Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

New ranching scenario #835

Merged
merged 2 commits into from
Nov 14, 2022
Merged
Show file tree
Hide file tree
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
New ranching scenario
  • Loading branch information
kostmo committed Nov 14, 2022
commit 75c32ae741f1faacbe9e39a7f7924028bd8a4357
1 change: 1 addition & 0 deletions data/scenarios/Challenges/00-ORDER.txt
Original file line number Diff line number Diff line change
Expand Up @@ -3,3 +3,4 @@ teleport.yaml
2048.yaml
hanoi.yaml
Mazes
Ranching
1 change: 1 addition & 0 deletions data/scenarios/Challenges/Ranching/00-ORDER.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
gated-paddock.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,255 @@
// Algorithm:
// ----------
// Maintain current direction until a wall is encountered.
// Then enter "wall-following mode".
// This mode presumes the wall is not a loop.
// Wall-following mode exploits recursion to keep track of how many left turns were made
// and then unwinds them again by ensuring each is paired with a right turn.
// Once the recursion is fully unwound, the robot proceeds along its original direction
// (though it may now be laterally displaced).
//
// (If it was a loop, then an "oriented breadcrumb" would need to be left.
// The breadcrumb is oriented in case a single-width passage is backtracked
// along the opposite wall.)

/** A "gate" is walkable, so we need to supplement the "blocked" check with this function.
Since fences are "unwalkable", they do not need to be mentioned in this function.
*/
def isFenced =
s <- scan forward;
return (
case s
(\_. false)
(\x. x == "gate")
);
end;

def isBlockedOrFenced =
b <- blocked;
f <- isFenced;
return (b || f);
end;

// Returns true if we've already placed two
// breadcrumbs on a given tile, false otherwise.
def leaveBreadcrumbs =

let bc1 = "fresh breadcrumb" in
let bc2 = "treaded breadcrumb" in

wasTraversedOnce <- ishere bc1;
if wasTraversedOnce {
_crumb <- grab;
make bc2;
place bc2;
return false;
} {
wasTraversedTwice <- ishere bc2;
if wasTraversedTwice {
return true;
} {
// Make sure nothing's in the way before we place
// our breadcrumb:
x <- scan down;
case x return (\y.
// If we're on a water tile, get rid of
// it with our special "drilling" recipe
if (y == "water") {
drill down;
// Nothing will remain on the ground.
// after making the "steam" via
// the drilling recipe.
} {
grab;
return ();
};
);

make bc1;
place bc1;
return false;
};
};
end;

def goForwardToPatrol = \wasBlocked.
b <- isBlockedOrFenced;
if b {
turn left;
goForwardToPatrol true;
turn right;
goForwardToPatrol false;
} {
if wasBlocked {
isLoop <- leaveBreadcrumbs;
if isLoop {
fail "loop";
} {};
} {};
move;
};
end;

/**
There should only be one place in the
code where an exception is thrown: that is,
if a treaded breadcrumb is encountered.
*/
def checkIsEnclosedInner =
try {
goForwardToPatrol false;
// Water is the outer boundary
hasWater <- ishere "water";
if hasWater {
return false;
} {
checkIsEnclosedInner;
};
} {
return true;
};
end;

def checkIsEnclosed =

// The "evaporator" drill is used
// to clear water tiles.
let specialDrill = "evaporator" in
create specialDrill;
install self specialDrill;

// NOTE: System robots can walk on water
// so we only need this if we want to
// demo the algorithm with a player robot.
// create "boat";
// install self "boat";

checkIsEnclosedInner;
end;

def boolToInt = \b. if (b) {return 1} {return 0}; end;

def countAdjacentBlockages =

turn left;
b1 <- isBlockedOrFenced;
c1 <- boolToInt b1;

turn left;
b2 <- isBlockedOrFenced;
c2 <- boolToInt b2;

turn left;
b3 <- isBlockedOrFenced;
c3 <- boolToInt b3;

turn left;
b4 <- isBlockedOrFenced;
c4 <- boolToInt b4;

return $ c1 + c2 + c3 + c4;
end;

// Step forward, observing left and right.
def observeLeftAndRight =
move;
turn left;
amBlockedLeft <- isBlockedOrFenced;
val1 <- boolToInt amBlockedLeft;

turn back;
amBlockedRight <- isBlockedOrFenced;
val2 <- boolToInt amBlockedRight;

turn right;
move;
return $ val1 + val2;
end;


/** If the four cardinal directions have at most
one blockage, then there will exist an orientation
where both that direction and its opposite direction
are clear.
So we can step that direction, check to the left and
right of us, then step in the opposite direction
and do the same. This allows us to check the 4
blocks that touch the corners of the center block.
*/
def countDiagonalBlockages =
// First, orient to the clear front-to-back path
amBlocked <- isBlockedOrFenced;
if amBlocked {turn left;} {};

// Second, step to both sides
fwdCount <- observeLeftAndRight;
backCount <- observeLeftAndRight;
return $ fwdCount + backCount;
end;

def isStandingOnBridge =
onFence <- ishere "fence";
onGate <- ishere "gate";
if (onFence || onGate) {
adjCount <- countAdjacentBlockages;
if (adjCount > 1) {
return true;
} {
diagCount <- countDiagonalBlockages;
return $ (adjCount + diagCount) > 1;
};
} {return false};
end;

def getValForSheepIndex = \predicateCmd. \i.
try {
// This will throw an exception if
// the sheep has already drowned.
r <- robotnumbered i;
didSucceed <- as r {predicateCmd};

boolToInt didSucceed;
} {
return 0;
}
end;

/**
There are 3 sheep.
They have indices 1, 2, 3.
(The base has index 0).

THIS DOES NOT WORK!
*/
def countSheepWithRecursive = \predicateCmd. \i.

if (i > 0) {
val <- getValForSheepIndex predicateCmd i;
recursiveCount <- countSheepWithRecursive predicateCmd $ i - 1;
return $ val + recursiveCount;
} {
return 0;
}
end;


def countSheepWith = \predicateCmd.

val1 <- getValForSheepIndex predicateCmd 1;
val2 <- getValForSheepIndex predicateCmd 2;
val3 <- getValForSheepIndex predicateCmd 3;
return $ val1 + val2 + val3;

end;


justFilledGap <- as base {
isStandingOnBridge;
};

if (justFilledGap) {
enclosedCount <- countSheepWith checkIsEnclosed;
return $ enclosedCount >= 1;
} {
return false;
}
Loading