Calibre conversion issue

[adam1234]

Hi,

I'm trying to convert a short story for reading on my Kindle. The story is a web page. Saving the web page as HTML and loading it into Calibre usually works.

When I save https://www.tor.com/2017/07/19/the-martian-obelisk/ (save as web page, html only, Firefox, mac), and add it as a book, then try to open it by double clicking, I get an error:

"Unable to expand "the martian Obelisk...Unknown.zip" It is in an unsupported format"

But the file is not a zip, it's HTML (with an inordinately long header block, but still, clean html that I can read with an editor, and I tried emptying the entire header block from the local archive.)

So I think there's a bug in perhaps the format recognizer code?

Either way, I'd appreciate advice on working around the issue.

[Quoth]

I copy/paste into LO Writer and fix formatting etc.
In this case I'd paste unformatted text, add image, save odt and save as docx for Calibre after sorting styles.

HTML from real web sites is unreliable.

[JSWolf]

Originally Posted by adam1234

Hi,

I'm trying to convert a short story for reading on my Kindle. The story is a web page. Saving the web page as HTML and loading it into Calibre usually works.

When I save https://www.tor.com/2017/07/19/the-martian-obelisk/ (save as web page, html only, Firefox, mac), and add it as a book, then try to open it by double clicking, I get an error:

"Unable to expand "the martian Obelisk...Unknown.zip" It is in an unsupported format"

But the file is not a zip, it's HTML (with an inordinately long header block, but still, clean html that I can read with an editor, and I tried emptying the entire header block from the local archive.)

So I think there's a bug in perhaps the format recognizer code?

Either way, I'd appreciate advice on working around the issue.

I'm using calibre 7.3. I saved as web page in Firefox. I then loaded the HMTL file as is. I then copied the image and pasted it as the cover. I converted it to ePub. Worked no problem.

I am using Widows 11 if that helps.

[adam1234]

Originally Posted by JSWolf

I'm using calibre 7.3. I saved as web page in Firefox. I then loaded the HMTL file as is. I then copied the image and pasted it as the cover. I converted it to ePub. Worked no problem.

I am using Widows 11 if that helps.

Did you save as a simple or complete page? Did you load the HTML as a book? (I used one book at a time)

Asking to help narrow it down.

[adam1234]

Originally Posted by Quoth

I copy/paste into LO Writer and fix formatting etc.
In this case I'd paste unformatted text, add image, save odt and save as docx for Calibre after sorting styles.

HTML from real web sites is unreliable.

Thanks - this is the first time I've ever had a problem, and the thing that caused me to create a thread was the "zip". The addition of a .zip to a filename that's not a zip seems like a bug.

We agree html can be all over the place. What I saw (after I deleted all the cruft in the <header> block) seemed like reasonable and simple html, but I didn't read down to see how much cruft came after the body.

[Karellen]

Originally Posted by adam1234

Hi,

I'm trying to convert a short story for reading on my Kindle.

Looks like an interesting short story, so I put it on my "to read" list.
It only took less than 10 minutes to create the epub. Most of that was fiddling with the cover. So try this...

[JSWolf]

Originally Posted by adam1234

Did you save as a simple or complete page? Did you load the HTML as a book? (I used one book at a time)

Asking to help narrow it down.

I did a Save page as and code Web page HTML only. I then dragged/dropped the HTML file into the calibre library window. I then selected it and hit E on the keyboard. I put the graphic in the story as the cover. Clicked OK and then converted to ePub.

[DNSB]

Originally Posted by adam1234

But the file is not a zip, it's HTML (with an inordinately long header block, but still, clean html that I can read with an editor, and I tried emptying the entire header block from the local archive.)

So I think there's a bug in perhaps the format recognizer code?

Either way, I'd appreciate advice on working around the issue.

For what it's worth, when calibre adds a saved web page, it will save the various bits and pieces in a .zip file. If you open the .zip file in an archiver, you will see a .htm file, a content.opf file, a mimetype file and a META-INF file. Oddly similar to what you see when opening an ePub file.

I did a quick copy and paste of the document and spent a minute doing cleanup to generate an ePub.

[adam1234]

Originally Posted by JSWolf

I did a Save page as and code Web page HTML only. I then dragged/dropped the HTML file into the calibre library window. I then selected it and hit E on the keyboard. I put the graphic in the story as the cover. Clicked OK and then converted to ePub.

Huh. I could swear I had converted files without the E step in the past but thank you!

[adam1234]

Thanks to @DNSB and @Karellen for doing the conversions for me!