k2pdfopt: optimizes PDFs for viewing on e-readers

[famfam]

Tried k2pdfopt with pdf-document in gothic script (german Frakturschrift). The results are to small for me for reading on my device Kobo Clara HD (1072 x 1448 resolution). So until now I have to use my tablet for reading it. That is why I am looking for a 'How to' for pdf in gothic script (german Frakturschrift). Did anyone test the k2pdfopt with Frakturschrift and got good results for his device?

[willus]

Originally Posted by famfam

Tried k2pdfopt with pdf-document in gothic script (german Frakturschrift). The results are to small for me for reading on my device Kobo Clara HD (1072 x 1448 resolution). So until now I have to use my tablet for reading it. That is why I am looking for a 'How to' for pdf in gothic script (german Frakturschrift). Did anyone test the k2pdfopt with Frakturschrift and got good results for his device?

If you can post the source file, or a couple pages from the source file, I can recommend some settings to you. Or PM it to me.

[famfam]

Originally Posted by willus

If you can post the source file, or a couple pages from the source file, I can recommend some settings to you. Or PM it to me.

You can find the source files (pdf) on archive.org.

In the meanwhile I had some succes with the following 2 books in frakturschrift(gothic script):
Eduard Bernstein:
1) Die Geschichte der Berliner Arbeiterbewegung Band 2
2) Sozialismus und Demokratie in der Englischen Revolution

I found them in the internet (legal sources, copyright out of date, author dead since 1932)

I can send you my results from k2pdfopt by pm. I would be happy, if you could find a better way with better results. (How to?)

[famfam]

Originally Posted by willus

If you can post the source file, or a couple pages from the source file, I can recommend some settings to you. Or PM it to me.

I try to append the (cropped and cleaned) source files here.

Sorry, did not work ->

Your submission could not be processed because a security token was missing.

If this occurred unexpectedly, please inform the administrator and describe the action you performed before you received this error.

[willus]

Originally Posted by famfam

I try to append the (cropped and cleaned) source files here.

Sorry, did not work ->

Your submission could not be processed because a security token was missing.

If this occurred unexpectedly, please inform the administrator and describe the action you performed before you received this error.

The files you are trying to send are probably too big. I'm not sure what the limit is. Is there a link you can send for any of the source files.

[famfam]

@willus
Here are the links:

1) Die Geschichte der Berliner Arbeiterbewegung Band 2

https://archive.org/download/diegesc...01berngoog.pdf

or (better)

https://archive.org/download/bub_gb_...knAQAAIAAJ.pdf

and

2) Sozialismus und Demokratie in der Englischen Revolution

https://archive.org/download/bub_gb_...NCAAAAYAAJ.pdf

I don't know how to send my results to you as pm. May be that they are too big. But I zipped them in part of 20 mb. But no success. Can I send them to you by email, and could you send me a pm with your email-address please.
And sure, I cleaned the originel files from empty or unneedet pages, cropped them and so on. Would be easier for you to work with.

[willus]

Originally Posted by famfam

@willus
Here are the links ...
I don't know how to send my results to you as pm. May be that they are too big. But I zipped them in part of 20 mb. But no success. Can I send them to you by email, and could you send me a pm with your email-address please.
And sure, I cleaned the originel files from empty or unneedet pages, cropped them and so on. Would be easier for you to work with.

I have no recommendation for the first file. It is so random--every page seems to be formatted differently. The second file is a more conventional book. I attached two conversions--one using "fit width" and one using larger font and word wrap. I just converted the first 25 pages. Here are the commands:

k2pdfopt -m 0.13in,0.2in,0.13in,0.45in -mode fw -ls- -p 1-25 -o conv_fw.pdf file2.pdf

k2pdfopt -m 0.13in,0.2in,0.13in,0.45in -mag 1.5 -p 1-25 -o conv_wrap.pdf file2.pdf

The -m option crops out the Google watermark.

[famfam]

Originally Posted by willus

I have no recommendation for the first file. It is so random--every page seems to be formatted differently.

I just found a better version of 'Die Geschichte der Berliner Arbeiterbewegung. Teil 2'.
And it was this version, I hat worked with. I only couldn't find the link to it.
Now I found the link in my jpdownloader.

https://archive.org/download/bub_gb_...knAQAAIAAJ.pdf

I now will test your settings.

My procedure is, to clean the files in 'foxit phantom pdf' or in 'adobe acrobat' (cropping is very fast in acrobat). For deleting headers or footers, that are close too to the text, I use the Foxit 'Comment rectangular function'. I also delete all pages (cover, title page, table of contents, dedication, copyright, index), except for the text pages, notes, footnotes, bibliography before editing with k2topdfopt. That is necessary for getting better results. After the k2pdfopt process I add Cover, title page, dedication, copyright. TOC I do manually with Foxit. OCR: Is it better to do ocr with k2pdfop or is it better to do ocr before with Foxit. I think for Frakturschrift I should do it with k2pdfopt and Tesseract traindata. So I did it.

So many questions. Why?
Because I want to know whether cleaning up with Foxit and cutting with Acrobat will unnecessarily inflate the file and how to get smaller files.
Thanks a lot for helping me finding better solutions.

[vds]

Hi, I've been trying to compile k2pdfopt on linux without success. I got the source from here and I'm following the steps described in the readme file.

First I run

gcc -Wall -Ofast -m64 -o k2pdfopt.o -c k2pdfopt.c

and it gives me the following error:

k2pdfopt.c:76:10: fatal error: k2pdfopt.h: No such file or directory
   76 | #include <k2pdfopt.h>
      |          ^~~~~~~~~~~~
compilation terminated.

Then I tried adding the other directories to the search path

gcc -I./willuslib/ -I./k2pdfoptlib/ -Wall -Ofast -m64 -o k2pdfopt.o -c k2pdfopt.c

and this first command runs successfully.
Then I run the second command:

g++ -Ofast -m64 -o k2pdfopt k2pdfopt.o -static -static-libgcc -static-libstdc++ -lk2pdfopt -lwillus -lgocr -ltesseract -lleptonica -ldjvu -lmupdf -lfreetype -ljbig2 -ljpeglib -lopenjpeg -lpng -lzlib -lpthread -lstdc++ -lc -lm

and it gives the following error:

/usr/sbin/ld: cannot find -lk2pdfopt
/usr/sbin/ld: cannot find -lwillus
/usr/sbin/ld: cannot find -lgocr
/usr/sbin/ld: cannot find -ltesseract
/usr/sbin/ld: cannot find -lleptonica
/usr/sbin/ld: cannot find -ldjvu
/usr/sbin/ld: cannot find -lmupdf
/usr/sbin/ld: cannot find -lfreetype
/usr/sbin/ld: cannot find -ljbig2
/usr/sbin/ld: cannot find -ljpeglib
/usr/sbin/ld: cannot find -lopenjpeg
/usr/sbin/ld: cannot find -lpng
/usr/sbin/ld: cannot find -lzlib
collect2: error: ld returned 1 exit status

Has anyone compiled it successfully on Linux? What am I doing wrong?

[famfam]

Just found a Magazine article with 5 columns (Die Zeit, 30.01.2020, Das Corona Virus).
So it would be great, if we could make it happen, to extend the column limit to 5.
Would that be possible without much effort?
Thank you so much.

[willus]

Originally Posted by famfam

Just found a Magazine article with 5 columns (Die Zeit, 30.01.2020, Das Corona Virus).
So it would be great, if we could make it happen, to extend the column limit to 5.
Would that be possible without much effort?
Thank you so much.

You should be able to do this in two passes without any modifications. Do the first pass with text re-flow disabled and limit to two columns max (e.g. -mode 2col). Then do another pass with 4 columns max. If you want to send me a link to the document, I can experiment for you.

[willus]

Originally Posted by vds

Hi, I've been trying to compile k2pdfopt on linux without success. I got the source from here and I'm following the steps described in the readme file.

...

What am I doing wrong?

The archive you linked to does not contain all of the source code for all of the libraries that it needs to create the final binary (except for k2pdfopt and willus--it has the source code for those). You have to build all of those libraries separately and then link to them (e.g. gocr, tesseract, leptonica, djvu, mupdf, etc.).

What version of Linux are you running? Do the i386 linux binaries not work for you?

[vds]

Originally Posted by willus

The archive you linked to does not contain all of the source code for all of the libraries that it needs to create the final binary (except for k2pdfopt and willus--it has the source code for those). You have to build all of those libraries separately and then link to them (e.g. gocr, tesseract, leptonica, djvu, mupdf, etc.).

What version of Linux are you running? Do the i386 linux binaries not work for you?

Thanks for the response. The binaries work fine for me, I'm trying to compile the sources because I had the idea to use emscripten to create a javascript version of k2pdfopt. I'm first trying to compile using gcc before moving to emscripten. I don't even know if it will be possible or if I will be able to do that but looks like something worth trying.

I will try again soon and post the results here.

[willus]

Originally Posted by vds

Thanks for the response. The binaries work fine for me, I'm trying to compile the sources because I had the idea to use emscripten to create a javascript version of k2pdfopt. I'm first trying to compile using gcc before moving to emscripten. I don't even know if it will be possible or if I will be able to do that but looks like something worth trying.

I will try again soon and post the results here.

I think given all of the k2pdfopt dependencies, it will be a small miracle if you get this to work. You might start by modifying the main include file to disable as many of the support libraries as possible (e.g. TESSERACT, DJVU, etc.) just to see if you can get a small subset to work.

[famfam]

Is tesseract integrated in the k2pdfopt-gui and if so version 3 or version 4? Do the traindata files have to be version 3 or version 4? In which folder must the traindata files be in Windows 10: 'Programs' (for 64 bit) or 'Programs (x86)' (for 32 bit)?
In my k2pdfopt-gui I get the error message,

Initializing OCR for 2 threads x x
Could not find Tesseract data (env var TESSDATA_PREFIX = (not assigned)).
Using GOCR v0.50.

What am I doing wrong?
Is my entry in the input window 'Env. var: TESSDATA_PREFIX = c: \ program files \ tesseract-ocr \ tessdata 'not correct?