Designed and coded solution to the fueling problem; introduced design for Huffman encoding.

greedy-algorithms/notes.md

@@ -126,8 +126,8 @@ Problem: [same as above]
 ## Choosing a Forward Edge
 
 *   Exhaustive search:
-    *   Generate …
-    *   Check that …
+    *   Generate all edges that travel up to distance `range` along the route
+    *   Check that the chosen edge travels farthest (always favors the current candidate)
 
 ## Example
 
@@ -143,12 +143,15 @@ Problem: [same as above]
 
     Location  Next Location
     --------  -------------
-    a         …
-    …
+    a         c
+    c         d
+    d         e
+    e         h
+    h         i
 
     Path
     ----
-    a → …
+    a → c → d → e → h → i
 
 --------------------------------------------------------------------------------
 
@@ -214,22 +217,22 @@ Problem: Given frequencies for a set of symbols, create a lossless binary encodi
 ## DAG
 
 *   Edges (actions):
-    *   …
+    *   Combining two sets of symbols
 *   Vertices (situations):
-    *   …
+    *   Symbol partitionnings
 *   Edge weights:
-    *   …
+    *   Total frequency of the combined sets
 *   Topological order:
-    *   …
+    *   By decreasing number of sets (parts in the paritioning)
 *   Goal:
-    *   …
+    *   Find the shortest path
 
 ## Choosing a Forward Edge
 
 *   Direct solution:
-    *   …
-    *   …
-    *   …
+    *   Choose the lowest-frequency set
+    *   Choose the second-lowest-frequency set
+    *   Choose the edge that combines those sets
 
 ## Example
 

greedy-algorithms/src/features/fueling/solver.js

@@ -7,7 +7,16 @@ export function planFuelings(gaps, range) {
     gaps.every((gap) => gap <= range),
     `Tried to find a fueling plan crossing a gap longer than the range ${range}.`,
   );
-  const results = [];
-  // TODO: stub
+  const destination = gaps.length;
+  const results = [0];
+  while (results[results.length - 1] !== destination) {
+    let best = undefined;
+    for (let distance = 0, candidate = results[results.length - 1];
+      distance <= range && candidate <= destination;
+      distance += gaps[candidate], ++candidate) {
+      best = candidate;
+    }
+    results.push(best);
+  }
   return results;
 }