day 6

Quick challenge today, just from the input alone it is easy to conclude
the naive solution will work fine even with a large windowing size.

day_6/src/main.rs

@@ -1,3 +1,57 @@
-fn main() {
+use std::{collections::HashSet, fs::read_to_string, ops::Add};
-    println!("Hello, world!");
+
+const PATH: &str = "src/input";
+
+const HEADER_PACKET_LENGTH: usize = 4;
+const MESSAGE_LENGTH: usize = 14;
+
+fn is_unique(chunk: &str) -> bool {
+    let mut set = HashSet::<u8>::new();
+    for c in chunk.bytes().into_iter() {
+        if set.contains(&c) {
+            return false;
+        } else {
+            set.insert(c);
+        }
+    }
+    return true;
+}
+
+fn main() {
+    assert!(HEADER_PACKET_LENGTH > 0);
+    assert!(MESSAGE_LENGTH > 0);
+
+    let datagram = read_to_string(PATH)
+        .expect("Error reading file")
+        .replace(char::is_whitespace, "")
+        .add(&"?".repeat(HEADER_PACKET_LENGTH))
+        .add(&"?".repeat(MESSAGE_LENGTH));
+
+    assert!(datagram.is_ascii());
+
+    let mut header_idx: usize = 0;
+    let mut message_idx: usize = 0;
+
+    for i in 0..datagram.len() {
+        // There are certainly more efficient methods that save computation by only
+        // computing whether the addition of one character and removal of the last
+        // are sufficient to be unique. This is easier though.
+        if header_idx == 0 {
+            let window_header = datagram.get(i..i + HEADER_PACKET_LENGTH).unwrap();
+            if is_unique(window_header) {
+                header_idx = i + HEADER_PACKET_LENGTH;
+            }
+        }
+        if message_idx == 0 {
+            let window_message = datagram.get(i..i + MESSAGE_LENGTH).unwrap();
+            if is_unique(window_message) {
+                message_idx = i + MESSAGE_LENGTH;
+            }
+        }
+        if header_idx != 0 && message_idx != 0 {
+            println!("PART 1 {}", header_idx);
+            println!("PART 2 {}", message_idx);
+            break;
+        }
+    }
 }