Solutions to 09

NOTES.md

@@ -24,6 +24,7 @@ Hmmmm, right now it seems even worse than GoLang. Though the Error handling is _
 * Switching on strings
 * [Iterating over values of an enum](https://zig.guide/language-basics/enums) - [this](https://ziggit.dev/t/iterating-over-a-packed-enum/6530) suggests that it's possible, but testing indicates that that only works at comptime.
 * [Sets](https://github.com/ziglang/zig/issues/6919) (though, as the top comment points out, that's not _too_ bad as you can abuse a HashMap for it)
+* [Descending ranges](https://github.com/ziglang/zig/issues/14701) - i.e. `for (10..0)`. Not a big deal, tbf, but annoying when needed!
 
 ## Not "missing", but...
 

inputs/09/test.txt

@@ -0,0 +1 @@
+2333133121414131402

solutions/08.zig

@@ -251,11 +251,11 @@ fn divide(num: i32, denom: u32) i32 {
 
 const expect = std.testing.expect;
 
-// test "part_one" {
+test "part_one" {
-//     const part_one_response = try part_one(true);
+    const part_one_response = try part_one(true);
-//     print("DEBUG - part_one_response is {}\n", .{part_one_response});
+    print("DEBUG - part_one_response is {}\n", .{part_one_response});
-//     try expect(part_one_response == 14);
+    try expect(part_one_response == 14);
-// }
+}
 
 test "greatest_common_factor" {
     try expect(gcf(18, 27) == 9);

solutions/09.zig

@@ -0,0 +1,197 @@
+const std = @import("std");
+const print = std.debug.print;
+const util = @import("util.zig");
+
+pub fn main() !void {
+    const response = try part_two(false);
+    print("{}\n", .{response});
+}
+
+// There's _probably_ a super-clever way of doing this without actually expanding out the disk map, using dual-pointers
+// and keeping track of how much of each "chunk" has been consumed at each step - but, after thinking it over for a bit,
+// I couldn't figure out how to determine when to terminate (i.e. when enough blocks have been moved to make everything
+// contiguous), so I'm going for the naive approach - expand out the disk-map, then swap blocks one-by-one.
+//
+// ...huh, that executed _way_ faster than I feared.
+fn part_one(is_test_case: bool) !u64 {
+    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
+    defer _ = gpa.deinit();
+    const allocator = gpa.allocator();
+
+    const input_file = try util.getInputFile("09", is_test_case);
+    const data = try util.readAllInputWithAllocator(input_file, allocator);
+    defer allocator.free(data);
+
+    var disk = std.ArrayList(?u32).init(allocator);
+    defer disk.deinit();
+
+    var is_in_disk = true;
+    var disk_id: u32 = 0;
+    for (data) |c| {
+        if (c == '\n') {
+            continue;
+        }
+        const count = c - 48; // Why 48? Just because, I guess :shrug: ASCII for `2` is 50, and so...
+        for (0..count) |_| {
+            if (is_in_disk) {
+                try disk.append(disk_id);
+            } else {
+                try disk.append(null);
+            }
+        }
+        if (!is_in_disk) {
+            disk_id += 1;
+        }
+        is_in_disk = !is_in_disk;
+    }
+    print("DEBUG - actual disk is {any}\n", .{disk.items});
+
+    // We could _probably_ iterate over the disk from both ends and _just_ sum up the calculated values to get the
+    // answer (which would be more space-efficient) - but that way I'd have no chance of debugging if something goes
+    // wrong. I'll instead construct the _actually_ defragged disk, so that I can visually compare it with how it's
+    // supposed to look.
+    var left_pointer: usize = 0;
+    var right_pointer: usize = disk.items.len - 1;
+    var defragged_disk = std.ArrayList(u32).init(allocator);
+    defer defragged_disk.deinit();
+    while (true) : (left_pointer += 1) {
+        const value_in_original_disk = disk.items[left_pointer];
+        if (value_in_original_disk != null) {
+            try defragged_disk.append(value_in_original_disk.?);
+        } else {
+            // Position of left_pointer is inside a space in the original disk =>
+            // Move the right_pointer left until it finds a block, then insert that into the defragged disk
+            while (disk.items[right_pointer] == null) {
+                right_pointer -= 1;
+            }
+            try defragged_disk.append(disk.items[right_pointer].?);
+            right_pointer -= 1;
+        }
+        if (right_pointer == left_pointer) {
+            break;
+        }
+    }
+    print("DEBUG - defragged disk is {any}\n", .{defragged_disk.items});
+
+    var total: u64 = 0;
+    var idx: u32 = 0;
+    while (idx < defragged_disk.items.len) : (idx += 1) {
+        total += idx * defragged_disk.items[idx];
+    }
+
+    return total;
+}
+
+fn part_two(is_test_case: bool) !u64 {
+    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
+    defer _ = gpa.deinit();
+    const allocator = gpa.allocator();
+
+    const input_file = try util.getInputFile("09", is_test_case);
+    const data = try util.readAllInputWithAllocator(input_file, allocator);
+    defer allocator.free(data);
+
+    var disk = std.ArrayList(?u32).init(allocator);
+    defer disk.deinit();
+
+    var is_in_file = true;
+    var file_id: u32 = 0;
+    var file_start_indices = std.AutoHashMap(u32, usize).init(allocator);
+    defer file_start_indices.deinit();
+    // If we wanted, we could probably store "DiskInfo" in a struct and thus in a single HashMap, but...ehhh...
+    var file_sizes = std.AutoHashMap(u32, usize).init(allocator);
+    defer file_sizes.deinit();
+    for (data) |c| {
+        if (c == '\n') {
+            continue;
+        }
+        const count = c - 48; // Why 48? Just because, I guess :shrug: ASCII for `2` is 50, and so...
+        if (is_in_file) {
+            try file_start_indices.put(file_id, disk.items.len);
+        }
+        for (0..count) |_| {
+            if (is_in_file) {
+                try disk.append(file_id);
+                try file_sizes.put(file_id, count);
+            } else {
+                try disk.append(null);
+            }
+        }
+        if (!is_in_file) {
+            file_id += 1;
+        }
+        is_in_file = !is_in_file;
+    }
+    const highest_file_id = file_id;
+    print("DEBUG - actual disk is {any}\n", .{disk.items});
+    print("DEBUG - highest_file_id is {}\n", .{highest_file_id});
+    print("DEBUG - file_start_indices is:\n", .{});
+    var it_dsi = file_start_indices.iterator();
+    while (it_dsi.next()) |kv| {
+        print("  {}: {}\n", .{ kv.key_ptr.*, kv.value_ptr.* });
+    }
+    print("DEBUG - file_sizes is:\n", .{});
+    var it_ds = file_sizes.iterator();
+    while (it_ds.next()) |kv| {
+        print("  {}: {}\n", .{ kv.key_ptr.*, kv.value_ptr.* });
+    }
+
+    // Zig does not support decreasing ranges :'(
+    // for (highest_file_id..0) |file_id_to_move| {
+    for (0..highest_file_id) |file_i| {
+        const file_id_to_move = highest_file_id - file_i;
+        print("DEBUG - trying to move file with id {}\n", .{file_id_to_move});
+        const file_id_as_u32: u32 = @intCast(file_id_to_move);
+        const original_start_index = file_start_indices.get(file_id_as_u32).?;
+        const file_size = file_sizes.get(file_id_as_u32).?;
+        var left_pointer: usize = 0;
+        // I don't _think_ we can be smarter about jumping forward more than 1 at a time - at least, not without
+        // returning information from within `hasSpaceToMove`. In particular, we can't jump forward `file_size` -
+        // consider trying to move a file of size 4 into the following disk:
+        //    X..X....X
+        // if we jumped forward 4 from index 1, we'd land partway through the following 4-gap, and miss it.
+        while (left_pointer < original_start_index) : (left_pointer += 1) {
+            if (hasSpaceToMove(disk, left_pointer, file_size)) {
+                for (0..file_size) |i| {
+                    disk.items[left_pointer + i] = file_id_as_u32;
+                    disk.items[original_start_index + i] = null;
+                }
+                break;
+            }
+        }
+    }
+
+    print("DEBUG - defragged disk is {any}\n", .{disk.items});
+
+    var total: u64 = 0;
+    var idx: u32 = 0;
+    while (idx < disk.items.len) : (idx += 1) {
+        total += idx * (disk.items[idx] orelse 0);
+    }
+
+    return total;
+}
+
+fn hasSpaceToMove(disk: std.ArrayList(?u32), pointer: usize, file_size: usize) bool {
+    for (0..file_size) |i| {
+        if (disk.items[pointer + i] != null) {
+            return false;
+        }
+    }
+    print("DEBUG - found space to move - to location {}\n", .{pointer});
+    return true;
+}
+
+const expect = std.testing.expect;
+
+// test "part_one" {
+//     const part_one_response = try part_one(true);
+//     print("DEBUG - part_one_response is {}\n", .{part_one_response});
+//     try expect(part_one_response == 1928);
+// }
+
+test "part_two" {
+    const part_two_response = try part_two(true);
+    print("DEBUG - part_two_response is {}\n", .{part_two_response});
+    try expect(part_two_response == 2858);
+}