ePub validation in epubcheck 1.1

[islandcharlie]

I have saved a Word 2011 (Mac) document as HTML, then converted with calibre to ePub. Subsequent validation check with epubcheck 1.1, supposedly required for submission to Apple's iBookstore, results in a report of a huge number of errors. What is the likely source? Is it a Word formatting issue? a word to html issue? I'm baffled. Help, please.

[JSWolf]

It is because Word puts in all kinds of garbage when you save as HTML. Even filtered HTML is not nice. You'll have to open the ePub and have a go at removing the mess that is Word.

[islandcharlie]

Thanks. But with over 400 error messages, that seems an almost insurmountable task. Is there another word processing program or environment in which this kind of thing might not occur? I can use Pages on the Mac which might be cleaner, I suppose. I'd welcome any ideas. Thanks again.

[Manichean]

Try saving as RTF in Word and converting that.

[JSWolf]

Originally Posted by islandcharlie

Thanks. But with over 400 error messages, that seems an almost insurmountable task. Is there another word processing program or environment in which this kind of thing might not occur? I can use Pages on the Mac which might be cleaner, I suppose. I'd welcome any ideas. Thanks again.

Did you save it as filtered HTML from Word? If not, try that and then convert. Then clean up the messy code you get. It can be done. I've done it before.

[theducks]

Originally Posted by islandcharlie

Thanks. But with over 400 error messages, that seems an almost insurmountable task. Is there another word processing program or environment in which this kind of thing might not occur? I can use Pages on the Mac which might be cleaner, I suppose. I'd welcome any ideas. Thanks again.

With 400 messages, many are going to be the same type of error.
learn to fix globally

Use Sigil (and Flightcrew that is built in).

Common faults are failures to conform to the rules. eg id=##
in Codeview
REGEX Search: id="(\d+)"
replace: id="b_\1" <- any letters are good
once you validat one fix meets you eyeball chech, switch to
All HTML files and replace all.
Always keep a prior version backup for those moments.

[JSWolf]

For me personally, I prefer using notepad++ and FlightCrew. I do need to update my copy of Sigil at some point. One major advantage of FlightCrew vs. ePubCheck is that FC's error messages are more human readable.

[Dopedangel]

try using libre office for doc to html conversion the html created by that is a lot cleaner than Microsoft office

[theducks]

Originally Posted by Dopedangel

try using libre office for doc to html conversion the html created by that is a lot cleaner than Microsoft office

Use with the Writer2EPUB plugin for a very good conversion

[islandcharlie]

Thank you all. I tried moving my book to Pages first. It exported to ePub cleanly, but I lost all my links, fonts, and appearance factors. The OpenOffice worked but left all my errors in place from MS Word. Finally, working with Sigil, I was able to eliminate all the errors in about an hour of work. It seems that Word's HTML output includes many attributes not allowed in XHTML, which is used by ePub. Sigil with onboard Flightcrew made it easy to delete them, and the results are excellent. Thanks again.

[islandcharlie]

Much later.
I have had good success using Sigil with its built-in FlightCew but with one problem. Writing in Word for Mac 2011, then saving as HTML then converting with Calibre, then testing and editing in Sigil/Flightcrew almost always works. But sometimes , in Sigil, when I go to validate with FlightCrew I get an error message saying "An exception occurred during validation: std::exception" I have been unable to get an answer from anyone as to what this means. Does anyone know?. The validation always stops at this point and will not continue.

[Manichean]

There's a Sigil forum over here. You might have better luck asking there.

[DoctorOhh]

Originally Posted by Manichean

There's a Sigil forum over here. You might have better luck asking there.

......

[theducks]

Originally Posted by islandcharlie

Much later.
I have had good success using Sigil with its built-in FlightCew but with one problem. Writing in Word for Mac 2011, then saving as HTML then converting with Calibre, then testing and editing in Sigil/Flightcrew almost always works. But sometimes , in Sigil, when I go to validate with FlightCrew I get an error message saying "An exception occurred during validation: std::exception" I have been unable to get an answer from anyone as to what this means. Does anyone know?. The validation always stops at this point and will not continue.

Flightcrew is also available as a standalone.