Use `black` to break long strings

[vil02]

As soon as black will support breaking long strings:

• revert changes in create_puzzle from feat: break long strings in _PUZZLE #139,
• use black to ensure that the output does not contain long lines (cf. Ensure that the output of create does not contain long lines #111).

Some related issues and PRs:

• Remove _split_long_str and use textwrap.wrap instead #150 and style: use textwrap.wrap instead of _split_long_str #151.

[DhanashreePetare]

Hello @vil02
I would like to work on the issue.

[vil02]

Hi @DhanashreePetare - go ahead! Let me know if you would need some support.

[DhanashreePetare]

Sure @vil02

…

On Wed, Oct 1, 2025 at 2:20 AM Piotr Idzik ***@***.***> wrote: *vil02* left a comment (vil02/puzzle_generator#140) <#140 (comment)> Hi @DhanashreePetare <https://github.com/DhanashreePetare> - go ahead! Let me know if you would need some support. — Reply to this email directly, view it on GitHub <#140 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BDXPXJFJ3WPCRPKAHCCVVT33VLUJHAVCNFSM6AAAAACH6AMORCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGNJTG43DANJZGY> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[cclauss]

Use ruff format instead. Faster because it is written in Rust.

[mathurojus]

Hi @vil02,

I’d like to help solve this issue.

Approach:

I will first check if the latest version of black supports automatic breaking of long string literals. If not, I’ll test ruff format as suggested, since it reportedly handles this better and is faster.

Once confirmed, I’ll remove all manual string-breaking logic (like textwrap.wrap or _split_long_str) from relevant modules, including create_puzzle.

I’ll update the formatting setup (such as .pre-commit-config.yaml or workflow configs) to use the chosen formatter for consistent handling.

The affected files will be reformatted so that outputs like _PUZZLE do not contain excessively long lines.

Finally, I’ll validate the results and open a PR referencing this issue and related discussions.

Please let me know if I have your permission to proceed, or if you have any additional preferences regarding the formatter/tooling!

Thanks!

[vil02]

@DhanashreePetare: what is the progress? Please note the comment from @mathurojus.

[vil02]

Use ruff format instead. Faster because it is written in Rust.

@cclauss: correct me if I am wrong: black is a tiny package in comparison to ruff. Since puzzle-generator depends on some formatting package, I would prefer to keep this dependency lightweight.

Furthermore, time spend on formatting the output of the puzzle-generator is not crucial.

[cclauss]

Please. Ruff is several times faster because it is a (smaller) compiled executable, while Black is interpreted.

% time uvx ruff --help

0.02s user 0.03s system 7% cpu 0.613 total

% time uvx black --help

0.28s user 0.12s system 6% cpu 6.086 total

Multiply that slowness (installing + running) across black + bandit + flake8 + bugbear + pytest-style + isort + pylint and you will see why projects that run pre-commit favor Ruff. Also add up all the separate config files because flake8 robustly refuses to use pyproject.toml

Plus, ruff is already installed for linting, so it does not need to be reinstalled for formatting.

[vil02]

@cclauss: what about systems not having rust? Tools like pylint, flake8 etc. are only the dev-dependencies - the users do not need to install them.

[cclauss]

You don’t need Rust locally because Ruff is pre-compiled. All you need is pip.

[mathurojus]

Hi @vil02,

I've completed my research on whether black or ruff format now support automatic breaking of long string literals:

Black (version 25.9.0):

• Does NOT support automatic breaking of long string literals
• According to the Black documentation, Black will only split strings that are already broken up into segments
• Quote from docs: "Black does not split string literals that are not already split"

Ruff Format:

• Also does NOT support automatic breaking of long string literals
• Ruff format maintains compatibility with Black's formatting decisions
• According to Ruff's documentation, it follows Black's approach for string handling

Conclusion:
Neither formatter currently supports this feature, so the manual string-breaking logic using textwrap.wrap (from PR #151) should remain in place. The current implementation is still the appropriate solution for ensuring that the output does not contain excessively long lines.

Recommendation:
Close this issue as "won't fix" (or similar) with a note that we'll revisit if/when Black or Ruff add support for automatic string breaking in the future.

Let me know if you'd like me to prepare documentation of these findings or add a code comment explaining why we use manual string breaking!

[cclauss]

I like the approach of continue to use textwrap and then run ruff format to clean up.