View Apache Arrow IPC (Feather) files with schema inspection and tabular data browsing.
This tool opens a .arrow or .feather file, reads its trailing footer and Flatbuffers-encoded schema, and lists every column with its Arrow type, nullability, and null count before rendering the record batches as a scrollable table. It exists for the moment you have an Arrow file and no Python environment handy to check what's in it. The most common surprise: Feather V1, the original pre-Arrow Feather layout, is a different on-disk format from Feather V2, which is just the Arrow IPC file format under another name, and V1 files may not open here.
A concrete input and expected output from the current implementation.
Input
A Feather (Arrow IPC) file with schema {id: int64, name: string, score: float64} and 3 rows: (1, 'Alice', 92.5), (2, 'Bob', 88.0), (3, 'Cara', null) Expected output
Schema panel: id -> int64, not null; name -> string (utf8), not null; score -> float64, nullable, 1 null. Data table (3 rows): 1/Alice/92.5, 2/Bob/88.0, 3/Cara/(null).
The viewer reports exactly the three declared columns with their Arrow types and flags the single null in score, then lists the three record-batch rows in file order.
The tool reads the trailing footer of the Arrow IPC file, a Flatbuffers-encoded block listing the schema and the byte offset of each record batch. It uses those offsets to locate batches directly instead of scanning the file start to finish, so the schema panel populates even before any row data is decoded.
Arrow stores each column as contiguous buffers: a validity bitmap plus one or more value buffers, per the Arrow columnar memory layout. To build the table view, the viewer decodes each column's buffers according to its declared type and null bitmap, then interleaves columns row by row for on-screen display.
No. Parquet is a separate columnar file format with its own footer and page layout, distinct from Arrow IPC. Convert it to Arrow or Feather first, for example with pyarrow, or use a Parquet-specific viewer instead.
No, the file is read and parsed locally in your browser and is not sent to a server. That matters if the data is sensitive, or if you want to inspect a file while offline.
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.