input-remapper/readme/usage.md

8.9 KiB

Usage

To open the UI to modify the mappings, look into your applications menu and search for 'Key Mapper'. You should be prompted for your sudo password as special permissions are needed to read events from /dev/input/ files. You can also start it via key-mapper-gtk.

Hitting a key on the device that is selected in the large dropdown on the top should display the key on the bottom of the window, and write it into the selected row (as shown in the screenshots).

In "Mapping", type the key to which you would like to map this key. More information can be found below.

Changes are saved automatically. Afterwards press the "Apply" button.

To change the mapping, you need to use the "Restore Defaults" button, so that the application can read the original keycode. It would otherwise be invisible since the daemon maps it independently of the GUI.

Troubleshooting

If stuff doesn't work, check the output of key-mapper-gtk -d and feel free to open up an issue here. Make sure to not post any debug logs that were generated while you entered private information with your device. Debug logs are quite verbose.

If key-mapper or your presets prevents your input device from working at all due to autoload, please try to unplug and plug it in twice. No injection should be running anymore.

Combinations

Select the key in your row (click here) and hold a few buttons down. Releasing them will make your text cursor jump into the mapping column to type in what you want to map it to.

Combinations involving Modifiers might not work. Configuring a combination of two keys to output a single key will require you to push down the first key, which of course ends up injecting that first key. Then the second key will trigger the mapping, because the combination is complete. This is not a bug. Otherwise every combination would have to automatically disable all keys that are involved in it.

For example a combination of LEFTSHIFT + a for b would write "B" insetad, because shift will be activated before you hit the "a". Therefore the environment will see shift and a "b", which will then be capitalized.

The first option to work around this issue is to use KP1 + a instead of LEFTSHIFT + a. You can disable KP1 by mapping it to disable, so you won't trigger writing a "1" into your focused application.

The second option is to release the modifier in your combination by writing the modifier one more time. This will write lowercase "a" characters.

Writing Combinations

You can write Control_L + a as mapping, which will inject those two keycodes into your system on a single key press. An arbitrary number of names can be chained using +.

UI Shortcuts

  • ctrl + del stops the injection (only works while the gui is in focus)
  • ctrl + q closes the application
  • ctrl + r refreshes the device list

Key Names

Check the autocompletion of the GUI for possible values. You can also obtain a complete list of possiblities using key-mapper-control --symbol-names.

The available symbols depend on the environments keyboard layout, and only those that don't require a combination to be pressed can be used without workarounds (so most special characters need some extra steps to use them). Furthermore, if your configured keyboard layout doesn't support the special character at all (not even via a combination), then it also won't be possible for key-mapper to map that character at all.

For example, mapping a key to an exclamation mark is not possible if the keyboard layout is set to german. However, it is possible to mimic the combination that would be required to write it, by writing Shift_L + 1 into the mapping.

This is because key-mapper creates a new virtual keyboard and injects numeric keycodes, and it won't be able to inject anything a usb keyboard wouldn't been able to. This has the benefit of being compatible to all display servers though.

Key-mapper only recognizes symbol names, but not the symbols themselfes. So for example, key-mapper might (depending on the system layout) know what a minus is, but it doesn't know -.

Key names that start with KEY_ are keyboard layout independent constants that might not result in the expected output. For example using KEY_Y would result in "z" if the layout of the environment is set to german. Using y on the other hand would correctly result in "y" to be written.







Advanced

Configuration Files

If you don't have a graphical user interface, you'll need to edit the configuration files.

The default configuration is stored at ~/.config/key-mapper/config.json, which doesn't include any mappings, but rather other parameters that are interesting for injections. The current default configuration as of 1.2.1 looks like, with an example autoload entry:

{
    "autoload": {
        "Logitech USB Keyboard": "preset name"
    },
    "macros": {
        "keystroke_sleep_ms": 10
    },
    "gamepad": {
        "joystick": {
            "non_linearity": 4,
            "pointer_speed": 80,
            "left_purpose": "none",
            "right_purpose": "none",
            "x_scroll_speed": 2,
            "y_scroll_speed": 0.5
        }
    }
}

preset name refers to ~/.config/key-mapper/presets/device name/preset name.json. The device name can be found with sudo key-mapper-control --list-devices.

Anything that is relevant to presets can be overwritten in them as well. Here is an example configuration for preset "a" for the "gamepad" device: ~/.config/key-mapper/presets/gamepad/a.json

{
    "macros": {
        "keystroke_sleep_ms": 100
    },
    "mapping": {
        "1,315,1+1,16,-1": "1",
        "1,307,1": "k(2).k(3)"
    }
}

Both need to be valid json files, otherwise the parser refuses to work. This preset maps the EV_KEY down event with code 307 to a macro and sets the time between injected events of macros to 100 ms. Note that a complete keystroke consists of two events: down and up. The other mapping is a key combination, chained using +.

Other than that, it inherits all configurations from ~/.config/key-mapper/config.json. If config.json is missing some stuff, it will query the hardcoded default values.

The event codes can be read using evtest. Available names in the mapping can be listed with key-mapper-control --symbol-names.

CLI

key-mapper-control

--command requires the service to be running. You can start it via systemctl start key-mapper or sudo key-mapper-service if it isn't already running (or without sudo if your user has the appropriate permissions).

Examples:

Description Command
Load all configured presets for all devices key-mapper-control --command autoload
If you are running as root user, provide information about the whereabouts of the key-mapper config key-mapper-control --command autoload --config-dir "~/.config/key-mapper/"
List available device names for the --device parameter sudo key-mapper-control --list-devices
Stop injecting key-mapper-control --command stop --device "Razer Razer Naga Trinity"
Load ~/.config/key-mapper/presets/Razer Razer Naga Trinity/a.json key-mapper-control --command start --device "Razer Razer Naga Trinity" --preset "a"
Loads the configured preset for whatever device is using this /dev path /bin/key-mapper-control --command autoload --device /dev/input/event5

systemctl

Stopping the service will stop all ongoing injections

sudo systemctl stop key-mapper
sudo systemctl start key-mapper
systemctl status key-mapper

Testing your Installation

The following commands can be used to make sure it works:

sudo key-mapper-service &
key-mapper-control --command hello

should print Daemon answered with "hello". And

sudo key-mapper-control --list-devices

should print Found "...", .... If anything looks wrong, feel free to create an issue.