Work through a debugging task

Explain a changed API response

Start here when a response parses but the consuming application behaves differently. Work on invented or sanitized records.

  1. Why a duplicate JSON key can change an order status

    Resolve conflicting keys with the producer.

  2. Keep a large order ID intact when processing JSON

    Preserve identifier digits before parsing.

  3. Find a string where your JSON Schema expects an integer

    Validate field types against the agreed schema.

  4. Compare JSON values without confusing types

    Compare the corrected response against the expected one.

When to stop

Keep the original response, the agreed contract and the final diff. Re-run the consuming application; schema success alone does not establish business correctness.

Decide whether a conversion preserves meaning

Choose the branch matching your source format; these are alternative checks, not three mandatory conversions.

  1. Keep XML attributes and empty repeated elements in JSON

    XML input: inspect attributes, empty elements and mixed content.

  2. Check strings and booleans when converting JSON to YAML

    JSON to YAML: check strings, booleans and ordered arrays.

  3. Infer optional fields without erasing null from JSON samples

    JSON to TypeScript: distinguish missing fields from explicit null.

When to stop

Save the result only after checking types with the parser or compiler used by the consuming application. Keep the original when the mapping discards document structure.

Prepare evidence for a reproducible bug report

Remove secrets without deleting the detail needed to reproduce an error. Query parameters can carry credentials too.

  1. Prepare a log excerpt for a support ticket

    Mask recognized credentials, then add application-specific fields and review the entire log.

  2. Find changed and repeated URL query parameters

    If the bug concerns a link, compare parameter values and order using invented URLs.

When to stop

Include tool URL, browser, a minimal synthetic input, steps, expected output and actual output. Verify the reduced example still fails before sending it; do not attach production credentials.

Open the report template

Individual cases and background guides

2026-09-30 · Worked example

Keep XML attributes and empty repeated elements in JSON

A simple XML converter can drop attributes on leaf nodes or overwrite an empty first value when a second element has the same name. Inspect both before using the output as test data.

Read more →
2026-09-30 · Worked example

Find changed and repeated URL query parameters

A search link behaves differently after being copied between environments. Looking only at a dictionary of query keys can hide repeated values, while comparing raw text can overstate differences in space encoding.

Read more →
2026-09-30 · Worked example

Compare JSON values without confusing types

Two responses look similar, but a numeric field became a string or an empty array became an object. These changes can break consumers even when both documents parse.

Read more →
2026-09-30 · Worked example

Why a duplicate JSON key can change an order status

A response contains two status fields. A parser may retain only one, so formatting with a parse-and-stringify shortcut can hide the conflict.

Read more →
2026-09-30 · Worked example

Keep a large order ID intact when processing JSON

An identifier exceeds JavaScript’s safe integer range. Parsing it as a Number can change its digits before you notice.

Read more →
2026-09-30 · Worked example

Find a string where your JSON Schema expects an integer

The response is valid JSON, yet age is "30" rather than 30. Syntax formatting cannot detect this contract mismatch.

Read more →
2026-09-30 · Worked example

Infer optional fields without erasing null from JSON samples

Two records have different shapes. Declaring every observed property as required produces a type that rejects the second record.

Read more →
2026-09-30 · Worked example

Check strings and booleans when converting JSON to YAML

Configuration values such as "0012" and "yes" can be misread if a conversion changes their types or a downstream parser uses different YAML rules.

Read more →
2026-09-30 · Worked example

Prepare a log excerpt for a support ticket

A useful request trace may contain a token, password or customer identifier. Removing the whole trace loses the request ID needed to investigate.

Read more →
Updated 2026-10-09

JSON formatting: syntax, types and safe identifiers

Separate syntax, duplicate keys, identifier precision and schema failures.

Read more →
Updated 2026-10-09

Base64 text conversion: UTF-8 examples and common errors

Reproduce a UTF-8 roundtrip and distinguish text errors from binary payloads.

Read more →
Updated 2026-10-09

UUIDs: choose a version and check v4 output

Check v4 output shape, estimate collision risk and preserve access controls.

Read more →
Updated 2026-10-09

Regex testing: boundaries, flags and near misses

Test extraction versus whole-field matching with positive and negative inputs.

Read more →
Updated 2026-10-09

Text hashing: reproduce SHA-256 and investigate differences

Compare exact UTF-8 input, line endings and trusted digest references.

Read more →