Dialogue aftermath.

This is what happens after you call the dialog_run(). Here's the entire code, including the parts mentioned before:

    int  result;

    sprint f(inp_buffer,"%s\n%s\ns\n%s\nOK\nCancel\n",
                        "First string field",
			"password",
			"Yes",
			"on");

    result = dialog_run ("foo", FOOVT, FOOLT, inp_buffer);

    if (result!=0) {
            error_log("Unable to run data entry subsystem");
	    return;
    }

    dialog_parse(inp_buffer);

    if (sameas(margv[6],"OK") || sameas (margv[4],margv[6])) {
            print ("String field 1 has value \"%s\"\n",margv[0]);
            print ("String field 2 has value \"%s\"\n",margv[1]);
	    print ("Toggle has value         \"%s\"\n",margv[2]);
	    print ("List has value           \"%s\"\n",margv[3]);
    } else print ("Cancel button pressed.\n");

    

This is where using inp_buffer shows its advantages: you can parse the entire result string with a single call to dialog_parse(), which is virtually the same function we use to parse a user's input. I like re-using code!

The rest should be pretty obvious. The last datum returned by dialog_run() is the label of one of three things:

• The label of button that ended the dialogue. This may change, depending on sysop customisation and internationalisation.

• The string "OK", if the user pressed the `save' shortcut, which is synonymous to `pressing' the OK button. You should always have an OK button, but even if you don't, this shortcut will still work in this way.

• The string "Cancel", if the user pressed the `cancel' shortcut.

Since our dialogue only has two buttons, OK and Cancel, we check for the `OK' shortcut and an actual button `press'. The latter is done by comparing the label of the OK button (whatever it is) against the label of the button that ended the dialogue.

Alphabetic index