View Apache Avro container files with embedded schema display and record browsing.
Avro is a binary serialization format with no plain-text form: records are encoded as typed values, and the schema describing them is embedded in the container file's header, not kept in a separate .avsc file. This tool parses that header, displays the writer schema, and decodes the file's records so you can inspect an .avro file without installing avro-tools or a JVM. All parsing happens locally in your browser; the file is never uploaded. The surprise most users hit: an .avro file is not readable as text, and the record structure you see is determined by whatever schema the producer embedded.
A concrete input and expected output from the current implementation.
Input
An Avro container file whose embedded writer schema is {"type":"record","name":"Reading","fields":[{"name":"sensor","type":"string"},{"name":"value","type":"double"},{"name":"ts","type":"long"}]} and which contains two records: {"sensor":"a1","value":21.5,"ts":1720000000000} and {"sensor":"b2","value":19.75,"ts":1720000000100}. Expected output
A schema panel showing the schema JSON, and a record list: Record 1: {"sensor":"a1","value":21.5,"ts":1720000000000} / Record 2: {"sensor":"b2","value":19.75,"ts":1720000000100} The container header stores the schema, and each record is binary-encoded in the field order the schema declares, so the viewer decodes both records with that same schema and shows each field with the type the schema assigns: strings as text, doubles as decimals, longs as integers.
An Avro container file begins with the four-byte magic Obj\x01, followed by a metadata map that holds the writer schema under the key avro.schema, then a 16-byte sync marker. Data follows as a sequence of blocks: each block carries the number of objects it contains, the byte length of its encoded payload, and ends with a sync marker. The viewer reads this structure to locate records and validate the framing.
Each record is a stream of binary values whose meaning comes only from the schema: int and long fields use variable-length zig-zag encoding, strings are length-prefixed UTF-8, and union fields begin with an index byte naming the branch present. The viewer applies the embedded schema field by field, so doubles display as numbers, bytes fields become readable, and the exact union branch is resolved. If the file declares a codec such as deflate or snappy, blocks are decompressed before decoding.
Check the magic bytes first: a valid container starts with the four bytes 'Obj' followed by 0x01, visible in a hex editor. Also check whether the file declares a codec the viewer does not implement, and whether it is a container at all: raw Avro-encoded records without a container header cannot be opened, because they carry no schema.
No. The viewer parses the container entirely in your browser; nothing is uploaded to a server, so you do not need to worry about sharing sensitive records with a third party, and the tool works with no network access after the page loads.
The example and behavioral notes were checked against the browser implementation. Standards and primary references below define the relevant format, formula, or platform behavior.
Convert CSV data into structured JSON arrays or objects with automatic type detection.
Transform JSON arrays and objects into downloadable CSV spreadsheet format.
Convert YAML configuration files to JSON format with validation and formatting.
Parse XML documents and convert them to equivalent JSON structure with attributes.
Convert tab-separated values to comma-separated format with proper escaping.
Convert numbers between binary, octal, decimal, and hexadecimal bases.