[GUI Plugin] Hyphenate This!

[Ruskie_it]

Originally Posted by silasg

But besides of that I wonder if ebooks bought at Amazon are not already hyphenized?! I really would expect that, if their reading devices need this to present a hyphened view of the text.

I'd expect that too, it simple is "good quality", but I can tell you they aren't. This is not such a big issue when you're font size is small enough to have lots of words in a row, but it shows from time to time.
However, if I am not mistaking, this does not depend on Amazon but on the publishers. Amazon is not "responsible" for the content of ebooks they sell, they do not tamper with them.

[Ruskie_it]

Hello, I noticed something very weird yesterday, and after a few test I managed to connect it directly to Hypenate This!
I have one book that I correct on Calibre while I am reading on my Kindle. I have noticed that when the book is soft-hypenated the search function doesn't work: I got "no match found" even when the word is right before my eyes. This happens both with Sigil and with the Calibre internal ebook editor. If I remove soft-hypenation with the plugin everything works. Then again if I re-hypenate it, search stops working again.
Is it normal? If not, how can this be avoided?

[shamanNS]

Well Amazon is responsible for failing (7 years and counting) to include "device level" hyphenation on eInk Kindles at least for english and couple of others which also have official dictionaries and UI translation [I'm certainly not holding my breath waiting to ever see hyphenation support for serbian language], especially when there is auto-hyphenation in Kindle for Android app (only in portrait orientation if I remember correctly?) which is written in Java just like the majority of eInk Kindle GUI, and I think Kindle for PC all has/had it at some point.

[NiLuJe]

@Ruskie_it: On which device/FW/language is that? I've never had any issues with search in soft-hyphenated English or French texts. Might be another story for non-latin scripts, though.

[shamanNS]

@NiLuJe: I think/guess that he is having search problem on PC side of things? That both Sigil and Calibre can't find the word typed into the Search box because actual word is broken-down by soft hyphens.
Kindle itself usually is smart enough to successfully search for other instances of the same word...

[JSWolf]

The solution is to use the plugin to remove the soft-hyphens and use the eBook on the PC side and when you are ready to send it to the Kindle, then include the hyphens. But don't hyphenate say an ePub to be converted to a Kindle format. Convert the ePub and then hyphenation the Kindle eBook.

[NiLuJe]

@shamanNS: Oh, right, thanks ;D. In which case, no idea, but that might be a good thing to report to both projects if it's confirmed .

[eschwartz]

Originally Posted by NiLuJe

@shamanNS: Oh, right, thanks ;D. In which case, no idea, but that might be a good thing to report to both projects if it's confirmed .

In an editor that is expected behavior. Regexes should not ignore invisible unicode chars.

[NiLuJe]

@eschwartz: Very good point. I (wrongly?) assumed that we were talking about the Calibre Reader, where this might be an issue, but for Sigil & the Calibre Editor, yup, very good point indeed .

[eschwartz]

I was privy to the other thread, in the editor subforum.

[mykey]

Can I use this plugin in CLI (command line)? I want to make it part of the auto converting ebooks workflow.

[imaginer]

Unfortunately, I could not register the bugreport page.

after hungarian hypen, using Hyphenate This!
correct in Calibre v.2.10: "ál*lat*kert a poggyá*szom*ban volt"
incorrect in Calibre v.2.12: "ál*lat*kert a pogy*yá*szom*ban volt"

(When we encounter two such letters consecutivly, we double only the first glyph. For example, busz + -szal = busszal.
The exception to this rule is when we have to wrap the word over two lines. When then write each letter in full:
Megyek busz-
-szal minden nap.
From:http://www.hungarianreference.com/Hu...unciation.aspx)

[Ruskie_it]

Sorry I forgot to check this thread back. Yes, Shaman was right, this was debated in another thread.

[silasg]

Originally Posted by Ruskie_it

I'd expect that too, it simple is "good quality", but I can tell you they aren't. This is not such a big issue when you're font size is small enough to have lots of words in a row, but it shows from time to time.
However, if I am not mistaking, this does not depend on Amazon but on the publishers. Amazon is not "responsible" for the content of ebooks they sell, they do not tamper with them.

Yes ok, I agree that this should fall into the responsibility of the publisher. But well, on the other hand: Amazon has such a big influence (file format, prices and so on...), why not using this to ensure some quality? I read a lot of German texts and, you know, we like joining words to produce quite long funny compounds like "Donaudampfschifffahrtskapitän" (ok ok, this _is_ a quite artificial example but you get the point). So it is good to know that I have to preprocess my ebooks in any case after buying it, since in that case I might consider buying them somewhere else, if there are some other reasons for that.

[Ruskie_it]

Originally Posted by silasg

Yes ok, I agree that this should fall into the responsibility of the publisher. But well, on the other hand: Amazon has such a big influence (file format, prices and so on...), why not using this to ensure some quality? I read a lot of German texts and, you know, we like joining words to produce quite long funny compounds like "Donaudampfschifffahrtskapitän" (ok ok, this _is_ a quite artificial example but you get the point). So it is good to know that I have to preprocess my ebooks in any case after buying it, since in that case I might consider buying them somewhere else, if there are some other reasons for that.

I guess thus is because to extend their sells they value much more not to loose potential publishers (relatively only a few) rather than not to loose a few readers because of the hyphenation, which normally is only an aesthetic problem. Makes sense, if not many readers start complaining.
Rather than asking myself why they don't force publishers to pre-hypenate their books, I wonder why they don't implement an algorithm that is able to handle "hardware" hyphenation, since this is not rocket science and almost every reader has it...