> </p>
> <p class="P-P2 c_odt4">Die Tür geht auf, die Frau kommt rein, sagt ärgerlich, was ich </p>
> <p class="P-P2 c_odt4">denn mit der Lampe gemacht hätte und das fände sie nicht </p>
> <p class="P-P2 c_odt4">sonderlich lustig.</p>
> <p class="P-P2 c_odt4"> </p>
>
>
> In the split files, we have
>
> index_split_000.xhtml
> <p class="p-p">Ich hatte mir zur Feier des Samstagnachmittags ein bisschen </p>
> <p class="p-p">von dem ganz besonders guten Lawendelduft rein gekippt und </p>
> <p class="p-p">so saß ich hinterher ziemlich entspannt und gut gelaunt vor </p>
> <p class="p-p">dem Fernseher, gerade dabei, die Welt mal wieder ein bisschen </p>
> <p class="p-p">rosiger zu betrachten.</p>
> <p class="p-p"> </p>
> <p class="p-p">Die Tür geht auf, die Frau kommt rein, sagt ärgerlich, was ich </p>
> <p class="p-p">denn mit der Lampe gemacht hätte und das fände sie nicht </p>
>
> ----------
>
> index_split_001.xhtml
> <p class="p-p">sonderlich lustig.</p>
> <p class="p-p"> </p>
>
>
> So the split is perfectly correct.
>
> If you are complaining that the split causes there to be blank space at
> the end of the page where there was no blank space to start with -- that
> cannot be helped, that is the nature of epub.
>
> Either turn off splitting as I told you to, or insert some page breaks
> in your document so that the splitter can break at those locations
> instead of at randomly chosen paragraphs.
>

[Kovid Goyal (kovid)]

So make the split size very large instead of zero, same effect.

[schubi (schubert-hd)]

Yes, this helped.

Thanks!

Kovid Goyal schrieb:
> So make the split size very large instead of zero, same effect.
>