bug#20385: [PATCH] Support curved quotes in doc strings

[Paul Eggert <eggert@cs.ucla.edu>]

On 05/15/2015 10:24 AM, Dmitry Gutov wrote:
> It still doesn't have the fancy quotes directly in the source code: 
> they seem to be generated dynamically

Yes, GCC uses "%< %>" in format strings to represent curved quotes, and 
Coreutils has a function 'quote (X)' that returns X surrounded by curved 
quotes.  However, these are both awkward compared to just using quotes.  
GCC and Coreutils use these awkward constructs because they can't assume 
decent support for Unicode. Emacs doesn't have this constraint, so it 
can do better.

> On the other hand, you'll be able to dynamically determine which 
> quotes to use, depending e.g. on the locale. 

That's an independent issue: it can be done no matter what quoting 
regime the source code uses.  But it's probably not worth doing 
separately.  Once we're worried about different locales, we'll need to 
translate the strings to Russian (or whatever) and at that point we can 
translate the quotes to use «Russian style».  (And yes, Emacs should do 
this, just as GCC and Coreutils do -- but one thing at a time.)

> Neither of those should be particularly hard to implement.

I'm not that worried about the cost of implementation now.  I'm worried 
about the long-term cost of having a quoting regime that requires 
documentation and explanation.  It's simpler and easier to explain if 
the doc string looks the same in the UI as it does while you're editing 
it.  For example, you can easily cut and paste from the UI into the doc 
string source when composing a new doc string, which is something that 
doesn't work well for either GCC or Coreutils.

> - Using unicode for markup is a complication (e.g. with certain mail 
> clients, but some other instances might come up).

It's a complication we already have, as we already use UTF-8 in a few 
doc strings.  For example, the documentation for prettify-symbols-mode 
uses UTF-8 curved double-quotes.  This usage will naturally grow in 
time.  Yes, there will be some email glitches but they'll be easy to 
iron out as we gain experience.

> - I don't like having two standards for quoting.

Nor do I.  This is a transition process, with the long-term goal of 
quoting via quotes rather than via grave accent and apostrophe. It's not 
practical to do this transition all at once, hence the transition 
period.  But in the long run there'll be just one standard way to quote 
(even if Emacs supports the obsolete way indefinitely).