Re: [Bug-zile] Why are \e and \r read-syntaxes marked as FIXME?

[Gary V . Vaughan]

Hi Reuben,

[[re-added bug-zile to Cc:]]

On 21 Aug 2011, at 19:21, Reuben Thomas wrote:
> On 20/08/2011, Gary V. Vaughan <address@hidden> wrote:
>> Considering Zile is emulating emacs to a greater degree, what is it
>> that needs fixing with emulating \e and \r read-syntaxes in keycode.c?
> 
> It's just that they're not strictly speaking key names, and they're
> stored in the key names array. Perhaps the fix is as simple as
> renaming things to do with "keys" to be to do rather with string
> read-syntax?

Okay that makes sense.  Thanks for the explanation :)

On the Zi branch, I'm trying to refactor all of the various references to
keys in curses and zile, and keynames, and read-syntaxes into two tables:

  1. keycodes.lua has a single table (not counting reverse lookup tables)
     for converting between read-syntax->internal representation->display
     syntax (for zi, I think we can profitably simplify away the arbirtary
     difference between read-syntax and display-syntax, but Zile needs to
     differentiate for compatibility with Emacs I think?).
  2. term_curses.lua has a single table (counting likewise) for converting
     between curses keys -> internal representation.

Things have started to simplify considerably now that I'm making the zi
internal representation of a key be a reference to its row in the keycodes
table, instead of a list of arbitrary numbers in main.lua and bitwise
flips for ctrl and alt.

Unfortunately, I've broken things horribly at the moment so I can't push,
but once I've reassembled the pieces, we ought be able to apply some of
the same logic to the other branches.

Cheers,
-- 
Gary V. Vaughan (gary AT gnu DOT org)