On Sat, Jan 30, 1999 at 12:18:36AM -0500, Larry S. Marso wrote:
> I received an answer to my question (thank you!) about the right latex for
> an period *not* at the end of a sentance.
>
> I don't understand why we have "end of sentance period" built into the
> Insert -> Special Character menu. We get that by typing "." don't we?
> Shouldn't the special character instead be "not end of sentance period"
> (described more eloquently than this, of course) -- which would correspond
> to a ".\ " in raw LaTeX? Any document with lots of "Mr." notations, names
> including "Inc." (sorry, we don't use Aktiengesselchaft here) needs this
> function. Hard spaces mess up line breaks badly.
Yes, we do need the "abbreviation period" IMO. However, your question
points out that we also need the "end of sentence period" --- can you
name the esoteric LaTeX code that forces inter-sentence space after a
period? Didn't think so. Again, for some odd reason, no one wanted
to add one of these.
Now that I write C++ for a living, I'll take a gander at adding this
to 1.0.1/1.1 at some point.
What follows is only a suggestion, so take it with a grain of NaCl.
After I finally grasped when LaTeX uses what sort of spacing after a
".", the algorithm struck me as odd. After all, it's not hard to look
at the next non-whitespace character after a "." and see if it's
uppercase. If so, that "." ends a sentence. If not, it's an
abbreviation. That kind of algorithm would be fairly easy to add to
the LyX LaTeX export. On re-importing, we'd have to put in the
distinction between abbreviation and sentence-ending periods that we
wrote to the LaTeX file. Yes, the resulting LyX doc would look like
an xmas tree with all of the multicolored dots, but that's the price
of import.
There are several ways we can implement "intelligent dots". Here are
some of my ideas:
Algorithm #0: Existing LaTeX
Not a suggestion per-se, but a statement of what we get right now.
Inter-Word Spacing:
"." follows any uppercase *letter*
Inter-Sentence Spacing:
Anywhere else
Customization Needed:
[x] Latinate Abbreviations: "e. g. ", "i. e. "
[x] Title Abbreviations: "Mr. Jones", "Dr. Smith", "Mrs. Peale"
[x] Name Descriptors: "Richard Hawkins, Esq. knows [...]"
[x] Sentence Ending in a Capitalized Word:
"[...] on Fiji. If you are [...]"
[x] Sentence Ending in a Capitalized Letter:
"[...] at NASA. Experts say [...]"
====================================================================
The remainder are things LyX would forcibly do "i. e. add the "@"
or "\ " upon LaTeX output. Each of the suggestions is
semi-orthogonal; I'm not considering combinations just yet.
====================================================================
Algorithm #1:
Inter-Word Spacing:
Anyplace except...
Inter-Sentence Spacing:
"." followed by any whitespace followed by an uppercase
letter.
Customization Needed:
[ ] Latinate Abbreviations: "e. g. ", "i. e. "
[x] Title Abbreviations: "Mr. Jones", "Dr. Smith", "Mrs. Peale"
[ ] Name Descriptors: "Richard Hawkins, Esq. knows [...]"
[ ] Sentence Ending in a Capitalized Word:
"[...] on Fiji. If you are [...]"
[ ] Sentence Ending in a Capitalized Letter:
"[...] at NASA. Experts say [...]"
Algorithm #2:
Inter-Word Spacing:
"." follows a capitalized *word*
Inter-Sentence Spacing:
Anywhere else
Customization Needed:
[x] Latinate Abbreviations: "e. g. ", "i. e. "
[ ] Title Abbreviations: "Mr. Jones", "Dr. Smith", "Mrs. Peale"
[ ] Name Descriptors: "Richard Hawkins, Esq. knows [...]"
[x] Sentence Ending in a Capitalized Word:
"[...] on Fiji. If you are [...]"
[x] Sentence Ending in a Capitalized Letter:
"[...] at NASA. Experts say [...]"
Algorithm #3:
Inter-Word Spacing:
Anyplace except...
Inter-Sentence Spacing:
"." follows a lowercase word
Customization Needed: Same as #3
Algorithm #4:
Inter-Word Spacing:
"." follows a "word" that is only 1 character long.
Inter-Sentence Spacing:
Anywhere else
Customization Needed:
[ ] Latinate Abbreviations: "e. g. ", "i. e. "
[x] Title Abbreviations: "Mr. Jones", "Dr. Smith", "Mrs. Peale"
[x] Name Descriptors: "Richard Hawkins, Esq. knows [...]"
[ ] Sentence Ending in a Capitalized Word:
"[...] on Fiji. If you are [...]"
[ ] Sentence Ending in a Capitalized Letter:
"[...] at NASA. Experts say [...]"
Algorithm #5:
Inter-Word Spacing:
"." follows any word that is shorter than 4 characters.
Inter-Sentence Spacing:
Anywhere else
Customization Needed:
[ ] Latinate Abbreviations: "e. g. ", "i. e. "
[ ] Title Abbreviations: "Mr. Jones", "Dr. Smith", "Mrs. Peale"
[ ] Name Descriptors: "Richard Hawkins, Esq. knows [...]"
[ ] Sentence Ending in a Capitalized Word:
"[...] on Fiji. If you are [...]"
[ ] Sentence Ending in a Capitalized Letter:
"[...] at NASA. Experts say [...]"
[x] Sentence Ending in a Short Word:
"[...] the place we went to. I thought [...]"
#1 is probably the easiest to implement, and would also execute the
fastest. Its downside is that it forces customization for the most
common form of abbreviation: titles on names. #4 has the same
drawback. #5 has a different fatal flaw: it will fail on some pretty
common sentences. #2 doesn't fail on name titles, but fails on many
other cases. We'd want this "magic period" to work properly with the
sorts of abbreviations and sentences that people use most often. None
of the ideas I've had fit that description perfectly. So, it boils
down to where do we want to force the user to add a special period?
Thoughts? Comments?
--
John Weiss
