Non-breaking space

[JSWolf]

I thought I had seen it all when it comes to bad practices making an eBook. Well, today I found something I have yet to (knowingly) come across. I've come across an eBook where there are non-breaking spaces in place of spaces that throws off the formatting and makes nice big gaps between words.

I kind of knew it had to be non-breaking spaces when the gaps were big enough to fit the word that started the next line without a problem. I then had a look at the XML file and yes, there are non-breaking spaces there.

So why would non-breaking spaces be put in an eBook where they have no need to be there? This is a publisher original and not some download some unknown source.

[Jellby]

Some crappy automatic conversion, maybe a layout-preserving OCR?

[JSWolf]

Originally Posted by Jellby

Some crappy automatic conversion, maybe a layout-preserving OCR?

No idea. But when I was reading, the gaps did show up very easily. More people making eBooks who have no clue how to do it. They also include covers that seems like they are scaled up in size or are very pixelated looking at least for the first two books in the series. This is the George Gently series by Alan Hunter. I've been watching the BBC TV series on PBS and finally decided to give the books a go.

[Tex2002ans]

Were these occurring towards the the end of paragraphs only? Or were the   just spread out willy nilly with no rhyme or reason?

A few months back in my LaTeX research, I stumbled upon this odd question about the "minimum length of the last line of a paragraph":

https://tex.stackexchange.com/questi...of-a-paragraph

There was also a similar idea for poetry for not wanting a single word on a line by itself, and trying to keep two or three words together if the line split:

<p>This is just a sample sentence that ends too&nbsp;soon.</p>

So that if the line does break, "too soon" would be pulled to the next line. (To some typographers that might look more pleasing than lines that are too short).

Certain languages/countries/style guides might have certain odd typographic rules as well. I stumbled upon this too:

One rule in French typography is that the last line of a paragraph should not be shorter than the double of the indentation of the next paragraph.

https://tex.stackexchange.com/questi...h-of-last-line

Could also just be some leftover typographic crud from InDesign/Quark output as well. The good ol' "we can just export this as EPUB because it has a 'Save As EPUB'!"

Does any metadata give any hints about the creation program? Anything you can deduce from the CSS/class names?

We need some examples and DETAILS!

[JSWolf]

The non-breaking spaces looked to be placed without any rhyme or reason.

Here are some samples. The bolding is mine.

<p id="rw-p_287847-00003">‘He went back to the fair after he’d been here,’&nbsp;continued the superintendent. ‘He had tea with his wife&nbsp;in his caravan and did his stunt at 6.15. He was due to do&nbsp;it again at 6.45. I had men there at 6.35, but he’d&nbsp;disappeared. The last person to see him was the mechanic&nbsp;who looks after the machines.’</p>

<p id="rw-p_287847-00010">‘Then for heaven’s sake why didn’t you grab him?’&nbsp;snapped Hansom.</p>

<p id="rw-p_287847-00012">Hansom snarled disgustedly. The superintendent&nbsp;brooded for a moment. ‘I don’t think there’s much doubt&nbsp;left that he’s our man,’ he said. ‘It looks as though we&nbsp;shan’t be needing you after all, Gently. I think we shall&nbsp;be able to pin something on young Huysmann and make&nbsp;it stick.’</p>

<p id="rw-p_287847-00021">‘You might print the door handle and the back of the&nbsp;chair that stands just inside,’ continued Gently, ‘and&nbsp;photograph the marks left on the carpet. Then again,’ he&nbsp;turned his thumb back with slow care, ‘you might&nbsp;wonder to yourself how the knife came to be in the chest&nbsp;in the hall. I can’t help you in the slightest. I’m still&nbsp;wondering myself …’</p>

[Jellby]

It looks like they appear at quite constant distances. I'd say that's where linebreaks occur in the printed version, and they were converted to  . A reason to never buy something from that publisher.

[WillAdams]

Probably the non-breaking spaces were inserted to influence how the line breaks --- I really wish there were a ``discretionary non-breaking space'' which would only be used to keep the preceding and following word from not hyphenating and encourage keeping them together (one can program that sort of thing in LaTeX, but it's not an option until some engineer at Adobe or Quark or Microsoft or Apple works it up elsewhere).

[AlexBell]

This may be off topic but for what it is worth I really don't like to see single ndashes or right quotation marks on a line by themselves; it just looks ugly to me. Most of the ebooks I do were written in the 19th century, and it was quite common to see something like:
'Lorem ipsum dolor sit - ' which might end up as
'Lorem ipsum dolor sit
- '
or (even worse) Lorem ipsum dolor sit -
'

I end to use Lorem ipsum dolor sit –’
which at worse displays as
Lorem ipsum dolor
sit –’

Any comments?

[JSWolf]

I would take

'Lorem ipsum dolor sit - '

and fix it to be

'Lorem ipsum dolor sit—'

. I would replace the ndash with an mdash, remove any spacing around the mdash (I don't like the spaces) and remove the space in from of the last apostrophe. Then it would break better. Also, I would not put in non-breaking spaces like that. It's not needed and can make things look bad.

[Tex2002ans]

Originally Posted by AlexBell

This may be off topic but for what it is worth I really don't like to see single ndashes or right quotation marks on a line by themselves; it just looks ugly to me. Most of the ebooks I do were written in the 19th century, and it was quite common to see something like:

[...]

Any comments?

Pick a Style Guide that says they should be Set Closed:

https://en.wikipedia.org/wiki/En_das...d_substitution

But, if you still want to keep the Set Open (as some languages require), the   is probably the best way to handle it in EPUB since the more complex spaces (hair space, thin space, etc. etc.) are not supported very well on devices.

I don't have any Regex on hand to handle all the situations (since I Set Close all em/en dashes), but it shouldn't be too hard to come up with a bunch of Regex to substitute spacing around dashes to Set Open.

I use this to Regex to Set Close all Em Dashes:

Search: [ ]*—[ ]*
Replace: —

I am also not too sure if it would be valid to only have   only BEFORE the Set Open en dash, and not BEFORE+AFTER (this will make sure the device doesn't see both words combined into one very large word, and mangle the justification algorithm).

Same sort of thing happens in certain languages with spacing before/after other punctuation. I was converting a French Canadian book once, and the typographic rules made my head hurt, and created ugly code.

Here is a paragraph from the French Canadian book:

<p>Pour comprendre le rôle de la monnaie, nous devons commencer par le commencement, et nous demander&nbsp;: après tout, pourquoi les hommes échangent-ils&nbsp;? L’échange est le fondement de notre vie économique. Sans échanges, aucune économie ne peut réellement se développer, la société même est impossible. À l’évidence, si un échange a lieu volontairement, c’est que les deux parties espèrent en bénéficier. Un échange, c’est un accord entre A et B pour céder les biens et services de l’un contre les biens et services de l’autre. Les deux en profitent parce que chacun accorde plus de valeur à ce qu’il reçoit qu’à ce qu’il cède. Par exemple, lorsque Crusoé échange du poisson contre du bois, il accorde plus de valeur au bois qu’il «&nbsp;achète&nbsp;» qu’au poisson qu’il «&nbsp;vend&nbsp;». Pour Vendredi c’est le contraire&nbsp;: il accorde plus de valeur au poisson qu’au bois. Mais depuis Aristote jusqu’à Marx, les hommes ont cru à tort que l’échange correspondait à une sorte d’identité de valeurs – que si une barrique de poisson était échangée contre dix rondins de bois, cela voulait dire qu’implicitement il y avait <em>égalité</em> entre les deux. Mais c’est le contraire. L’échange a lieu justement parce que chaque partie, <a id="page_14"></a>parmi ses préférences, range les deux produits dans un ordre <em>différent</em>.</p>

As you can see, they even use &nbsp; before colons, and question marks.

Side Note: Alex, I sent you an email a few months back, never got a response, did you change your email or something? Maybe it got lost in the spam folder, I did attach A TON of stuff to get your input/ideas on.

[Jellby]

Originally Posted by AlexBell

I end to use Lorem ipsum dolor sit&nbsp;&ndash;&rsquo;

In English books, I use " &ndash; " in the middle of a sentence and "&mdash;" (no spaces) at the beginning or end:

Bla &ndash; foo &ndash; bar, &lsquo;Nevermore&mdash;&rsquo; quoth the raven.

An alternate style would be:

" &ndash; " -> "&mdash;"
"&mdash" -> "&mdash;&mdash;"

Originally Posted by Tex2002ans

As you can see, they even use &nbsp; before colons, and question marks.

Yes, every "double" sign. Some crappy magazines don't even use non-breaking spaces, and you find the question mark alone at the beginning of a line!

For best effect, use & #8239; (narrow no-break space).

[mrmikel]

This style is very common in English books printed about 100 years ago, but not common in American books. Makes a bit of a quandary when converting them.

[JSWolf]

Originally Posted by mrmikel

This style is very common in English books printed about 100 years ago, but not common in American books. Makes a bit of a quandary when converting them.

I'm of the no spaces around em-dash and no spaces around ellipses. It just looks off to have the spaces. Also, as for the em-dash vs. the n-dash, I don't care for the n-dash and prefer to go with the em-dash.

[Hitch]

Originally Posted by AlexBell

This may be off topic but for what it is worth I really don't like to see single ndashes or right quotation marks on a line by themselves; it just looks ugly to me. Most of the ebooks I do were written in the 19th century, and it was quite common to see something like:
'Lorem ipsum dolor sit - ' which might end up as
'Lorem ipsum dolor sit
- '
or (even worse) Lorem ipsum dolor sit -
'

I end to use Lorem ipsum dolor sit&nbsp;&ndash;&rsquo;
which at worse displays as
Lorem ipsum dolor
sit&nbsp;&ndash;&rsquo;

Any comments?

Yes. It doesn't fix it in ADE or several other readers, anyway, as ADE, et al, will break bloody closing quotation marks away from the end of a line, even if they are after a period, emdash, etc. Using the nbsp doesn't solve that issue, and can cause more, like the George Gently discussion, because using them creates faux-words where they are least needed. For every place that an nbsp magically works to prevent the emdash, et al, from breaking away, it will cause another reader to have massive rivers of white because it does work in other locations.

Personally, I think the best bet is to join the punctuation (emdashes, endashes and ellipses being the big offenders) to the prior word, and leave a space after. This gives you a "split the baby" approach. I find an emdash starting a new line, or an ellipsis, far odder than ending one. FWIW; everyone has their own opinions on this topic.

Hitch

[AlexBell]

Originally Posted by Tex2002ans

Side Note: Alex, I sent you an email a few months back, never got a response, did you change your email or something? Maybe it got lost in the spam folder, I did attach A TON of stuff to get your input/ideas on.

Thanks for the suggestions.

And no, I have no recollection of getting a long email from anyone relating to ebooks. I'm sure I would have at least acknowledged it even if I didn't deal with it in detail. Apologies anyway, and could you send it again? And let me know by private mail that it's been sent?