PDF to ePub problem

[Dark123]

I converted something in PDF to ePub (This only happened with 1 file)
This is how is came out.
Hello this is a
test.

Anyone know how to fix this?

[ac4lt]

I suspect you can fix this by going to the "pdf input" section on the conversion dialog and changing the line unwrapping value to 0.5. That worked for me.

[Dark123]

Thank you ac4lt. That worked.

[Vargagy]

Hi,

I am new user of Calibre, and new member in forum. Regret my english isn't very good, my original language is hungarian. Sorry!

My deal is to convert Adobe Indesign CS2 origin books to epub format. CS3 and CS4 contains epub export, but CS2 not. Avoid of problem I exported text from InDesign in PDF format, and converted the PDF by Calibre.

In the Calibre converted epub file are broken lines, the last word of paragraph stays in a stand alone line, after this line begins in a newer line the next paragraph. The broken line isn't in all paragraph, only in the few, and I can not found some sytem in that.

If I change something in the layout (e.g. shorter lines) and tha structure of paragraph, the error on the earlier place go out, but on newer place coming in.

Do You found already this effect? What can I do?

[JSWolf]

The problem is there is NO program that can convert PDF to any other format without errors. So once you convert the PDF, the only choice you have is to a/b compare the PDF to the output and fix all the errors in the conversion. I know it's tedious. But, there's nothing else you can do.

[bthoven]

Originally Posted by ac4lt

I suspect you can fix this by going to the "pdf input" section on the conversion dialog and changing the line unwrapping value to 0.5. That worked for me.

Hi Ac4lt...thanks for your tip, it works for me too!

By the way, what this setting is doing?

[ac4lt]

I *think* it's deciding at what percentage of a line length it needs to pull in the next line, but I'm not exactly sure how it's implemented.

[AnemicOak]

Originally Posted by Vargagy

My deal is to convert Adobe Indesign CS2 origin books to epub format. CS3 and CS4 contains epub export, but CS2 not. Avoid of problem I exported text from InDesign in PDF format, and converted the PDF by Calibre.

Does CS2 have export to XHTML? I can't remember if that was added in CS2 or CS3. If so you could do that and use Sigil to format your ePub. Or perhaps exporting to XML and converting that to HTML and going from there or something? Not sure if there's a converter that will go from XML straight to ePub.

[alecE]

Fwiw, I've converted 20+ pdf titles to epub. I haven't found any foolproof, quick way of doing this. My rather laborious procedure now is:
1. If necessary, remove the pdf restrictions which prevent exporting as text;
2. Use the File | Save As Text menu option to create a .txt version;
3. Using the text editor of your choice, remove page headers, page numbers, extraneous front matter. Then find and change quotes, apostrophes and accents etc., to html named entities (e.g., " --> “
4. Create an epub file using Sigil (thank you so, so much Valloric)
5. Load into Calibre (thank you Kovid);
6. Transfer to Sony reader;
7. Read, enjoy, bookmark the typos;
8. Back to Sigil, correct all the typos etc., repeat steps 4 to 7.

This works, it's labour intensive, and during the process you may lose the will to live.

[kjk]

Originally Posted by alecE

Fwiw, I've converted 20+ pdf titles to epub. I haven't found any foolproof, quick way of doing this. My rather laborious procedure now is:
1. If necessary, remove the pdf restrictions which prevent exporting as text;
2. Use the File | Save As Text menu option to create a .txt version;

This works, it's labour intensive, and during the process you may lose the will to live.

ahahaha I feel your pain.
I am curious though, why do you go all the way back to text? Don't you lose italics, bold, etc.? Or do you put that back in during editing?

[chaley]

I have had good luck converting single-column PDFs by:
1. Crop margins of the PDF to the text area.
2. Save as RTF
3. Use Calibre to convert RTF to EPUB
4. Use Sigil to fix line breaks across page breaks (and a few others)

This method conserves all the character formatting, and (with one exception) the 10 or so files I have converted hasn't resulted in a large amount of repair work.

The one case involved fixing up chapter headings. The original document had chapters in the form
ROMAN_NUMERAL.
CHAPTER TITLE
followed by a few empty lines. The easiest way for me to fix this was to use VIM's global search & replace on the .html that came out of the epub conversion. I used a regexp that matched the two lines and replaced them with a single line wrapped in <h1></h1> tags. I tried using sigil, but couldn't figure out how to make a multi-line match expression (I admit I didn't look for a long time).

[Vargagy]

Originally Posted by AnemicOak

Does CS2 have export to XHTML? I can't remember if that was added in CS2 or CS3. If so you could do that and use Sigil to format your ePub. Or perhaps exporting to XML and converting that to HTML and going from there or something? Not sure if there's a converter that will go from XML straight to ePub.

Yes, the actual solution of problem is the simple exporting of text in HTML format. This is going to convert into the Calibre. The result ePub is correct in Calibre reader and some other ePub reader, but - it isn't' understable - not correct in the Adobe Digital Edition, all in east-europe used akcent-characters (e.g. á, é, ö, ü) substitued with "?"

[hidari]

Originally Posted by Vargagy

Yes, the actual solution of problem is the simple exporting of text in HTML format. This is going to convert into the Calibre. The result ePub is correct in Calibre reader and some other ePub reader, but - it isn't' understable - not correct in the Adobe Digital Edition, all in east-europe used akcent-characters (e.g. á, é, ö, ü) substitued with "?"

Just in Case you want to do it a more simple way: I myself have converted quite a few books from PDF to epub. like alicE said convert it to a text format.

Then, I pop it into Calibre. Conver it to epub and it's ready to read.

Not sure about you but I care not if there are mistakes in quotes/indentations etc...... Yes I am a heathen I drink instant coffee. BUT for me content is more important than the bits and bobs...

sum up:

1. convert PDF to text
2. Convert .txt format to epub with Calibre
3. Read..........

[alecE]

Originally Posted by kjk

ahahaha I feel your pain.
I am curious though, why do you go all the way back to text? Don't you lose italics, bold, etc.? Or do you put that back in during editing?

I like to go all the way back to text as that gives me the greatest control. With respect to italics etc., yes, I then have to put those back in, but, so many titles are poorly edited to begin with that very often there are errors in the formatting which have to be corrected anyway.
With respect to the pain, whilst asprin helps, a single malt whisky is even better.

[avresbo]

Does calibre recognize *.doc?
Why replace special caracters when using *.txt? The original is correct.
Thank you for your help.