bert
6506af879a
Simplify y-hit iterator
2021-12-18 18:15:11 +01:00
bert
101ebee505
Use dedicated iterator instead of range
2021-12-18 18:15:11 +01:00
bert
cc81a7012b
Use math instead of binary search
2021-12-18 17:21:43 +01:00
bert
9c299f140c
Tighter bounds for the range of y
2021-12-18 15:01:47 +01:00
bert
ba1b7b693e
Use reusable parser wrapper more
2021-12-18 14:54:05 +01:00
bert
7d331f9131
Implement day 18 2021
2021-12-18 14:46:27 +01:00
bert
dfd8b2b985
Partially smart, partially brute-force day 17
2021-12-17 10:04:36 +01:00
bert
0f6167e90f
Merge common parts of parsers
2021-12-16 19:26:10 +01:00
bert
3e2a3f5206
Avoid allocating and tracking bits manually
2021-12-16 19:12:34 +01:00
bert
8cc2245492
Optimize hex parser
2021-12-16 18:46:19 +01:00
bert
fb167ef899
Implementation day 16
...
Not clean but it works.
2021-12-16 09:55:53 +01:00
bert
4240f8fc8c
Avoid queueing worse routes
2021-12-15 18:40:20 +01:00
bert
9a4ac427e0
Use Dijkstra instead of A*
...
The available distance heuristic doesn't save enough steps to offset its
overhead.
2021-12-15 18:29:08 +01:00
bert
c3929a56d8
Slow implementation for day 15
2021-12-15 07:54:14 +01:00
bert
9e37026f30
Remove unnecessary indexing
2021-12-14 19:54:45 +01:00
bert
a926243f0d
Improve "winning" logic
...
You don't need to check every possible win condition, just the one you
touched.
2021-12-14 19:51:57 +01:00
bert
a35ae82548
Replace unnecessary HashMap and allocations
2021-12-14 19:39:00 +01:00
bert
8a0b72f111
Implementation day 14 2021
2021-12-14 09:25:38 +01:00
bert
b5866b2d8a
Modify points in-place
...
Thanks to skius from libera##rust for the idea
2021-12-13 22:55:14 +01:00
bert
554683bc64
Actually apply nom
2021-12-13 22:20:02 +01:00
bert
1031a8cdbb
Avoid allocating the final buffer twice
2021-12-13 21:40:00 +01:00
bert
5c764f6627
Don't use silly hashsets
2021-12-13 21:34:48 +01:00
bert
ffe8d27469
Clean up day 13 a little.
2021-12-13 08:52:57 +01:00
bert
d471f170b3
Very dirty solution day 13
2021-12-13 08:50:00 +01:00
bert
07cbf6cf53
Fix upcoming clippy warning
2021-12-12 15:56:05 +01:00
bert
6e3252ce5a
Add formal benchmarking code
2021-12-12 15:33:52 +01:00
bert
0897e2e907
Implement day 12 part 2
2021-12-12 11:31:54 +01:00
bert
4d1fdd9cc0
Implement day 12 part 1
2021-12-12 11:09:14 +01:00
bert
84c160cf54
Simplify part 2
2021-12-11 17:20:22 +01:00
bert
440d454911
Implement day 11
2021-12-11 13:56:10 +01:00
bert
50cd6d8171
Implementation day 10 2021
2021-12-10 10:46:02 +01:00
bert
dde9c0adbf
Update to RC version of clap
...
Deriving a parser has become an opt-in feature so we need to enable
that.
2021-12-09 12:14:15 +01:00
bert
e0e1bc26e8
Simplify implementation day 8
...
The second part doesn't actually need to start the search at the low
points; just iterating everything and keeping track of visited spaces is
enough.
Now that the iterator is only used in part 1, we inline the iterator to
remove some overhead from the code.
2021-12-09 12:02:07 +01:00
bert
77ce31980b
Implement day 9
2021-12-09 11:56:29 +01:00
bert
8c78106846
Bit-optimize day 8
...
The state of a seven segment display can be stored in a byte after all.
Using NonZeroU8 makes the Options smaller too.
2021-12-08 13:52:08 +01:00
bert
22f767e8df
Implement day 08 2021
2021-12-08 13:31:33 +01:00
bert
3434966ac2
Just use the median, obviously
2021-12-07 13:22:43 +01:00
bert
c5f66fcc09
Avoid more allocations in day 7 part 1
2021-12-07 11:41:19 +01:00
bert
c02f1e11c5
Convert recursion to iteration
2021-12-07 11:00:26 +01:00
bert
d099614217
Improve part 2 with binary search
2021-12-07 10:40:34 +01:00
bert
766ee91719
Brute-force day 7 part 2
2021-12-07 10:12:01 +01:00
bert
2f3eb50a5b
Implement day 7 part 1
2021-12-07 09:45:58 +01:00
bert
b369f9d36a
Implement day 6 2021
2021-12-06 09:21:00 +01:00
bert
1433b0cdbe
Implement day 5
...
Nom is really nice and fast, why did I write parsers manually before.
2021-12-05 11:31:14 +01:00
bert
5e52da6e6b
Use iterator instead of range
2021-12-04 11:57:01 +01:00
bert
e50b812aed
Day 4: more efficiently ignore completed cards
2021-12-04 11:40:40 +01:00
bert
fdef10a78e
Less awkward line length hack
2021-12-04 11:28:32 +01:00
bert
392aefb32d
Less allocations in day 3 part 2
...
By working with binary numbers as integers rather than byte strings, we
don't need to allocate a Vec for each of them, reducing us to just the
allocations in the outer Vec.
2021-12-04 11:13:54 +01:00
bert
fb358be8f0
Implementation day 4 2021
2021-12-04 10:43:02 +01:00
bert
612d3ecb6b
Tricky solution day 03, could possibly be improved
2021-12-03 09:21:13 +01:00
bert