[Tex2002ans]

Originally Posted by graycyn

<p><snip text>that young person answered respectfully,⁠⁠⁠—</p>

After the comma, your code accidentally has 3 WORD JOINERS in it.

Cut that down to 1 (or 0).

This is one of the dangers of using some of these invisible Unicode characters:

• It becomes very tough to see when you've made errors
• You may accidentally cause MORE bugs/harm than the problem you're trying to solve.

Originally Posted by graycyn

I'd prefer that em-dashes don't break off and end up on a line by themselves, because it could be distracting.

I'd still stick with the advice I've been saying for years:

In ebooks, don't worry about quirks in EM DASH linebreaks.

Leave it up to each device's/program's linebreak algorithms.

Originally Posted by graycyn

I am using the Word Joiner HTML entity ( & # 8 2 8 8 ; ) between the comma and the em-dash. I'm assuming that there is no need for the word joiner between the word and the comma?

No need between word and comma.

Originally Posted by graycyn

[...] Will all that hold together with *bags* if the joiner is used between the right quote and the em-dash or do I need to have the word joiner between the word AND each punctuation mark?

<p><snip text>mumbling something about “bothering old bags,”—</p>

or this:

<p>after the other—”</p>

Do I need the word joiner before AND after the em-dash?

Again, I wouldn't worry about this word joiner stuff.

But, if you had to pull my teeth, this is where WORD JOINER would be placed around the EM DASH:

• Example #1: BEFORE
• Example #2: BEFORE + AFTER

* * *

Side Note: Similar issue happens with odd linebreaking around the ELLIPSIS character + other punctuation.

For more info on that, see my posts in:

• 2021: "How to search for ellipses"

and all the linked threads which go into extreme detail.

Side Note #2: Most devices handle the ELLIPSIS character properly though, where the EM DASH is much trickier, because many languages use them as "Quotation Dashes":

• 2019: "No break space and alignment"

You also have languages that use every single possible combination of quotation marks... so you can't just say EM DASH + RIGHT DOUBLE QUOTE stick together 100% of the time.

Examples of EM DASHES in the middle of text are common, like:

• The man—“Tiny”—was the size of a mountain. (English)
• The man—”Tiny”—was the size of a mountain. (Finnish)

so if you "fixed" linebreaking for dialogue, you'd "break it" for parentheticals!

(Not to mention, the linebreaking algorithms should/would have to be context- + language-aware!)

Originally Posted by JSWolf

Does a printed copy of the 1910 edition do the same thing?

Yes, the printed copy has that. See Page 99 in this Archive.org version where graycyn's first example was.

Originally Posted by JSWolf

Is it correct to have the em-dash like it is ion the eBook?

Looks like every piece of Dialogue got its own new paragraph, so they split:

• Action/Beats
  
  • + EM DASH
• Dialogue

Example:

Ben shivered as if the mere thought hurt him, as he said, in a grim tone,—

“They might have cut Sanch's off.”

[...]

Ben looked discouraged, but Miss Celia cheered him up again by saying,—

“Yes, he could. My father had a friend who left a little dog in Paris; [...]”

Usage of that kind wouldn't be done in modern Style Guides, but back then they did lots of different things.

Nowadays, it would most likely be done like this:

Ben shivered as if the mere thought hurt him.

“They might have cut Sanch's off,” Ben said in a grim tone.

[...]

Ben looked discouraged, but Miss Celia cheered him up again.

“Yes, he could,” Miss Celia said. “My father had a friend who left a little dog in Paris; [...]”

or:

“They might have cut Sanch's off,” Ben grimly said, shivering as if the mere thought hurt him.

[...]

Ben looked discouraged, but Miss Celia cheered him up again.

“Yes, he could,” Miss Celia said. “My father had a friend who left a little dog in Paris; [...]”

or:

Ben shivered as if the mere thought hurt him. “They might have cut Sanch's off.”

[...]