Iterate on a tolerant XML parser

Published by Armin Ronacher

This credits the original publisher. Better Loop membership or a shared assessment is not implied.

The public work

Ronacher published a tolerant XML parser after earlier parsing approaches failed. He describes directing the initial implementation, reviewing it, requesting further fixes and an extensive test suite, and releasing the resulting library.

What to notice

Specify recovery behavior for malformed inputs, then turn failures found during use into tests and focused revisions.

Keep the context

The author reports a useful experiment, acknowledges possible remaining bugs, and does not endorse it as a production-ready parser. No independent rerun or measured human improvement is established.

AI use: Documented in the source.

The author explicitly describes using Claude for implementation, tests and revisions; the exact Claude model and host are unspecified.

A useful public example is not an assessment of a reader, a publisher or a Better Loop member.

Authored practice suggestion

Try the idea.
Check your own work.

Use material you are allowed to work with. This suggestion is preparation; it does not record a completed task or an improvement.

A prompt to adapt privately

Choose a small public parsing function and three malformed inputs. State the intended recovery behavior before asking for changes. Request a minimal patch and regression cases. Review whether each expected result matches the intended behavior, including one input the parser should reject.

A check to adapt

Each malformed input has an explicit expected outcome, and the proposed changes preserve valid-input behavior.

Open a related practice brief

Private practice needs no account. Your configured model provider may process the evidence you select.

See the full skill-to-sharing flow

Another way into the work.

Expose clipboard formats before debugging paste

Willison built a browser clipboard viewer while investigating rich-text paste. His public write-up links the code and conversation; the conversation shows requests to display available formats and then make long output easier to inspect.

Published bySimon Willison

Curated public exampleNotebook

Code & software

Check the expected values in generated tests

OpenAI publishes a multi-step test-writing notebook with a concrete Pig Latin function and generated pytest cases. Several printed expected values conflict with the function, making this a worked example of why generated tests themselves need review.

Published byOpenAI Cookbook

Curated public exampleCode