additional breakline after every line (need to wrap)

[gnychis]

It seems that after I convert from PDF to ePub, the conversion has trouble wrapping lines. That is, between every line of text I have an additional blank line (whitespace). It turns out to look like double spaced text. I think this is the opposite problem as unwrapping the text, where I am looking to re-wrap some of it.

I found the un-wrap helper in the structure part of the conversion, is there something for wrapping?

[gnychis]

to provide a little more information, I think this is created by the fact that the PDF uses hard line breaks after each line.

So, if I look at the HTML I see something like this for a single paragraph and transition to the next paragraph (pardon the foreign language):

Οκύριος και η κυρία Ντάρσλι, που έμεναν στο νούμε-<br>
ρο 4 της οδού Πριβέτ, έλεγαν συχνά, και πάντα με<br>
υπερηφάνεια, πως ήταν απόλυτα φυσιολογικοί άν-<br>
θρωποι, τίποτα περισσότερο ή λιγότερο. Ήταν οι τελευταίοι<br>
άνθρωποι που θα περίμενε κανείς να δει ανακατεμένους σε<br>
κάτι παράξενο ή μυστήριο, απλώς και μόνο γιατί και οι ίδιοι<br>
πίστευαν πως δεν υπήρχαν αληθινά τέτοιες ανοησίες στη<br>
ζωή.<br>   ###### TRANSITION TO NEXT PARAGRAPH.....
Ο κύριος Ντάρσλι ήταν ο διευθυντής ενός εργοστασίου<br>
με το όνομα «Γκράνινγκς», το οποίο έφτιαχνε γεωτρύπανα.<br>

[ldolse]

Make sure the line un-wrap factor under pdf input is set to the default of 0.45. If that still doesn't work file a bug at bugs.calibre-ebook.com with an example file.

[gnychis]

It was the default, thanks I can build an example file

[gnychis]

I have created an example and filed a ticket, thanks!

http://bugs.calibre-ebook.com/ticket/7761

[ldolse]

The problem is because the current line unwrapping function requires a Roman alphabet. Handling Greek will require the new pdf engine. There isn't much you can do from the GUI, but you could use the command line with debug output to get an html conversion with the new pdf engine. It won't finish a complete conversion, it will error out, but there will be usable html in the debug directory.

The new engine should un-wrap the lines, but in my experience it un-wraps too much - YMMV. The command line should be something like this:[code]ebook-convert book.pdf book.epub --new-pdf-engine -v --debug-pipeline=/somedirectory/[code] That will vary slightly by platform, filenames, etc.