I've only just realised that there's a false-positive bug in the
latin.c solver framework.
It's designed to solve puzzles in which the solution is a latin square
but with some additional constraints provided by the individual
puzzle, and so during solving, it runs a mixture of its own standard
deduction functions that apply to any latin-square puzzle and extra
functions provided by the client puzzle to do deductions based on the
extra clues or constraints.
But what happens if the _last_ move in the solving process is
performed by one of the latin.c built-in methods, and it causes a
violation of the client puzzle's extra constraints? Nothing will ever
notice, and so the solver will report that the puzzle has a solution
when it actually has none.
An example is the Group game id 12i:m12b9a1zd9i6d10c3y2l11q4r . This
was reported by 'groupsolver -g' as being ambiguous. But if you look
at the two 'solutions' reported in the verbose diagnostics, one of
them is arrant nonsense: it has no identity element at all, and
therefore, it fails associativity all over the place. Actually that
puzzle _does_ have a unique solution.
This bug has been around for ages, and nobody has reported a problem.
For recursive solving, that's not much of a surprise, because it would
cause a spurious accusation of ambiguity, so that at generation time
some valid puzzles would be wrongly discarded, and you'd never see
them. But at non-recursive levels, I can't see a reason why this bug
_couldn't_ have led one of the games to present an actually impossible
puzzle believing it to be soluble.
Possibly this never came up because the other clients of latin.c are
more forgiving of this error in some way. For example, they might all
be very likely to use their extra clues early in the solving process,
so that the requirements are already baked in by the time the final
grid square is filled. I don't know!
Anyway. The fix is to introduce last-minute client-side validation:
whenever the centralised latin_solver thinks it's come up with a
filled grid, it should present it to a puzzle-specific validator
function and check that it's _really_ a legal solution.
This commit does the plumbing for all of that: it introduces the new
validator function as one of the many parameters to latin_solver, and
arranges to call it in an appropriate way during the solving process.
But all the per-puzzle validation functions are empty, for the moment.
Applications of the associativity rule were iterating over only n-1 of
the group elements, because I started the loops at 1 rather than 0. So
the solver was a bit underpowered, and would have had trouble with
some perfectly good unambiguous game ids such as 6i:2a5i4a6a1s .
(The latin.c system is very confusing for this, because for historical
reasons due to its ancestry in Solo, grid coordinates run from 0 to
n-1 inclusive, but grid _contents_ run from 1 to n, using 0 as the
'unknown' value. So there's a constant risk of getting confused as to
which kind of value you're dealing with.)
This solver weakness only affects puzzles in 'identity hidden' mode,
because in normal mode, the omitted row and column are those of the
group identity, and applications of associativity involving the
identity never generate anything interesting.
In identity-hidden mode, as soon as you find any table entry that
matches the element indexing its row or column (i.e. a product of
group elements of the form ab=a or ab=b), then you immediately know
which element is the group identity, and you can fill in the rest of
its row and column trivially.
But the Group solver was not actually able to do this deduction.
Proof: it couldn't solve the game id 4i:1d1d1d1, which is trivial if
you take this into account. (a^2=a, so a is the identity, and once you
fill in ax=xa=x for all x, the rest of the grid is forced.)
So I've added dedicated piece of logic to spot the group identity as
soon as anything in its row and column is filled in. Now that puzzle
can be solved.
(A thing that I _haven't_ done here is the higher-level deduction of
determining the identity by elimination. Any table entry that
_doesn't_ match either its row or column rules out both the row and
the column from being the group identity - so as soon as you have
enough table entries to rule out all but one element, you know the
identity must be the remaining one. At the moment, this is just doing
the simple version of the deduction where a single table entry tells
you what _is_ the identity.)
One slightly tricky part is that because this new identity inference
deduction fills in multiple grid entries at a time, it has to be
careful to avoid triggering an assertion failure if the puzzle is
inconsistent - before filling in each entry, it has to make sure the
value it wants to fill in has not been ruled out by something it
filled in already. Without that care, an insoluble puzzle can cause
the solver to crash - and worse, the same can happen on an insoluble
_branch_ of the guesswork-and-backtracking tree in Unreasonable mode.
In both cases, the answer is to make sure you detect the problem
before hitting the assertion, and return -1 for 'inconsistent puzzle'
if you spot it. Then any guesswork branch that ends up in this
situation is cleanly abandoned, and we move on to another one.
I still don't actually have a solver design, but a recent conversation
sent my brain in some more useful directions than I've come up with
before, so I might as well at least note it all down.
I noticed this during the bool trawl: both of these games have an
array of flags indicating which grid squares are immutable starting
clues, and copy it in every call to dup_game, which is completely
unnecessary because it doesn't change during play. So now each one
lives in a reference-counted structure, as per my usual practice in
similar cases elsewhere.
This is the main bulk of this boolification work, but although it's
making the largest actual change, it should also be the least
disruptive to anyone interacting with this code base downstream of me,
because it doesn't modify any interface between modules: all the
inter-module APIs were updated one by one in the previous commits.
This just cleans up the code within each individual source file to use
bool in place of int where I think that makes things clearer.
This commit removes the old #defines of TRUE and FALSE from puzzles.h,
and does a mechanical search-and-replace throughout the code to
replace them with the C99 standard lowercase spellings.
encode_params, validate_params and new_desc now take a bool parameter;
fetch_preset, can_format_as_text_now and timing_state all return bool;
and the data fields is_timed, wants_statusbar and can_* are all bool.
All of those were previously typed as int, but semantically boolean.
This commit changes the API declarations in puzzles.h, updates all the
games to match (including the unfinisheds), and updates the developer
docs as well.
This function gives the front end a way to find out what keys the back
end requires; and as such it is mostly useful for ports without a
keyboard. It is based on changes originally found in Chris Boyle's
Android port, though some modifications were needed to make it more
flexible.
It's silly to have every puzzle using latin.c separately specify in
its .R file the list of additional modules that latin.c depends on, or
for that matter to have them all have to separately know how to adjust
that for the STANDALONE_SOLVER mode of latin.c.
So I've centralised a new pair of definitions into the core Recipe
file, called LATIN and LATIN_SOLVER, and now a client of latin.c need
only ask for that to get all the necessary dependencies too.
Also, while I'm here, I've moved the non-puzzle-specific 'latincheck'
test program out of unequal.R into the central Recipe.
I went through all the char * parameters and return values I could see
in puzzles.h by eye and spotted ones that surely ought to have been
const all along.
This allows me to use different types for the mutable, dynamically
allocated string value in a C_STRING control and the fixed constant
list of option names in a C_CHOICES.
Now midend.c directly tests the returned pointer for equality to this
value, instead of checking whether it's the empty string.
A minor effect of this is that games may now return a dynamically
allocated empty string from interpret_move() and treat it as just
another legal move description. But I don't expect anyone to be
perverse enough to actually do that! The main purpose is that it
avoids returning a string literal from a function whose return type is
a pointer to _non-const_ char, i.e. we are now one step closer to
being able to make this code base clean under -Wwrite-strings.
To do this, I've completely replaced the API between mid-end and front
end, so any downstream front end maintainers will have to do some
rewriting of their own (sorry). I've done the necessary work in all
five of the front ends I keep in-tree here - Windows, GTK, OS X,
Javascript/Emscripten, and Java/NestedVM - and I've done it in various
different styles (as each front end found most convenient), so that
should provide a variety of sample code to show downstreams how, if
they should need it.
I've left in the old puzzle back-end API function to return a flat
list of presets, so for the moment, all the puzzle backends are
unchanged apart from an extra null pointer appearing in their
top-level game structure. In a future commit I'll actually use the new
feature in a puzzle; perhaps in the further future it might make sense
to migrate all the puzzles to the new API and stop providing back ends
with two alternative ways of doing things, but this seemed like enough
upheaval for one day.
When filling in a cyclic subgroup or one of its cosets, I've often
found I wanted to set an entire diagonal to the same thing at once
(usually SW-NE, but the other way round too in non-abelian groups),
and it's a pain having to do that to each square individually.
Restricting multiple selection to diagonals makes it easy to get the
selection I really wanted.
When you drag group elements around, previous dividers are meant to
dissolve whenever the same two elements are no longer on each side of
it. One case in which this didn't happen was that of dragging an
element from the left of a divider to the far right column - the
divider became invisible, but would then startlingly reappear if you
drag that element back to the left of whatever it was left of before.
[originally from svn r10051]
to it giving each game's "internal" name (as seen in the source file,
.R etc) and also a brief description of the game. The idea of the
latter is that it should be usable as a comment field in .desktop
files and similar.
[originally from svn r9858]
puzzle backend function which ought to have it, and propagate those
consts through to per-puzzle subroutines as needed.
I've recently had to do that to a few specific parameters which were
being misused by particular puzzles (r9657, r9830), which suggests
that it's probably a good idea to do the whole lot pre-emptively
before the next such problem shows up.
[originally from svn r9832]
[r9657 == 3b250baa02a7332510685948bf17576c397b8ceb]
[r9830 == 0b93de904a98f119b1a95d3a53029f1ed4bfb9b3]
new_desc. Oddities in the 'make test' output brought to my attention
that a few puzzles have been modifying their input game_params for
various reasons; they shouldn't do that, because that's the
game_params held permanently by the midend and it will affect
subsequent game generations if they modify it. So now those arguments
are const, and all the games which previously modified their
game_params now take a copy and modify that instead.
[originally from svn r9830]
in the 'unfinished' directory for a while, and has now been finished
up thanks to James Harvey putting in some effort and galvanising me to
put in the rest. This is 'Pearl', an implementation of Nikoli's 'Masyu'.
The code in Loopy that generates a random loop along grid edges to use
as the puzzle solution has been abstracted out into loopgen.[ch] so
that Pearl can use it for its puzzle solutions too. I've also
introduced a new utility module called 'tdq' (for 'to-do queue').
[originally from svn r9379]
midend_status(), and given it three return codes for win, (permanent)
loss and game-still-in-play. Depending on what the front end wants to
use it for, it may find any or all of these three states worth
distinguishing from each other.
(I suppose a further enhancement might be to add _non_-permanent loss
as a fourth distinct status, to describe situations in which you can't
play further without pressing Undo but doing so is not completely
pointless. That might reasonably include dead-end situations in Same
Game and Pegs, and blown-self-up situations in Mines and Inertia.
However, I haven't done this at present.)
[originally from svn r9179]
thereafter read. Most of these changes are just removal of pointless
stuff or trivial reorganisations; one change is actually substantive,
and fixes a bug in Keen's clue selection (the variable 'bad' was
unreferenced not because I shouldn't have set it, but because I
_should_ have referenced it!).
[originally from svn r9164]
state is in a solved position, and a midend function wrapping it.
(Or, at least, a situation in which further play is pointless. The
point is, given that game state, would it be a good idea for a front
end that does that sort of thing to proactively provide the option to
start a fresh game?)
[originally from svn r9140]
elements to toggle thick lines in the grid. Helps to delineate
subgroups and cosets, so it's easier to remember what you can
legitimately fill in by associativity.
(I should really stop fiddling with this game's UI; it's far too silly.)
[originally from svn r9084]
much easier to keep track of things if, once you've identified a
cyclic subgroup, you can move it into a contiguous correctly ordered
block.
[originally from svn r9075]
- move critical correctness checks out of diagnostic ifdefs (ahem)
- move declarations to before conditionally compiled code (we don't
build in C99 mode round here)
- tidy up an unsightly blank line while I'm here.
[originally from svn r8969]
immediately obvious which element of the group is the identity - at
least two elements including the identity have their rows and
columns completely blanked.
[originally from svn r8810]
you are given a partially specified Cayley table of a small finite
group, and must fill in all the missing entries using both Sudoku-
style deductions (minus the square block constraint) and the group
axioms. I've just thrown it together in about five hours by cloning-
and-hacking from Keen, as much as anything else to demonstrate that
the new latin.c interface really does make it extremely easy to
write new Latin square puzzles.
It's not really _unfinished_, as such, but it is just too esoteric
(not to mention difficult) for me to feel entirely comfortable with
adding it to the main puzzle collection. I can't bring myself to
throw it away, though, and who knows - perhaps a university maths
department might find it a useful teaching tool :-)
[originally from svn r8800]
