On Thu, 21 Apr 2022 12:33:06 +0400 "Tigran S." <tisaruk...@gmail.com> wrote:
> Hello, > I am testing a new keyboard layout that has seven alphabetic dead keys (T, > S, D, G, K, Z, and C). To type a word that has two consecutive dead keys, I > have to type a space between them. For example, to input the Armenian word > թխկի, I have to type t khki, with a space between t and k. But in macOS, I > can type that word without a space between t and k dead keys. Also in > macOS, when pressing a dead key, the corresponding first letter is input > immediately and highlighted. Is it possible to have such an implementation > of dead keys in Linux? > > The repository of the keyboard layout is located at > https://github.com/dotland/mnemonic-kb-hy. Hi, I'm very far from being knowledgeable on this topic, but I believe the following: What you are describing would have to be an input method, not a keymap nor a compose map. The reason for this is that once a character appears on screen, it has left the keymap processing for good. So it's not possible for a keymap to change what has already shown up on screen. The only thing a keymap can do is postpone the emission of a character until there are enough key-presses to generate one (or more), and then it's done, no taking back. Input methods on the other hand should be able to do everything you want, including visual feedback during the composition of a character, word, or sentence. Applications must explicitly support an input method protocol, but I think all major toolkits do. Thanks, pq
pgp2JB0p61HLh.pgp
Description: OpenPGP digital signature
