ePub to Mobi Conversion in Calibre Problem

[FredUruncle]

I've been using Calibre for years and never encountered any real conversion problems. However, lately, I've tried to convert some ePub books to Mobi to put them on my Kindle PaperWhite and, while they seem to convert just fine as far as Calibre is concerned when I view the Mobi file either in Calibre or on my Kindle all I see are rows and rows of dots.

I also tried to convert the ePub to just text or just PDF with the same result. Is there something special about the ePub file that would break Calibre for all conversions?

Thanks!

[kovidgoyal]

https://www.mobileread.com/forums/sho...d.php?t=186697

[FredUruncle]

Kovid,

Thanks for the quick reply. I was hoping that the "rows of dots" would be enough to indicate something silly I was doing. I've actually stumbled across a dozen or more ePubs where this happens lately. I'll try and give you more complete information when I get time to collect it all.

[FredUruncle]

Originally Posted by kovidgoyal

https://www.mobileread.com/forums/sh...d.php?t=186697

Kovid,

I guess I'll finally have to try and get to the bottom of this. Today I bought a kindle book from Amazon. I used the Alf tools in Calibre to strip the DRM. Why? Well, I can obviously read the book on my Kindle Paperwhite but my eyes are not so good and I often convert my .azw files to ePub so I can load them into the wonderful IOS app "Voice Dream" which does a wonderful job of reading the books to me and also has many neat tools to focus my attention as I read the words while they are being read to me.

Anyway, as before, I'm having a terrible time converting things in Calibre. While Calibre says it has converted the book and the Table to Contents is all there and readable, the actual text body, regardless of output format looks like this....

......................................
......................................
......................................
......................................

etc., etc. The resultant file, regardless of output format, looks this way in the Mac Kindle reader, Calibre Viewer, on on the Kindle itself if converted to .mobi. I've had this problem for a while now with downloaded .azw files.

Note that the book I bought today converted perfectly on one of those online converter sites. I've also used the EPubsoft converter trial app and it has solved this problem in the past.

Something is apparently very wrong in my Calibre setup. I've tried the debug file creation but reading it doesn't help me much.....no skill here. I'll try and upload something useful.

Thanks for reading the diatribe...

[rashkae]

The only way I've found so far to replicate this effect is if I add a Search and Replace filter with a . in each field.

[theducks]

Could that be a line-height: 1 (or smaller number) stacking the <p>lines on top of each other?

[pdurrant]

When a DRMed Kindle file is imported to calibre (when Alf's tools installed), the file type should change to .mobi, .azw3 or .htmlz.

If it doesn't, the DRM has not been removed, and you need to get help at Alf's blog, or Harper's github repository.

[FredUruncle]

Originally Posted by rashkae

The only way I've found so far to replicate this effect is if I add a Search and Replace filter with a . in each field.

God Bless you

When I first read your reply, I thought to myself that I had never done any search and replace and I even looked at that TAB in the conversion window to make sure. I didn't see it so I moved on. I redid the book several times using the BUG option. I was able to find that the input RAW-HTML result was fine but the parsed data was full of dots. On a lark I used the metadata.opf as the input to the conversion. When I tried to convert this to an EPUB it failed but the error report gave the changes from default values and low and behold, it said I was trying to search and replace a "." with ". " I went back to the Search and Replace page and there it was. Do you know how hard it is to even notice that you have two periods in the search/replace boxes?

I didn't realize that these definitions were persistent across all formats forever. Now, I recall that the Dream Voice app I mentioned in an earlier post was having trouble reading a file and basically, that file had no space after the period at the end of each sentence. I used this search and replace and fixed that file and then promptly forgot about it. It sat there causing me grief for months.

All is well now.... The only remaining confusion is that when, on the search and replace tab, I hit "Restore Defaults" its putting periods there. At any rate, I know how to make it work and I'll figure that out as I need to ....

[FredUruncle]

Originally Posted by theducks

Could that be a line-height: 1 (or smaller number) stacking the <p>lines on top of each other?

Thanks for the reply. I've figured it how with help from you folks. Fonts and line heights were my first thought also, but they were default. It was and errant search and replace that was killing me!

[FredUruncle]

Originally Posted by pdurrant

When a DRMed Kindle file is imported to calibre (when Alf's tools installed), the file type should change to .mobi, .azw3 or .htmlz.

If it doesn't, the DRM has not been removed, and you need to get help at Alf's blog, or Harper's github repository.

Thanks for chiming in. The DRM was indeed being removed so that wasn't the issue this time. I was able to convert the file without error but it was just all periods. I made the mistake of assuming this was a formatting/font issue defaulting to periods on view when it was a really a search/replace issue that replaced every character with an actual ".". Anyway, I"m good....

Thanks again....

[rashkae]

Originally Posted by FredUruncle

All is well now.... The only remaining confusion is that when, on the search and replace tab, I hit "Restore Defaults" its putting periods there. At any rate, I know how to make it work and I'll figure that out as I need to ....

You have to go into preferences -> Common Options, From there, you will change the *default* Search and Replace.

As a further note, the search portion of Search and replace is a Regex expression. In Regex, . is a symbol that means *any* character. If you really wanted to search replace every . in a conversion, you would search for \. . (The backlash escapes the special meaning of characters, so \. means to search for . literally.)

[FredUruncle]

Originally Posted by rashkae

You have to go into preferences -> Common Options, From there, you will change the *default* Search and Replace.

As a further note, the search portion of Search and replace is a Regex expression. In Regex, . is a symbol that means *any* character. If you really wanted to search replace every . in a conversion, you would search for \. . (The backlash escapes the special meaning of characters, so \. means to search for . literally.)

Yes, I figured the former out shortly after posting the last time. The further gotcha is that Calibre seems to remember what you used the first time on each book in the library, so when reconverting all the "rows of dots books" I have to remember to hit "Restore to Defaults" for each book to get it to remove the replace rule each time. Not a biggie. Once I've re-done the bad conversations, new books will be OK from the beginning.

And, thanks for the Regex tip about dealing with .'s. I think I looked that up when I was doing it that one time that messed me up but I had forgotten it so the reminder is welcome.

Thanks All of the tips....