⚔️
UE5 Action-Combat 3C Demo
- 构建 C++ 优先的输入管线：Enhanced Input 喂给自定义 UInputComponent 子类，把原始输入分类成 gameplay intent，用一层很薄的 Blueprint 转换层保持每角色绑定的数据驱动。Built a C++-first input pipeline: Enhanced Input feeding a custom UInputComponent subclass that classifies raw input into gameplay intents, with a thin Blueprint translation layer keeping per-character bindings data-driven.
- 实现带 Custom Movement Modes 的 CharacterMovementComponent 子类，包括当角色以一定角度阈值靠近墙面时自动进入攀爬模式的 wall climbing。Implemented a CharacterMovementComponent subclass with Custom Movement Modes, including wall climbing that auto-enters climb mode when the character approaches a surface within an angle threshold.
- 实现预测性 Foot IK —— 手 K 的 foot-state 与落地时机动画曲线驱动 IK Rig 节点，加上骨盆偏移 —— 让脚步在不平地形上提前预判，而不是接触后再纠正。Implemented predictive foot IK — hand-authored foot-state and plant-time animation curves driving an IK Rig node, plus pelvis offset — so foot placement anticipates uneven terrain instead of correcting after contact.
- 搭建 locomotion Animation Blueprint（嵌套 Idle ↔ Move 状态机 + Walk/Run 子状态）与 GAS 骨架（AbilitySystemComponent、AttributeSet、技能驱动的攻击）。Built the locomotion Animation Blueprint (nested Idle ↔ Move state machine with Walk/Run sub-states) and a Gameplay Ability System skeleton (AbilitySystemComponent, AttributeSet, ability-driven attacks).
- 研究成品架构作为参考 —— Epic 的 Lyra 示例 (PawnData、GameplayTag registry、AnimNotify CDO 模型) 以及对某商业 UE ARPG 的资产级分析。Studied shipped-title architecture as reference — Epic's Lyra sample (PawnData, GameplayTag registry, AnimNotify CDO model) and asset-level analysis of a commercial UE ARPG.
C++BlueprintGASIK Rig