Add orthographic cameras.

Orthographic cameras are specified by setting the `orthographic` attribute of the `<camera>` element. The `fovy` attribute is still used to specify the field-of-view, but its semantic is different for orthographic cameras. For orthographic cameras, the field-of-view is expressed in units of length, rather than degrees.

Other related changes:

* Fix bug in the ordering of `cam_xxx` elements in `mjModel`.
* Make camera visualization translucent only when the frustum is visualized.
* Added a button to `simulate` to toggle between perspective and orthographic free cameras.

https://youtu.be/ZXBTEIDWHhs

PiperOrigin-RevId: 642293435
Change-Id: Id090a421ad88ab404b5b27ddbbd6bfc81ad49bc5

test/xml/xml_native_reader_test.cc

@@ -1249,6 +1249,36 @@ TEST_F(XMLReaderTest, InvalidSkinGroup) {
   EXPECT_THAT(
       error.data(),
       HasSubstr("skin group must be between 0 and 5\nElement 'skin', line 7"));
+}
+
+TEST_F(XMLReaderTest, Orthographic) {
+  static constexpr char xml[] = R"(
+  <mujoco>
+    <visual>
+      <global orthographic="true" fovy="5"/>
+      <map znear="0.01"/>
+    </visual>
+
+    <default>
+      <camera orthographic="true"/>
+    </default>
+
+    <worldbody>
+      <camera name="fovy=1" pos=".5 .5 2" orthographic="true" fovy="1"/>
+      <camera name="fovy=2" pos="0 0 2" fovy="2"/>
+    </worldbody>
+  </mujoco>
+  )";
+  std::array<char, 1024> error;
+  mjModel* model = LoadModelFromString(xml, error.data(), error.size());
+  EXPECT_THAT(model, NotNull()) << error.data();
+
+  EXPECT_EQ(model->vis.global.orthographic, 1);
+  EXPECT_EQ(model->cam_orthographic[0], 1);
+  EXPECT_EQ(model->cam_orthographic[1], 1);
+  EXPECT_EQ(model->cam_fovy[0], 1);
+  EXPECT_EQ(model->cam_fovy[1], 2);
+
   mj_deleteModel(model);
 }