Unity small map production [2]

The project team has added some advanced functions to the small map module , Here is a record of the requirements and production process

demand

The following figure shows the effect of my understanding of requirements errors , When the touch is outside the map , The actual position of the camera should be the same as the blue dot square in the figure , This is the center of the map , Blue dot square , Touch point three o'clock line , I feel that this situation may be useful in the future , So keep .

Ideas

On the basis of one , take UI The local coordinates of the mapping object are limited . Because the actual map was originally made by UI Remapping by mapping the local coordinates of the object .

Code

Time relationship directly give the code ,Test1 Just phased ,Test2 Is the final version of the experiment
The problems encountered are also commented in the code , And this one demo Only applicable to large maps and UI Of mapping objects pivot All are 0.5,0.5 The situation of , In other cases, you should calculate the rectangle by yourself Rect In the center of

using System.Collections;
using System.Collections.Generic;
using UnityEngine;

public class Test1 : MonoBehaviour
{
    
    [SerializeField]
    private RectTransform mapRect;

    
    [SerializeField]
    private RectTransform smallMapRect;
    
    

    [SerializeField]
    private bool onlyMarginOffset;
    
    
    [SerializeField]
    private float offset;
    
    
    [SerializeField]
    private float currentOffset;
    
    // Start is called before the first frame update
    void Start()
    {
    
        if (onlyMarginOffset)
        {
    
            // Try one   Use the diagonal length as the offset value   Performance is not what you want 
            float halfDiagonal = Mathf.Sqrt(Mathf.Pow(smallMapRect.rect.width / 2f, 2) + Mathf.Pow(smallMapRect.rect.height / 2f, 2));
            currentOffset = halfDiagonal;
        }
        else
        {
    
            // Try two   Use a fixed value as the offset value   Performance is not what you want 
            currentOffset = offset;
        }
    }

    // Update is called once per frame
    void Update()
    {
    
        if (Input.GetMouseButtonDown(0))
        {
    
             bool isIn = RectTransformUtility.RectangleContainsScreenPoint(mapRect, Input.mousePosition);
             print(" isIn " + isIn);

             if (!isIn)
             {
    
                 // Pay attention to the difference between this paragraph and the next paragraph , It should be noted that 
                 //1.  Coordinate system , All coordinates involved in ray and assignment need to be in the same coordinate system , That makes sense   Be clear Overlay Of Canvas and Camera Of Canvas Of position The difference between   The current script is only applicable to Overlay
                 //2.  If the ray is from Collider From inside , Then the point of collision can only be the launch point , This is Unity Of Raycast Of API The rules of ,
                 // The rule is to carry out step-by-step detection along the path from the starting point to the end point of the ray , Once a point is detected at a Collider Inside , Stop the detection , And return collision information 
                 // So it's a more accurate method of judging whether it's done internally or not to the edge 
                 // So the introduction Physics2D.Raycast The ray formed by the parameter of must be from collider From the outside to the inside 
                 // Vector3 dir = Input.mousePosition - mapRect.position;
                 // RaycastHit2D hit = Physics2D.Raycast(mapRect.position, dir);
                 
                 Vector3 dir =  mapRect.position - Input.mousePosition;
                 RaycastHit2D hit = Physics2D.Raycast(Input.mousePosition, dir);
                 
                 if (hit.collider != null)
                 {
    
                     // Try one and two   Code for 
                     Vector3 dirVector = new Vector3(hit.point.x , hit.point .y, mapRect.position.z) - mapRect.position;
                     Vector3 normalizedDirVector = dirVector.normalized;
                     float newMagnitude = (dirVector.magnitude - currentOffset);
                     Vector3 newDirVector = normalizedDirVector * newMagnitude;
                     smallMapRect.position = mapRect.position + newDirVector;
                     
                 }

                    
             }
             
             
             
             
        }
    }
}

using System.Collections;
using System.Collections.Generic;
using UnityEngine;

public class Test2 : MonoBehaviour
{
    
    [SerializeField]
    private RectTransform mapRect;

    
    [SerializeField]
    private RectTransform smallPointRect;

    
    private BoxCollider2D smallPointCollider2D;
    
    
    private BoxCollider2D mapCollider2D;


    private Vector2 lastHitMapBorderPos;

    private bool rayCastNextFrame;
    
    // Start is called before the first frame update
    void Start()
    {
    
        smallPointCollider2D = smallPointRect.GetComponent<BoxCollider2D>();
        mapCollider2D = mapRect.GetComponent<BoxCollider2D>();
    }

    // Update is called once per frame
    void Update()
    {
    
        if (Input.GetMouseButton(0))
        {
    
             bool isIn = RectTransformUtility.RectangleContainsScreenPoint(mapRect, Input.mousePosition);
             print(" isIn " + isIn);

             if (!isIn)
             {
    
                 // Option three 
                 // Diagonal judgment 
                 
                 
                 Vector3 intoMapDir =  mapRect.position - Input.mousePosition;
                 smallPointCollider2D.enabled = false;
                 mapCollider2D.enabled = true;
                 RaycastHit2D hit = Physics2D.Raycast(Input.mousePosition, intoMapDir);
                
                 
                 // Option four 
                 
                 if (hit.collider != null)
                 {
    
                     smallPointRect.position = hit.point;
                     
                     mapCollider2D.enabled = false;
                     smallPointCollider2D.enabled = true;

                     
                     Vector3 outOfMapDir = Input.mousePosition - mapRect.position ;


                     lastHitMapBorderPos = hit.point;
                     RaycastHit2D hit2 = Physics2D.Raycast(mapRect.position, outOfMapDir);
                     
                     
                         
                     if (hit2.collider != null )
                     {
    
                         float minusDeltaDistance = (hit2.point - hit.point).magnitude;

                         Vector2 normalizedDir = (hit.point - hit2.point).normalized;
                         
                         Vector2 currentDirVector = new Vector3(hit.point.x, hit.point.y,  mapRect.position.z)  - mapRect.position;

                         float actualMagnitude = currentDirVector.magnitude - minusDeltaDistance;

                         Vector2 optimizedVector = actualMagnitude * normalizedDir;
                         
                         Vector2 optimizedPos = mapRect.position + new Vector3(optimizedVector.x, optimizedVector.y, 0);

                         smallPointRect.position =
                             new Vector3(optimizedPos.x, optimizedPos.y, smallPointRect.position.z);
                     }
                     
                 }

                    
             }
             
             
             
             
        }
    }
}

here Test2 At that time, scheme 3 was like this , Divide the center point of the rectangle and the four corners of the rectangle into four triangles , Judge the center point 360 The angle divided by these four triangles , According to the angle between the vector from the mouse to the center point and a line of the four triangles in the picture , Determine which triangle it belongs to , Then, according to the triangle, the edges of the large map are directly pushed inward, and the length of the width of the small map is limited . But the feeling is too complicated , We adopted plan 4

I encountered a problem of non response on the way , It was thought that the range update of the collision body would not be effective until the next frame after setting the position , It is found that the collision body is not opened , Then it's all right .

There are other collision bodies and I don't want to add layer Under the circumstances , use RayCastAll Add traversal to see whether the name of the collision object is the name of the specified object .

---

Demo