On 03/20/2016 12:49 PM, Dmitry Gutov wrote:
> On 03/20/2016 09:47 PM, Jostein Kjønigsen wrote:
> 
>>> diff --git a/lisp/progmodes/js.el b/lisp/progmodes/js.el
>>> index 15a52ba..dffdb77 100644
>>> --- a/lisp/progmodes/js.el
>>> +++ b/lisp/progmodes/js.el
>>> @@ -578,6 +578,7 @@ js-mode-syntax-table
>>>       (c-populate-syntax-table table)
>>>       (modify-syntax-entry ?$ "_" table)
>>>       (modify-syntax-entry ?` "\"" table)
>>> +    (modify-syntax-entry ?\C-M " " table)
>>>       table)
>>>     "Syntax table for `js-mode'.")
>>>
>>
>> That works for me, at least in js-mode.
>>
>> While not a general solution to all weird files one might encounter
>> while working in Emacs, it at least makes js-files behave as expected.
>> And in a fairly reusable manner too: I can copy this code and I'm home
>> free.
>>
>> Would pushing this up to cc-mode's c-populate-syntax-table function also
>> be considered an option? Or does that just sound on overall way too
>> risky?
> 
> I don't see any problem there, but that's up to Alan.

It don't have any problem teaching specific modes about carriage
returns, but this approach doesn't scale. A better approach, I think,
would be to extend something like ws-butler for dealing with
mixed-line-ending files. On load, we'd transform `\r\n' -> `\n', but
*remember* which `\n's represent `\r\n` and perform the reverse
translation when writing. This way, you'd preserve clowny extra carriage
returns from other people, but see a clean, unified view in Emacs.