It is currently Thu Jun 20, 2024 2:54 pm

All times are UTC - 8 hours [ DST ]

Post new topic Reply to topic  [ 4 posts ] 
Author Message
 Post subject: convert gobooks embedded svgs to image
Post #1 Posted: Mon Jun 12, 2023 10:25 am 

Posts: 1
Liked others: 0
Was liked: 0
The ebooks from have interactive diagrams, which is great but not supported on all readers. I have such a book and unfortunately the embedded svgs are unreadable on my Kobo reader (I'm using koreader). I'd like to convert all the svgs to some other supported image format, Has anyone done this? Any tools that might be useful?
Thank you

 Post subject: Re: convert gobooks embedded svgs to image
Post #2 Posted: Sat Jun 24, 2023 4:58 am 
Lives in gote

Posts: 598
Location: Adelaide, South Australia
Liked others: 211
Was liked: 267
Rank: Australian 2 dan
GD Posts: 200
Similar issues here. The images are "supported" on a lot of Android ebook readers, but not resizable, so full board diagrams don't fit on the screen.

The epub format is basically zipped HTML. In principle, it should be possible to unzip it, reformat a bit (add appropriate "size" tags to images, add navigation links, break things up so that problems and answers don't appear on the same page) and either read it in a web browser or convert to PDF.

 Post subject: Re: convert gobooks embedded svgs to image
Post #3 Posted: Sat Dec 02, 2023 9:05 pm 
Lives in gote

Posts: 598
Location: Adelaide, South Australia
Liked others: 211
Was liked: 267
Rank: Australian 2 dan
GD Posts: 200
And here's a first attempt at python code to do just that. Offered here without warranty or support, not guaranteed to work on all epub files.

Paste the code below into a file called "" or similar (sorry, attachments don't seem to be working in this forum nowadays, just need to use old fashioned copy-paste), and run it from the command line. Of course you'll need Python installed on your computer, and need to know how to open a command line and run a program from there. You'll also need to install the "zipfile" and "bs4" python libraries.

It's only been tested on my Ubuntu box, but in theory it should work on Windows or Mac just the same. The output is a folder containing an "index.html" file that you can open in a web browser, and a bunch of other files with the actual book pages.

Interactivity is a bit hit and miss (seems to work for some web browsers but not others), but at least you can open the book and see all the diagrams on any device this way. I've added large amounts of whitespace so that you won't see problem diagrams and answers on the same page, at the cost of a bit of extra scrolling. Near the end, change the "padding-top" number from 1000 to something smaller if you don't like this.

# Convert an epub from to a set of HTML pages that are easy to navigate

import sys, os
import re

from zipfile import ZipFile
from bs4 import BeautifulSoup
import copy # bs won't let you insert the same tag in two places: need deep copies!

if (len(sys.argv) != 2):
  sys.exit("Usage: python filename.epub")
input_filename = sys.argv[1]

if not os.path.exists(input_filename):
  sys.exit(f'File {input_filename} does not exist')

if (len(input_filename)>5) & (input_filename[-5:] == ".epub"):
  outdir = os.path.basename(input_filename[:-5])
  sys.exit(f"Filename {input_filename} isn't of the form something.epub")

if os.path.exists(outdir):
  sys.exit(f"Output location {outdir} already exists")
print(f"Creating output in {outdir}")
with ZipFile(input_filename, 'r') as zObject:

manifest_file = outdir + "/OPS/book.opf"
with open(manifest_file) as fp:
  soup = BeautifulSoup(fp, 'html.parser')

manifest = soup.find('manifest')
itemlist = manifest.findAll('item')
# The manifest contains all pages, not necessarily in reading order, plus other assets.
# Each manifest item has an id and an href.  The href is the filename of the xhtml.

spine = soup.find('spine')
idlist = [i.attrs["idref"] for i in spine.findAll('itemref')]
# The spine contains the ids (but not filenames) of all pages in reading order

long_idlist = [i.attrs["id"] for i in itemlist]
long_hreflist = [i.attrs["href"] for i in itemlist]
id_to_href = dict(zip(long_idlist, long_hreflist))

def id_to_filename(pageid):
# This is *.html for the converted page, not *.xhtml for the original!
  oldname = id_to_href[pageid]
  return(oldname[:-5] + oldname[-4:])

def prev_page(pageid):
  n = idlist.index(pageid)
  return None if n==0 else idlist[n-1]

def next_page(pageid):
  n = idlist.index(pageid)
  return None if n==len(idlist)-1 else idlist[n+1]

indexfile = open(outdir+"/index.html", 'w')
for p in idlist:
  indexfile.write(f'<a href=OPS/{id_to_filename(p)}>{p}</a><br>\n')

def convert_page(pageid):
  infile = outdir+"/OPS/"+id_to_href[pageid]
  outfile = outdir+"/OPS/"+id_to_filename(pageid)
  with open(infile) as f:
    soup = BeautifulSoup(f, 'html.parser')

  nextid = next_page(pageid)
  if nextid is not None:
    nexttag = soup.new_tag("a", href=id_to_filename(nextid))
    soup.body.insert(0, copy.copy(nexttag))
    soup.body.insert(0, " -- ")
    soup.body.append(" -- ")

  uptag = soup.new_tag("a", href="../index.html")
  uptag.string = "Up"
  soup.body.insert(0, copy.copy(uptag))
  soup.body.insert(0, " -- ")
  soup.body.append(" -- ")

  previd = prev_page(pageid)
  if previd is not None:
    prevtag = soup.new_tag("a", href=id_to_filename(previd))
    soup.body.insert(0, copy.copy(prevtag))

  # Reduce images to half size and make them resizable
  for imgtag in soup.findAll("svg"):
    boxstring = imgtag.attrs["viewbox"] # viewBox in original HTML, but bs converts to lower case and is case sensitive!
    # boxstring should be of the form "0 0 m n" where m and n are whole numbers
    x, y = boxstring.split()[2:]
    xhalf = int(x)*2 # to make the image half size, you need to double the viewbox size
    yhalf = int(y)*2
    imgtag["viewbox"] = f'0 0 {xhalf} {yhalf}'
    imgtag["width"] = str(xhalf)
    imgtag["height"] = str(yhalf)

  # Fix internal hyperlinks to point to the edited pages not the original
  for reftag in soup.findAll("a"):
    reftag["href"] = reftag["href"].replace(".xhtml", ".html")

  with open(outfile, 'w') as f:

  # Clean up unwanted old files

for p in idlist:

cssfile = outdir+"/OPS/css/style.css"
with open(cssfile, 'r') as f:
  css =
css = re.sub('^p {.*',
             'p { line-height: 1.3em !important; width: 50%; margin: 0 auto; }',
             css, flags=re.MULTILINE) # need flags so that ^ will match start of line
css = re.sub('page-break-before: always', 'padding-top: 1000px', css)
with open(cssfile, 'w') as f:

 Post subject: Re: convert gobooks embedded svgs to image
Post #4 Posted: Fri Dec 29, 2023 1:43 pm 
User avatar

Posts: 13
Liked others: 5
Was liked: 1
The author of announced, that from now on one can get the so called static version of purchased earlier interactive ebooks.
Paste your email here to get the static ones:

Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 4 posts ] 

All times are UTC - 8 hours [ DST ]

Who is online

Users browsing this forum: Google [Bot] and 1 guest

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group