ImDrawData: Slight refactor so internal logic uses same logic as AddDrawList(). (#6406, #4879, #1878)

# Conflicts:
#	imgui.cpp
#	imgui_internal.h

imgui_draw.cpp

@@ -1817,15 +1817,52 @@ void ImDrawData::Clear()
     OwnerViewport = NULL;
 }
 
+// Important: 'out_list' is generally going to be draw_data->CmdLists, but may be another temporary list
+// as long at it is expected that the result will be later merged into draw_data->CmdLists[].
+void ImGui::AddDrawListToDrawDataEx(ImDrawData* draw_data, ImVector<ImDrawList*>* out_list, ImDrawList* draw_list)
+{
+    if (draw_list->CmdBuffer.Size == 0)
+        return;
+    if (draw_list->CmdBuffer.Size == 1 && draw_list->CmdBuffer[0].ElemCount == 0 && draw_list->CmdBuffer[0].UserCallback == NULL)
+        return;
+
+    // Draw list sanity check. Detect mismatch between PrimReserve() calls and incrementing _VtxCurrentIdx, _VtxWritePtr etc.
+    // May trigger for you if you are using PrimXXX functions incorrectly.
+    IM_ASSERT(draw_list->VtxBuffer.Size == 0 || draw_list->_VtxWritePtr == draw_list->VtxBuffer.Data + draw_list->VtxBuffer.Size);
+    IM_ASSERT(draw_list->IdxBuffer.Size == 0 || draw_list->_IdxWritePtr == draw_list->IdxBuffer.Data + draw_list->IdxBuffer.Size);
+    if (!(draw_list->Flags & ImDrawListFlags_AllowVtxOffset))
+        IM_ASSERT((int)draw_list->_VtxCurrentIdx == draw_list->VtxBuffer.Size);
+
+    // Check that draw_list doesn't use more vertices than indexable (default ImDrawIdx = unsigned short = 2 bytes = 64K vertices per ImDrawList = per window)
+    // If this assert triggers because you are drawing lots of stuff manually:
+    // - First, make sure you are coarse clipping yourself and not trying to draw many things outside visible bounds.
+    //   Be mindful that the lower-level ImDrawList API doesn't filter vertices. Use the Metrics/Debugger window to inspect draw list contents.
+    // - If you want large meshes with more than 64K vertices, you can either:
+    //   (A) Handle the ImDrawCmd::VtxOffset value in your renderer backend, and set 'io.BackendFlags |= ImGuiBackendFlags_RendererHasVtxOffset'.
+    //       Most example backends already support this from 1.71. Pre-1.71 backends won't.
+    //       Some graphics API such as GL ES 1/2 don't have a way to offset the starting vertex so it is not supported for them.
+    //   (B) Or handle 32-bit indices in your renderer backend, and uncomment '#define ImDrawIdx unsigned int' line in imconfig.h.
+    //       Most example backends already support this. For example, the OpenGL example code detect index size at compile-time:
+    //         glDrawElements(GL_TRIANGLES, (GLsizei)pcmd->ElemCount, sizeof(ImDrawIdx) == 2 ? GL_UNSIGNED_SHORT : GL_UNSIGNED_INT, idx_buffer_offset);
+    //       Your own engine or render API may use different parameters or function calls to specify index sizes.
+    //       2 and 4 bytes indices are generally supported by most graphics API.
+    // - If for some reason neither of those solutions works for you, a workaround is to call BeginChild()/EndChild() before reaching
+    //   the 64K limit to split your draw commands in multiple draw lists.
+    if (sizeof(ImDrawIdx) == 2)
+        IM_ASSERT(draw_list->_VtxCurrentIdx < (1 << 16) && "Too many vertices in ImDrawList using 16-bit indices. Read comment above");
+
+    // Add to output list + records state in ImDrawData
+    out_list->push_back(draw_list);
+    draw_list->_PopUnusedDrawCmd();
+    draw_data->CmdListsCount++;
+    draw_data->TotalVtxCount += draw_list->VtxBuffer.Size;
+    draw_data->TotalIdxCount += draw_list->IdxBuffer.Size;
+}
+
 void ImDrawData::AddDrawList(ImDrawList* draw_list)
 {
-    IM_ASSERT(Valid);
     IM_ASSERT(CmdLists.Size == CmdListsCount);
-    draw_list->_PopUnusedDrawCmd();
-    CmdLists.push_back(draw_list);
-    CmdListsCount++;
-    TotalVtxCount += draw_list->VtxBuffer.Size;
-    TotalIdxCount += draw_list->IdxBuffer.Size;
+    ImGui::AddDrawListToDrawDataEx(this, &CmdLists, draw_list);
 }
 
 // For backward compatibility: convert all buffers from indexed to de-indexed, in case you cannot render indexed. Note: this is slow and most likely a waste of resources. Always prefer indexed rendering!