Clean TicTacToe games up #1311
Open
Clean TicTacToe games up #1311
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The max game length of an ultimate TTT relies on the max length of its meta TTT game. Encapsulate this instead of re-implementing TicTacToe::MaxGameLenth in UltimateTTT::MaxGameLenth. Signed-off-by: odanrc <odanrc@yahoo.com.br>
The max game length of an phantom TTT relies on the max length of its underlying TTT game. Encapsulate this instead of re- implementing TicTacToe::MaxGameLenth in PhantomTTT::MaxGameLenth. Signed-off-by: odanrc <odanrc@yahoo.com.br>
This helps abstract away how the board is implemented. Signed-off-by: odanrc <odanrc@yahoo.com.br>
For now this is just a stump, but the idea is to eventually grow this class to make it encapsulate more functionality that is common to game boards. Signed-off-by: odanrc <odanrc@yahoo.com.br>
kNumCells and kNumSubgames have been used interchangeably. This was only correct because they coincidentally match. Fix the misuses to reference the correct constant. Signed-off-by: odanrc <odanrc@yahoo.com.br>
This function allows searching for a line on the board. Its check were hardcoded to a 3x3 board. Make it more generic by automating the process. Signed-off-by: odanrc <odanrc@yahoo.com.br>
Increase the semantic of relying on the size of the board by creating a respective function. This improves code readability. It also reduces/centralizes the use of the kNumCells constant, making it easier to reach the goal of eventually turning it into a variable. Signed-off-by: odanrc <odanrc@yahoo.com.br>
The dimensions of a grid board should be known upon instantiation, and they must be provided by the caller in order to make the class more generic. For now we use a vector as the underlying container, but we should use inplace_vector once supported. Signed-off-by: odanrc <odanrc@yahoo.com.br>
Use consts instead of hardcoding values and make the code easier to follow through. Signed-off-by: odanrc <odanrc@yahoo.com.br>
This information may be needed by users. Signed-off-by: odanrc <odanrc@yahoo.com.br>
We want to generalize the board size, so use accessor methods to get the number of actions instead. Signed-off-by: odanrc <odanrc@yahoo.com.br>
The use of kNumCells as a constant forces a TTT configuration. Replace these hardcoded uses by dynamic values to allow other TTT sizes. Signed-off-by: odanrc <odanrc@yahoo.com.br>
The use of kNumCells as a constant forces a TTT configuration. Replace these hardcoded uses by dynamic values to allow other TTT sizes. As a side effect, UltimateTTTState is now a friend of TTTState, since the ultimate TTT is a TTT. This makes it easier to share members between these highly related classes. Signed-off-by: odanrc <odanrc@yahoo.com.br>
Instead of accessing the global constants, use the board-specific parameters so that we can eventually replace the constants by variables. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Make it compatible with TTT, so that it can call TTT´s respective functions instead of duplicating them. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Call the parent function instead of reimplementing it, to avoid accidentally introducing bugs and to reduce the dependency on the constants. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Rewrite this function to avoid reimplementing TTT::ToString. To do so we create a Display class that is able to manipulate strings that represent objects in a 2D space. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Use the functions that extract the dimensions of the board instead of relying on global constants. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Generalize the code to support a different number of rows and columns. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
This will allow us to convert the board to string in any class that uses a board, which is the case of PhantomTTT. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Create a function that returns what are the available legal actions for a given board. Notice that this is likely a workaround, and it is possible that what we need is to keep track of the view of each player as a state instead of a board. For the time being this works to reduce code duplication. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
This is a tiny step to generalize the GridBoard class to make it easier to be used by other games. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
This will help reduce the use of CellState. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
This will allow avoid using CellState directly. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
Prefer using a more generic framework. Signed-off-by: Daniel R. Carvalho <odanrc@yahoo.com.br>
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
|
I am confused by this section of the CLA check: ' Uncheck it and re-create the offending commit, ´. What does re-create mean in this context? Force push, or delete the PR and push again? |
|
Hi @odanrc, Thanks! Can I double-check that you have followed the link at the top of this screenshot and then, once you've signed the CLA, followed the link at the bottom of this screenshot (After "New contributors": ...)
You can find that page by clicking on the "failed invocation" link in this message. |
|
Hey, @lanctot . It looks like my CLA was voided because I didn't make clear my position in the company. Once I get a confirmation of acceptance I will get back to this PR. |
|
Hi @odanrc , I had to fix a breakage in the wheel tests. Could you pull changes from master and push the merge commit? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Hey!
I wanted to contribute to another project in my spare time, and this one seemed interesting; however, the website suggested checking out tic tace toe (TTT) for perfect information, which was a great suggestion, but then recommended poker and other games for references of imperfect information. I believe this is counterproductive: TTT is much simpler, and if one uses it to learn how to set up a perfect info game, then it should be a smaller step to follow through with phantom TTT, specially given that the poker code is not easily readable for a newcomer.
The problem is that phantom TTT is not a commonly known variant - can we even find a description longer than a paragraph online? - so I decided to clean it up a bit as I understood its code. I also noticed that TTT and variant's code embedded multiple assumptions about TTT's behavior. My goal with this set of commits (and PRs) is to generate abstractions and make TTT (and variants) more generic and easier to understand. These are the results of these commits:
For future PRs I plan to apply these same steps to other games, making implementing games more trivial and standardized.
Possible long-term goals:
Regarding this PR, it gives an overview of how the TTT and variants look like after the changes. I will likely break it down into multiple PRs, but I want to know what is the reviewers' preference first. I tried to make every commit as self-contained and small as possible, so they should be quite easy to review individually (and I recommend doing so). Do you prefer PRs per commit (or group of very similar commits), per feature, or per number of lines?
Note: On the bright side, the tests caught bugs quite well; on the dark side, though, I often had no idea where to look at to figure out what was breaking, specially when having to deal with python's dynamic typeness. I managed to get through by dumping variables' contents until I found which txt was breaking. Is there a document/tutorial describing how to find what/where things broke? If so, then maybe it should be made more visible in the Developer Guide section. If not, I suggest creating one.