Cleaning and validating data locally in your browser

1. Preserve a source copy

Work on a copy and record the expected format. A text transformation can be correct syntactically while still changing meaning. For sensitive material, close unrelated extensions and use a browser profile you trust.

Metyris tools process entered content in the current tab and do not send it to a Metyris application API. Ordinary page and asset requests still reach the infrastructure that delivers the site.

2. Validate JSON before reformatting

A formatter parses JSON, then writes the parsed value with readable indentation or minimal whitespace. It rejects comments, trailing commas, single-quoted strings and incomplete values because these are outside standard JSON.

Syntax validation is not schema validation. It cannot confirm that required fields exist or that values have the meaning an application expects. Large integers beyond JavaScript’s safe range and duplicate object keys can also change during parse-and-serialize workflows.

3. Clean text deliberately

Remove duplicate lines only after deciding whether case and surrounding spaces matter. Sort only records whose ordering is independent of meaning. Keep line breaks when paragraphs or row boundaries are significant.

Review a before-and-after sample before applying the same cleanup to a large dataset.

4. Use encoding and hashes for the right jobs

Base64 and URL encoding are reversible representations, not security controls. A SHA-2 hash is one-way and helps compare byte-for-byte integrity, but it does not encrypt content or prove the publisher’s identity by itself.

• Validate structure before formatting or minifying.
• Keep an original when numeric precision or duplicate keys may matter.
• Compare hashes only when the expected digest came through a trusted channel.