ePub to HTML links all broke

[perry59]

I am trying to use calibre to convert some ePubs to html. Calibre does the conversion but when I open the resulting html file all the links are kinda goofed (they are pointing to web addresses rather than the content on my hard drive). Also, no pictures appear. Calibre creates the images subdirectory and puts the pictures in there but their names do not match the codes in the html file.
Am I doing something wrong?

The reason I am doing this is because, while I love Calibre's management and conversion capabilities I want to use Blio as my reader. I just love its interface. The rest of Blio is not so good (and their support sucks). It only works with .xps files, if it would read ePubs I wouldnt have this problem. So I am trying to convert ePubs to html which I can then open in MS word and save out as xps files for Blio.
Thanks for any tips

[user_none]

What format are you using for HTML output (ZIP or HTMLZ)?

Originally Posted by perry59

they are pointing to web addresses rather than the content on my hard drive.

This is how your input is already. External links are left as is because it's not pointing to the document to begin with.

Originally Posted by perry59

Also, no pictures appear. Calibre creates the images subdirectory and puts the pictures in there but their names do not match the codes in the html file.

My guess is they are pointing to the website just like your links.

Originally Posted by perry59

Am I doing something wrong?

Probably not. It sounds like a case of garbage in garbage out.

[perry59]

I didnt realize there was a zip output option. I output to htmlz then renamed it a zip file so I could access the contents. Yes the picture links were pointing to the website to (gutenberg) but the links are really weird with a lot of "@" in them. When I get home I'll compare the htmlz output to what I find by just renaming the epub and looking at its internal html. I would just use the content of the epubs but they typically break a book into many html files which would be a pain to stitch together.
Thanks

[user_none]

@ is not a character conversion should be inserting in links. It sounds like there is an issue with the book you are converting.

[dwig]

Originally Posted by user_none

@ is not a character conversion should be inserting in links. It sounds like there is an issue with the book you are converting.

I've seen "@" symbols in image filenames in ePubs from PG many times. It seems that they convert their web server's full pathname to the image file into a filename and replace the illegal "/" symbols with ampersands. The result is a very long filename beginning with "www" and including 4-8 ampersands. One image in a book I recently reformated is named "www.gutenberg.org@dirs@3@2@1@0@32104@32104-h@images@image_f.jpg" in the ePub.

[user_none]

Originally Posted by dwig

I've seen "@" symbols in image filenames in ePubs from PG many times.

That's my point. An @ symbol will not be created by calibre as it rewrites links in a consistant manner. The only way for a link to contain an @ symbol is if the original link is an external link that includes it. HTMLZ's rewriting of links does not touch external links.