Add Root struct

Server handles listening for wayland signals. It delegates input handling
to Seat and output handling to Root.

1 files changed, 25 insertions, 61 deletions

diff --git a/src/server.zig b/src/server.zig
index 81dfe46..06185db 100644
--- a/src/server.zig
+++ b/src/server.zig
@@ -2,6 +2,7 @@ const std = @import("std");
 const c = @import("c.zig").c;
 
 const Output = @import("output.zig").Output;
+const Root = @import("root.zig").Root;
 const Seat = @import("seat.zig").Seat;
 const View = @import("view.zig").View;
 
@@ -10,91 +11,71 @@ pub const Server = struct {
 
     allocator: *std.mem.Allocator,
 
+    root: Root,
+    seat: Seat,
+
     wl_display: *c.wl_display,
     wlr_backend: *c.wlr_backend,
     wlr_renderer: *c.wlr_renderer,
 
-    wlr_output_layout: *c.wlr_output_layout,
-    outputs: std.TailQueue(Output),
-
     listen_new_output: c.wl_listener,
 
     wlr_xdg_shell: *c.wlr_xdg_shell,
     listen_new_xdg_surface: c.wl_listener,
 
-    // Must stay ordered, first in list is "on top" visually
-    views: std.TailQueue(View),
-
-    seat: Seat,
-
-    pub fn create(allocator: *std.mem.Allocator) !Self {
-        var server: Self = undefined;
-        server.allocator = allocator;
+    pub fn init(self: *Self, allocator: *std.mem.Allocator) !void {
+        self.allocator = allocator;
 
         // The Wayland display is managed by libwayland. It handles accepting
         // clients from the Unix socket, manging Wayland globals, and so on.
-        server.wl_display = c.wl_display_create() orelse
+        self.wl_display = c.wl_display_create() orelse
             return error.CantCreateWlDisplay;
-        errdefer c.wl_display_destroy(server.wl_display);
+        errdefer c.wl_display_destroy(self.wl_display);
 
         // The wlr_backend abstracts the input/output hardware. Autocreate chooses
         // the best option based on the environment, for example DRM when run from
         // a tty or wayland if WAYLAND_DISPLAY is set.
         //
-        // This frees itserver.when the wl_display is destroyed.
-        server.wlr_backend = c.zag_wlr_backend_autocreate(server.wl_display) orelse
+        // This frees itself.when the wl_display is destroyed.
+        self.wlr_backend = c.zag_wlr_backend_autocreate(self.wl_display) orelse
             return error.CantCreateWlrBackend;
 
         // If we don't provide a renderer, autocreate makes a GLES2 renderer for us.
         // The renderer is responsible for defining the various pixel formats it
         // supports for shared memory, this configures that for clients.
-        server.wlr_renderer = c.zag_wlr_backend_get_renderer(server.wlr_backend) orelse
+        self.wlr_renderer = c.zag_wlr_backend_get_renderer(self.wlr_backend) orelse
             return error.CantGetWlrRenderer;
         // TODO: Handle failure after https://github.com/swaywm/wlroots/pull/2080
-        c.wlr_renderer_init_wl_display(server.wlr_renderer, server.wl_display); // orelse
+        c.wlr_renderer_init_wl_display(self.wlr_renderer, self.wl_display); // orelse
         //    return error.CantInitWlDisplay;
 
         // These both free themselves when the wl_display is destroyed
-        _ = c.wlr_compositor_create(server.wl_display, server.wlr_renderer) orelse
+        _ = c.wlr_compositor_create(self.wl_display, self.wlr_renderer) orelse
             return error.CantCreateWlrCompositor;
-        _ = c.wlr_data_device_manager_create(server.wl_display) orelse
+        _ = c.wlr_data_device_manager_create(self.wl_display) orelse
             return error.CantCreateWlrDataDeviceManager;
 
-        // Create an output layout, which a wlroots utility for working with an
-        // arrangement of screens in a physical layout.
-        server.wlr_output_layout = c.wlr_output_layout_create() orelse
-            return error.CantCreateWlrOutputLayout;
-        errdefer c.wlr_output_layout_destroy(server.wlr_output_layout);
-
-        // Don't register the wl_listeners yet as they must first be pointer-stable
-        server.outputs = std.TailQueue(Output).init();
-        server.listen_new_output.notify = handle_new_output;
-
-        server.views = std.TailQueue(View).init();
-        server.wlr_xdg_shell = c.wlr_xdg_shell_create(server.wl_display) orelse
+        self.wlr_xdg_shell = c.wlr_xdg_shell_create(self.wl_display) orelse
             return error.CantCreateWlrXdgShell;
-        server.listen_new_xdg_surface.notify = handle_new_xdg_surface;
 
-        return server;
-    }
+        try self.root.init(self);
 
-    pub fn init(self: *Self) !void {
         self.seat = try Seat.create(self);
         try self.seat.init();
 
         // Register our listeners for new outputs and xdg_surfaces.
-        // This can't be done in create() as wl_signal_add() creates a pointer
-        // to the wl_list link in our wl_listener, a pointer that would be
-        // broken when returning from create();
+        self.listen_new_output.notify = handle_new_output;
         c.wl_signal_add(&self.wlr_backend.events.new_output, &self.listen_new_output);
+
+        self.listen_new_xdg_surface.notify = handle_new_xdg_surface;
         c.wl_signal_add(&self.wlr_xdg_shell.events.new_surface, &self.listen_new_xdg_surface);
     }
 
     /// Free allocated memory and clean up
-    pub fn destroy(self: Self) void {
+    pub fn destroy(self: *Self) void {
         c.wl_display_destroy_clients(self.wl_display);
         c.wl_display_destroy(self.wl_display);
-        c.wlr_output_layout_destroy(self.wlr_output_layout);
+        self.root.destroy();
     }
 
     /// Create the socket, set WAYLAND_DISPLAY, and start the backend
@@ -148,11 +129,7 @@ pub const Server = struct {
     fn handle_new_output(listener: ?*c.wl_listener, data: ?*c_void) callconv(.C) void {
         const server = @fieldParentPtr(Server, "listen_new_output", listener.?);
         const wlr_output = @ptrCast(*c.wlr_output, @alignCast(@alignOf(*c.wlr_output), data));
-
-        // TODO: Handle failure
-        const node = server.outputs.allocateNode(server.allocator) catch unreachable;
-        node.data.init(server, wlr_output) catch unreachable;
-        server.outputs.append(node);
+        server.root.addOutput(wlr_output);
     }
 
     fn handle_new_xdg_surface(listener: ?*c.wl_listener, data: ?*c_void) callconv(.C) void {
@@ -162,24 +139,11 @@ pub const Server = struct {
         const wlr_xdg_surface = @ptrCast(*c.wlr_xdg_surface, @alignCast(@alignOf(*c.wlr_xdg_surface), data));
 
         if (wlr_xdg_surface.role != c.enum_wlr_xdg_surface_role.WLR_XDG_SURFACE_ROLE_TOPLEVEL) {
+            // TODO: log
             return;
         }
 
-        // Create a View to handle this toplevel surface
-        const node = server.views.allocateNode(server.allocator) catch unreachable;
-        node.data.init(server, wlr_xdg_surface);
-        server.views.append(node);
-    }
-
-    /// Finds the top most view under the output layout coordinates lx, ly
-    /// returns the view if found, and a pointer to the wlr_surface as well as the surface coordinates
-    pub fn desktop_view_at(self: *Self, lx: f64, ly: f64, surface: *?*c.wlr_surface, sx: *f64, sy: *f64) ?*View {
-        var it = self.views.last;
-        while (it) |node| : (it = node.prev) {
-            if (node.data.isAt(lx, ly, surface, sx, sy)) {
-                return &node.data;
-            }
-        }
-        return null;
+        // toplevel surfaces are tracked and managed by the root
+        server.root.addView(wlr_xdg_surface);
     }
 };