How to repair broken lines?

[Jane Eyre]

Thank you but... I know nothing about regex.

[Turtle91]

Originally Posted by Jane Eyre

Thank you but... I know nothing about regex.

It’s just a fancy find/replace with some variables thrown in where you want multiple “finds”. Luckily you have been given an example or two where you can simply copy/paste the “find” portion, and then copy/paste the “replace” portion.

Then step thru the find replace one item at a time - that covers 99.99% of JSWolfe’s insistence on the “manual” method.

I also highly advise taking 10 minutes to learn the very basics of regex… it makes editing soooo much easier!!!

[Jane Eyre]

Originally Posted by Turtle91

It’s just a fancy find/replace with some variables thrown in where you want multiple “finds”. Luckily you have been given an example or two where you can simply copy/paste the “find” portion, and then copy/paste the “replace” portion.

Then step thru the find replace one item at a time - that covers 99.99% of JSWolfe’s insistence on the “manual” method.

I also highly advise taking 10 minutes to learn the very basics of regex… it makes editing soooo much easier!!!

I played for 30 seconds with Calibre, but nothing happened. So frustrating..
Useless. I won't try anything more.

[Turtle91]

lolz! That sounds a lot like "I sat in the seat and poked a couple of buttons but the car didn't drive anywhere!!"

If you quit too early on learning anything, then you will never learn anything... at least that is what I teach my grandkids when they give me answers like that! The short version is "Use your own brain, not mine. I already know how to do it!"

Cheers,

[Jane Eyre]

No. I don't want to waste my (little) daytime to do nonsense tasks. I would rather use it to READ the books that I already have

[KevinH]

Then read them as pdfs and be done.

[PeterT]

You might want to experiment with "Heuristic processing" options during the conversion; especially the "unwrap lines" setting. See https://manual.calibre-ebook.com/con...tic-processing

[Jane Eyre]

I'd better purchase a paperback copy. It must cost very little money.
Tanks to everybody!

Originally Posted by PeterT

You might want to experiment with "Heuristic processing" options during the conversion; especially the "unwrap lines" setting. See https://manual.calibre-ebook.com/con...tic-processing

Too hard for me, because of a book which I don't even want to (re)read, truly.
It's been a sort of nostalgy, really.

[DNSB]

I would agree with @gbm and @Karellen. You can automate much of the correction with @gbm's code or a similar variant to locate a paragraph ending in a lower case letter with the next paragraph starting with a lower case letter, @Karellen's code for locating a paragraph ending in a - and locating a paragraph ending in a ;:,. After that search and replace which is saved as an automated search/replace in Sigil, I then check for paragraphs starting with a lower case letter and paragraphs ending in a lower case letter and clean those up as needed. About 95% of the work is done by the automated search.

For the first search, I started with the Join Paragraph sample search in Sigil's saved searches:

Find: ([[:alpha:],])</p>\s*<p\b[^>]*>
Replace: \1

and added code similar to @Karellens for the more common non-alphabetic characters.


The last book I ran though the process dropped from 600 paragraphs split incorrectly to 14 that needed manual fixing plus a few random incorrect fixes in the title and copyright pages. The joys of an author after the rights were reverted, getting the electronic versions of their books in PDF format which needed to be converted to ePub and docx format.