codecrafters-io/build-your-own-git

Definition for the git challenge.

build-your-own-git's Issues

Tester says "internal error" on stage 5 for a user error

Logs:

Dec 19 12:49:00 148e5315b37068 vector [stage-5] Creating some files & directories
Dec 19 12:49:00 148e5315b37068 vector [stage-5] $ ./your_git.sh write-tree
Dec 19 12:49:00 148e5315b37068 vector [your_program] aec2d24a22463ac63c7ec9684903f7f5ba292c79
Dec 19 12:49:00 148e5315b37068 vector [stage-5] Reading file at .git/objects/ae/c2d24a22463ac63c7ec9684903f7f5ba292c79
Dec 19 12:49:00 148e5315b37068 vector Error [stage-5] CodeCrafters internal error. Error reading .git/objects/ae/c2d24a22463ac63c7ec9684903f7f5ba292c79: open /tmp/worktree814524593/.git/objects/ae/c2d24a22463ac63c7ec9684903f7f5ba292c79: no such file or directory
Dec 19 12:49:00 148e5315b37068 vector [stage-5] Test failed

Here's our code that's supposed to filter user vs. internal error:

	if err == os.ErrNotExist {
		return fmt.Errorf("Did you write the tree object? Did not find a file in .git/objects/<first 2 chars of sha>/<remaining chars of sha>")
	} else if err != nil {
		return fmt.Errorf("CodeCrafters internal error. Error reading %v: %v", relativePath, err)
	}

Looks like this specific instance of "no such file or directory" isn't covered by os.ErrNotExist?

Files generated by CMake are present in directory for stage 5 (write tree)

I think the test cases are wrong for C++ since I had to hardcode ignoring all of the files related to CMake. Maybe these files don't exist for whatever language you wrote them in?

Thanks to @ljdannull for highlighting this!

"Read tree" stage tests are flaky

the tests are flaky - they fail and if i run it again with no changes, they pass
[stage-4] Running ./your_git.sh ls-tree --name-only 8eeb4ae8872e3393d1530df8f536f56255b69000
[your_program] doo
[your_program] dumpty
[your_program] Traceback (most recent call last):
[your_program]   File "<frozen runpy>", line 198, in _run_module_as_main
[your_program]   File "<frozen runpy>", line 88, in _run_code
[your_program]   File "/app/app/main.py", line 105, in <module>
[your_program]     main()
[your_program]   File "/app/app/main.py", line 83, in main
[your_program]     mode, path = entry.split(b" ")
[your_program]     ^^^^^^^^^^
[your_program] ValueError: too many values to unpack (expected 2)
[stage-4] Expected "doo\ndumpty\nhumpty\n" as stdout, got: "doo\ndumpty\n"
[stage-4] Test failed
[stage-4] Running tests for Stage #4: Read a tree object
[stage-4] Running ./your_git.sh init
[your_program] Initialized git directory
[stage-4] Writing a tree to git storage..
[stage-4] Running ./your_git.sh ls-tree --name-only 0e0e083fd1cd81bb3df04d53e726e2a6a763e6d0
[your_program] doo
[your_program] dumpty
[your_program] yikes
[your_program] Traceback (most recent call last):
[your_program]   File "<frozen runpy>", line 198, in _run_module_as_main
[your_program]   File "<frozen runpy>", line 88, in _run_code
[your_program]   File "/app/app/main.py", line 105, in <module>
[your_program]     main()
[your_program]   File "/app/app/main.py", line 83, in main
[your_program]     mode, path = entry.split(b" ")
[your_program]     ^^^^^^^^^^
[your_program] ValueError: not enough values to unpack (expected 2, got 1)
[stage-4] Test passed.

Looks like we aren't testing the exit code?