Code · Made a little simpler

NDJSON Log Stream Parser

Locate malformed log records by physical source line. Filter valid JSON values by level, severity or source text and export matches in source order with original number tokens and separators.

Free to useNo account neededProcessed on your device
Your workspaceProcessed on your device

Review a saved UTF-8 log. Filters select valid records from the file you supply.

Add this tool to your website

Paste this code into your page to add the tool and a visible ToolMellow credit link.

HTML embed code
Preview embed

NDJSON Log Stream Parser: how to use it

  1. Open a local UTF-8 log or paste one JSON value per line. Choose whether blank lines should count as errors.
  2. Enter an optional level or message substring, then select Parse log.
  3. Read the reported line and column of invalid records. Download matching records or a diagnostics report with counts and positions.

Limits and notes

Reads up to 8 MB of UTF-8 and 100,000 physical lines, with up to 250,000 UTF-16 code units per record. Accepts LF and CRLF, a leading BOM and a final record without a newline. Connects to no live log source. Shows the first 40 matches and 100 diagnostics.

How your privacy works →

NDJSON Log Stream Parser: common questions

Where does the level match come from?

The filter checks the first top-level string value of level or severity, using case-insensitive exact equality after trimming. Nested fields and numeric codes need a source text search instead.

Can I export good records when one is malformed?

Yes. The matching export includes only valid selected records and keeps their original tokens and separators. The diagnostics export gives counts and source positions.

Does the tool accept a blank line or a null record?

The parser counts blank lines and can flag them as errors. JSON null is a valid value and appears when it matches the chosen filters.

Guides