Now you are good to go !
- Goto the sudoku-solver repository
- Click on the Fork button in the upper right corner.
Introduce your self to GIT, make sure you use an email associated with your GitHub account.
git config --global user.name "John Doe"
git config --global user.email [email protected]
Make sure you clone the forked repository.
git clone https://github.com/<your username>/sudoku-solver.git
Switch sudoku-solver to your fork
git remote set-url origin https://github.com/<your username>/sudoku-solver.git
Setup sudoku-solver to be able to fetch from the master
git remote add upstream https://github.com/The-Codesis/sudoku-solver.git
Follow this tutorial carefully at pyimagesearch.com
Make sure you are on the same virtual environment in which you installed OpenCV
Tensorflow 2.0 comes with keras already, so you don't need to install it.
Follow this tutorial Install tensorflow 2.0.
Setting up Tensorflow 2.0 for GPU is a little tricky, so you can skip doing that to avoid difficulties in getting started.
Go to the folder where you cloned the project
cd sudoku-solver
Install the requirements from requirements.txt
- Switch to you virtual environment
pip install -r requirements.txt
When you add a new feature always create an issue first, this allows others to comment and give you tips. It will also help us keep track of what people are adding and with new releases helps us to write new release notes and give you credit for your work.
Secondly, always work in a branch, never work on the master branch. Keep your master branch in sync with the master of the official sudoku-solver repository. This makes the pull requests (you do want your work to be in the main branch right?) easier for us.
Finally, try to keep your branches focused on fixing/adding only one feature and try not to fall in the trap of doing a lot of things in a single branch. This will not only make it harder for us to process your pull request but makes it take longer before you can submit your pull request. Small pull requests are more likely to be looked at faster and pulled into the main branch faster.
Here is a simplified workflow on how to add a new feature:
Syncing your forked repository to the original repository is an important step before submitting any pull request to the original repository. So, it goes something like, you fetch from the original repository (Upstream Repository) to your working area (your local copy) and merge it with the fork's master branch and then you push it to your forked Github repository.
You can see all your remotes with git remote -v, if you don't have upstream set as https://github.com/The-Codesis/sudoku-solver.git, set up the original repository as your upstream.
git remote add upstream https://github.com/The-Codesis/sudoku-solver.git
By now, you have set the upstream as the original repository i.e. https://github.com/The-Codesis/sudoku-solver.git and the origin as the forked repository i.e. https://github.com/<your username>/sudoku-solver.git.
Now, fetch all of the changes from the original repository (note that commits to the original repository will be stored in a local branch called, upstream/master).
git fetch upstream
Since you want fork's master branch and the original's master branch should be synced, so make sure you are on fork's master branch and merge the changes from the upstream/master into your fork's local master branch (you may need to resolve the conflicts).
git checkout master
git merge upstream/master
At this point, your local branch is synced to the original repository's master branch. To update your forked Github repository, you need to push your changes to the forked Github repository.
git push origin master
git fetch only downloads the changes from the upstream and doesn't integrate those into your forked repository. You have to use an additional command git merge to do so. Alternatively, git pull not only downloads the changes but also integrate those into your forked repository.
Set up the original repository as your upstream.
git remote add upstream https://github.com/The-Codesis/sudoku-solver.git
Now, fetch and merge the changes from the original repository simultaneously.
git pull upstream master
Update your forked Github repository.
git push origin master
To summarize, see the section below.
Update your master (both locally and on GitHub)
git fetch upstream
git checkout master
git merge upstream/master
git push
Rebasing and merging are both used for integrating changes from one branch into another—but they do it in different ways.
Instead of
git merge upstream/master
you can also use
git rebase upstream/master
git merge will integrate the histories of the upstream branch and your master branch, giving you a branch structure that looks like:
Merging preserves history since, the existing branches are not changed in any way.Merging adds a new commit to the history.
git rebase will move the entire upstream branch on the tip of your master branch, giving you a branch structure that looks like:
Rebasing re-writes the history by creating brand new commits for each commit in the original branch.
git rebase is usually preferred over git merge because
- It results in a perfectly linear and readable history.
- It eliminates the unnecessary merge commits required by git merge.
- It can remove undesired commits or squash two or more commits into one.
A good practice is to call the branch in the form of GH- followed by the title of the issue. This makes it easier to find out the issue you are trying to solve and helps us to understand what is done in the branch. Calling a branch my-work is confusing. Names of a branch can not have a space, and should be replaced with a hyphen.
git checkout -b GH-issuenumber-title-of-issue
Do you work, and commit as you see fit. Make your commit messages helpful.
In general commit messages follow these guidelines
A brief commit message in one or two lines
Followed By
Detailed explanation of your commit
eg : If this commit fixes a bug in the code then describe the bug, what was causing it, your solution or
any other relevant thing one should know if someone else works on this project.
See this commit format
You can also refer this article on how to write commit messages.
To squash means commonly to compact multiple commits into one to make it more concise, readable and not to pollute main branch’s history. To achieve that, interactive mode of git rebase command used.
To squash the last n commits into one, run the following command:
git rebase -i HEAD~n
That will open up a text-editor with something similar to the following:
pick commit_1
pick commit_2
pick commit_3
...
pick commit_n
# Bunch of comments
Leave the first commit alone, and change the rest of the pick to squash or s.
pick commit_1
squash commit_2
squash commit_3
...
squash commit_n
# Bunch of comments
Save and exit the editor. Now, the editor will reopen right away suggesting to reword commit messages. Make the changes if you want. The commit message will include all the lines that are not starting with # character. Once again, save and exit the editor.
Your terminal now would show a success message including Successfully rebased and updated <branch name> and the git log would show a compacted history with only one commit.
If this is the first time pushing to GitHub you will need to extended command, otherwise, you can simply do a git push.
git push -u origin GH-issuenumber-title-of-issue
-
When finished create a pull request from your branch to the main sudoku-solver repository.
-
When making a pull request use
Closes #(issue_number)in the description of your PR, so that GitHub automatically associates your PR with that issue. -
In the description of your PR, mention what bug or feature this provides. Give any relevant information, that you think maintainer should know like :
-
Tests you did to check your new PR or code
-
Your approach to the problem
-
Any other relevant information that you think one might keep in mind for future
Make it clear,complete and simple so your PR gets merged easily !
Refer this Git Fork and Pull Request Cheat Sheet