Move to double to avoid rounding error

This rounding errors seems to happen when adding floats very close to 1
to a relatively large number.

325.0 + 0.99999034 = 326.0

This is a problem as this sample should be counted at the 325 pixel.
However this will be a lesser problem when filtering is implemented.

1 files changed, 10 insertions, 0 deletions

diff --git a/src/core/vector2.rs b/src/core/vector2.rs
index 858068e..b3fa443 100644
--- a/src/core/vector2.rs
+++ b/src/core/vector2.rs
@@ -4,6 +4,7 @@
 use crate::{Float, Number};
 use std::ops::{Sub, Add};
 use std::fmt;
+use std::cmp::min;
 
 #[derive(Clone, Copy)]
 pub struct Vector2<T: Number> {
@@ -84,6 +85,15 @@ impl From<Vector2f> for Vector2i {
     }
 }
 
+impl Vector2i {
+    pub fn cap(&self, x: i32, y: i32) -> Self {
+        Self::new_xy(
+            min(self.x, x),
+            min(self.y, y),
+            )
+    }
+}
+
 
 #[cfg(test)]
 mod tests {